The Metre of Old Saxon Poetry: The Remaking of Alliterative Tradition (Medieval Literature)

THE METRE OF OLD SAXON POETRY The Remaking of Alliterative Tradition Seiichi Suzuki

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page i

The Metre of Old Saxon Poetry: The remaking of alliterative tradition

This is a comprehensive study of Old Saxon metre, with a particular emphasis on the Heliand, an alliterative epic of the Gospel story and the most extensive work of Old Germanic poetry. Through a detailed description of the metre in its own terms and a systematic comparison with the Old English alliterative tradition, especially Beowulf, this book shows how the Heliand poet introduced a wealth of metrical innovations, reorganising the traditional scheme underneath an overarching principle of artistic design. After setting out the literary, metrical, linguistic, and practical bases, the author moves on to consider the Heliand metre in depth, looking at its synchronic and diachronic, formal and functional, descriptive and interpretive properties; he identifies a set of metrical types, determines their distributional constraints, and establishes their paradigmatic and syntagmatic organisation. He also deals with resolution and alliteration, and the composition of hypermetric verses and lines; the book’s conclusions suggest the wider implications of the study as a whole. The volume is completed with two appendices, on the scansion of foreign names, and on the metre of the Old Saxon Genesis. SEIICHI SUZUKI is Professor of Old Germanic Studies, Kansai Gaidai University, Japan.

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page ii

To Winfred P. Lehmann, my mentor and model in metrics and linguistics

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page iii

The Metre of Old Saxon Poetry: The remaking of alliterative tradition

Seiichi Suzuki

D. S. BREWER

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page iv

© Seiichi Suzuki 2004 All Rights Reserved. Except as permitted under current legislation no part of this work may be photocopied, stored in a retrieval system, published, performed in public, adapted, broadcast, transmitted, recorded or reproduced in any form or by any means, without the prior permission of the copyright owner First published 2004 D. S. Brewer, Cambridge ISBN 1 84384 014 6

D. S. Brewer is an imprint of Boydell & Brewer Ltd PO Box 9, Woodbridge, Suffolk IP12 3DF, UK and of Boydell & Brewer Inc. PO Box 41026, Rochester, NY 14604-4126, USA website: www.boydellandbrewer.com A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Suzuki, Seiichi, 1956The metre of Old Saxon poetry : the remaking of alliterative tradition / Seiichi Suzuki. p. cm. Includes bibliographical references and index. ISBN 1-84384-014-6 (Hardback : alk.paper) 1. Old Saxon language–Versification. 2. Heliand. I. Title. PF3995.S89 2004 829.1⬘009–dc22 2003025773

This publication is printed on acid-free paper Printed in Great Britain by St Edmundsbury Press Ltd, Bury St Edmunds, Suffolk

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page v

Contents Tables Preface Symbols and Abbreviations

viii xv xviii

1. Bases of Old Saxon metre: an introduction 1.1. The literary basis: the Heliand 1.2. The metrical basis: the Old English alliterative tradition 1.3. The phonological basis: a weakening of stress 1.3.1. Restoration of syncopated vowels 1.3.2. Svarabhakti vowels 1.3.3. West Germanic Gemination 1.3.4. Summary 1.4. The practical basis for scanning Old Saxon verse: notations, foreign names, and the quað he phrase

1 1 7 11 12 13 22 23

2. Metrical types and positions: levelling and reorganisation 2.1. Type A1 (/ ⫻ / ⫻) 2.1.1. The configuration PS#Px 2.1.2. The configurations PSx#Px and PS#xPx 2.1.3. The configurations PXx#Px, Pxx#Px, PX#xPx, and Px#xPx 2.1.4. The configuration Pxx#Px 2.1.5. The configurations Px#Px, PX#Px, P#xPx, and P#xx . . . Px 2.1.6. The configuration PxSx 2.1.7. The graded patterning of the first drop of type A1 2.1.7.1. A synchronic perspective 2.1.7.2. A diachronic perspective 2.1.8. The configuration PS#px 2.2. Type A2 (/ ⫻ / \) 2.2.1. The configurations Px . . . PS, Px . . . P#x, and Px . . . PX 2.2.2. The configuration Px . . . Pxx 2.3. Type A3 (⫻ ⫻ / ⫻)

29 29 29 30 33 36 40 46 46 46 53 61 66 66 73 76

23

MOSP-Prelims.qxd

vi

5/11/04

2:51 PM

Page vi

Contents

2.4. Type B1 (⫻ / ⫻ /) 2.4.1. The first drop of type B1 2.4.2. The second drop of type B1 2.4.2.1. The range of variation on the number of syllables 2.4.2.2. The configurations x . . . PS#P and x . . . PS#px 2.4.2.3. The configurations x . . . Px#xP, x . . . Pxx#xP, x . . . PS#xP, x . . . Psx#(x)P, x . . . PSx#(x)P, and x . . . PXx#(x)P 2.4.2.4. The configurations x . . . P#x#P, x . . . P#xx#P, x . . . P#x#x#P, x . . . P#x#xP, and x . . . Px#xx#P 2.5. Type B3 (⫻ ⫻ ⫻ /) 2.6. Type C (⫻ / / ⫻) 2.6.1. The variation on the first drop of type C 2.6.2. The variation on the second lift of type C 2.6.3. The variation on the second drop of type C 2.7. Type D (/ / ⫻ ⫻) 2.8. Type D* (/ ⫻ / ⫻ ⫻) 2.9. Type E (/ \ ⫻ /) 2.10. Type E* (/ ⫻ \ ⫻ /) 2.11. Anacrusis 2.11.1. Anacrusis as a full metrical position: a synchronic account 2.11.2. Anacrusis reconfigured: a diachronic account 2.12. The paradigmatic and syntagmatic organisation of the metre: a recapitulation 2.12.1. The system of metrical types 2.12.2. The blurred distinction between normal and heavy drops 2.12.3. The distinction between verse-initial and non-initial normal drops 2.12.3.1. Quantitative and qualitative distinctions 2.12.3.2. Correlation with verse distribution 2.12.4. On the treatment of the disyllabic ending -xx# 2.12.5. On the treatment of quasi compounds

175 175 181 182 182 184 187 190

3. Resolution and alliteration: repatterning and reconstitution 3.1. Resolution 3.1.1. Suspension of resolution 3.1.1.1. The second lift of subtype A1s (PS#px) 3.1.1.2. The second lift of type C (x . . . P#px; x . . . Psx) 3.1.1.3. The second lift of subtype D1 (Psxx) 3.1.1.4. The first drop of subtype D2a (P#P#px; P#Psx)

194 194 194 194 197 199 200

80 80 87 87 89 94 105 111 113 113 117 123 125 136 145 156 160 161 172

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page vii

Contents 3.1.1.5. The first drop of type E (PsxP) 3.1.1.6. Implementation versus suspension of resolution: a synchronic and diachronic account 3.1.2. Loss of Kaluza’s Law and its consequences: a further dimension on resolution 3.1.3. Apparent conflict between implementation and suspension of resolution: on the scansion of the configurations Px . . . Psx and Px . . . P#px 3.1.4. Resolution and variants of metrical types 3.2. Alliteration 3.2.1. The metrical basis of alliteration 3.2.1.1. Introductory remarks 3.2.1.2. Metrical types and alliteration 3.2.1.3. Anacrusis and alliteration 3.2.1.4. Resolution and alliteration 3.2.2. The lexical and syntactic basis of alliteration 3.2.3. The phonological basis of alliteration

vii 201 203 211 217 223 235 235 235 237 239 243 255 287

4. Hypermetric verses and lines: diversification and restructuring 4.1. The composition of hypermetric verses: a synchronic perspective 4.2. Diversification and restructuring of hypermetric verses: a diachronic perspective

295 295

5. The remaking of alliterative tradition: gradation and harmonisation

330

Appendix 1. Foreign names Appendix 2. The metre of the Old Saxon Genesis

345 353

References Index to the scansion of the Heliand Index to the scansion of the Old Saxon Genesis Index of authors Index of subjects Index of verses cited for discussion or exemplification

366 370 478 485 487 495

323

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page viii

Tables* Table 1.1. The organisation of metrical units in Old English metre Table 1.2. The inventory of metrical types and subtypes in Old English metre Table 1.3. The system of metrical types in Old English metre Table 1.4. Distribution of breaking, smoothing, and vowel epenthesis in Old English dialects and Old Saxon Table 2.1. Distribution of PS#Px (including Psx#Px) Table 2.2. Distribution of PS#Px (excluding Psx#Px) Table 2.3. Distribution of PSx#Px Table 2.4. Distribution of PS#xPx Table 2.5. Distribution of Psx#Px Table 2.6. Distribution of PXx#Px (PXx# ⫽ non-compound) Table 2.7. Distribution of Pxx#Px (Pxx# ⫽ non-compound) Table 2.8. Distribution of PXx#xPx Table 2.9. Distribution of Pxx#xPx (excluding Pxxx#Px) Table 2.10. Distribution of PSx#xPx Table 2.11. Distribution of PX#xPx Table 2.12. Distribution of Px#xPx and P#xxPx Table 2.13. Distribution of Px#xPx (#x ⫽ prefix) Table 2.14. Distribution of Px#xPx (#x ⫽ independent word) Table 2.15. Distribution of P#xxPx (#x ⫽ prefix) Table 2.16. Distribution of P#xxPx (#x ⫽ independent word) Table 2.17. Distribution of Pxx#Px Table 2.18. Distribution of Pxx#xPx Table 2.19. Distribution of Pxxx#Px Table 2.20. Distribution of PXxx#Px Table 2.21. Distribution of Px#Px Table 2.22. Distribution of PX#Px Table 2.23. Distribution of Px#Px (excluding PX#Px) Table 2.24. Distribution of PX#Px and Px#Px in Beowulf Table 2.25. Distribution of P#xPx (#x ⫽ prefix or independent word) Table 2.26. Distribution of P#xPx (#x ⫽ prefix) Table 2.27. Distribution of P#xPx (#x ⫽ independent word) Table 2.28. Distribution of P#xxPx, P#xxxPx, P#xxxxPx, P#xxxxxPx, and P#xxxxxxPx Table 2.29. Distribution of PxSx (including two instances of PxxSx) Table 2.30. List of major variants of type A1 * Unless otherwise noted, all tables concern the Heliand.

8 9 9 19 30 30 31 31 32 33 34 34 34 35 35 36 37 38 38 39 39 40 40 40 40 41 42 44 44 44 44 45 47 47

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page ix

Tables Table 2.31. Cumulative effects of the three parameters Table 2.32. Distribution of Px#xxPx, PX#xxPx, Pxx#xxPx, and PXx#xxPx, compared with Px#xPx, PX#xPx, Pxx#xPx, and PXx#xPx, respectively Table 2.33. Linguistic realisations of the first drop of type A1, with reference to the medial and word-final syllables Table 2.34. Linguistic realisations of the first drop of type A1 in Beowulf, with reference to the medial and word-final syllables Table 2.35. Options for reorganisation of the Heliand metre with respect to the association of medial and word-final syllables with the first drop of type A1 Table 2.36. The categorisation of the first drop of types A1 and A2a in the traditional metre Table 2.37. Distribution of PS#px Table 2.38. Distribution of Px . . . PS and Px . . . P#P Table 2.39. Distribution patterns of P#x . . . PS and P#x . . . Px Table 2.40. Distribution of Px . . . P#x (#x ⫽ non-lexical word) Table 2.41. Distribution of Px#PX Table 2.42. Distribution of Px . . . Pxx Table 2.43. Distribution pattern of types A3, B1, and C according to the number of unstressed syllables before the first lift Table 2.44. Distribution pattern of types A3, B1, and C in the a-verse according to the number of unstressed syllables before the first lift Table 2.45. Distribution pattern of types A3, B1, and C in the a-verse in Beowulf according to the number of unstressed syllables before the first lift Table 2.46. Distribution of the varying number of syllables in the first drop of type B1 for the a-verse with double alliteration, the a-verse with single alliteration, and the b-verse Table 2.47. Verse distribution according to the number of syllables in the first drop of type B1 Table 2.48. Distribution of x . . . P#xP according to the number of syllables in the first drop Table 2.49. Distribution of x . . . Px#P according to the number of syllables in the first drop Table 2.50. Distribution of x . . . PX#P according to the number of syllables in the first drop Table 2.51. Distribution of x . . . Pxx#P according to the number of syllables in the first drop Table 2.52. Distribution of x . . . PXx#P according to the number of syllables in the first drop Table 2.53. Distribution of x . . . Pxx#P according to the number of syllables in the first drop Table 2.54. Distribution of x . . . PxS according to the number of syllables in the first drop Table 2.55. Distribution of x . . . Px#xP according to the number of syllables in the first drop

ix 51

52 54 55

55 56 61 66 68 68 70 74 77 78

79

81 81 82 83 83 84 84 85 85 86

MOSP-Prelims.qxd

x

5/11/04

2:51 PM

Page x

Tables

Table 2.56. Distribution of the varying number of syllables in the second drop of type B1 for the a-verse with double alliteration, the a-verse with single alliteration, and the b-verse Table 2.57. Verse distribution according to the number of syllables in the second drop of type B1 Table 2.58. Distribution of x . . . PS#P (to be modified as Table 2.62 below) Table 2.59. Distribution of x . . . PS#px and PS#px Table 2.60. Distribution of x . . . PS#px according to the number of syllables in initial position Table 2.61. Distribution of type A1 with anacrusis according to the number of anacrustic syllables Table 2.62. Distribution of x . . . PS#P (revised) Table 2.63. Distribution of PS#xP and x . . . PS#xP Table 2.64. Distribution of Psx#(x)P and x . . . Psx#(x)P Table 2.65. Distribution of PSx#(x)P and x . . . PSx#(x)P Table 2.66. Distribution of PXx#(x)P and x . . . PXx#(x)P Table 2.67. Distribution of type B1 variants x . . . Px#P, x . . . PX#P, x . . . Pxx#P, and x . . . PXx#(x)P Table 2.68. Distribution of the varying number of verse-initial unstressed syllables in x . . . PXx#(x)P Table 2.69. Distribution of the varying number of verse-initial unstressed syllables in x . . . Pxx#(x)P and x . . . Pxx#(x)P Table 2.70. Distribution of the varying number of verse-initial unstressed syllables in x . . . PSx#(x)P Table 2.71. Parameter-based evaluation of various posttonic forms as candidates for realising the second drop of type B1 Table 2.72. Distribution of type B1 variants according to the association of the second drop with the variety of word-internal constituents Table 2.73. Distribution of x . . . P#x#P Table 2.74. Distribution of x . . . P#xx#P (or x . . . P#Px#P), x . . . P#x#xP (or x . . . P#P#xP), and x . . . P#x#x#P (or x . . . P#P#x#P) Table 2.75. Distribution of the varying number of syllables in the first drop of type C for the a-verse with double alliteration, the a-verse with single alliteration, and the b-verse Table 2.76. Verse distribution according to the number of syllables in the first drop of type C Table 2.77. Distribution of major variants of type C Table 2.78. Distribution of major variants of type C in Beowulf Table 2.79. Distribution of the type C variant ending in -xx# Table 2.80. Distribution of major variants of type D Table 2.81. Distribution of major variants of types D1, D2a, and D2b in Beowulf Table 2.82. Distribution of major variants of type D* (subtypes D*1, D*2a, and D*2b) Table 2.83. Proportion of subtypes D*1, D*2a, D*2b to subtypes D1, D2a, D2b, respectively

87 87 90 93 93 94 94 96 96 97 98 99 100 100 101 104 105 105 106

114 114 118 122 124 127 134 138 140

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page xi

Tables Table 2.84. The varying number of syllables in the first drop of type D* Table 2.85. Distribution of type D* according to the number of syllables in the first drop Table 2.86. The varying number of syllables in the first drop of subtype D* in Beowulf Table 2.87. Distribution of subtype D* according to the number of syllables in the first drop in Beowulf Table 2.88 (⫽ Table 2.42). Distribution of Px . . . Pxx Table 2.89. Distribution of major variants of type E Table 2.90. Distribution of major variants of type E in Beowulf Table 2.91. Distribution of the varying number of syllables in the second drop of type E Table 2.92. Distribution of the varying number of syllables in the second drop of type E in Beowulf Table 2.93. Distribution of type E* Table 2.94. Distribution of major variants of types A1, A2, D, D*, E, and E* with and without anacrusis Table 2.95. Distribution of major variants of types A1, A2a, A2b, D1, D2a, and D2b, and subtypes D*1, D*2a, and D*2b, with and without anacrusis in Beowulf Table 2.96. Distribution of types A1, A2, D, E, and E* with or without anacrusis Table 2.97. Verse distribution according to the number of syllables used as anacrusis Table 2.98. The system of metrical types in the Heliand metre Table 2.99. Frequency of types A2, A3, and B3 in the Heliand and Beowulf Table 2.100. Frequency of types D* and E* in the Heliand and their corresponding configurations in Beowulf Table 2.101. Distribution of types A1, B1, and C according to the numbers of syllables in the first drop Table 2.102. Distribution of the varying number of syllables in the non-final normal drop in relation to the distinction between the a-verse and the b-verse Table 2.103. Distribution of x . . . PSx#(x)P (PSx# ⫽ quasi compound; type E with anacrusis) and x . . . PXx#(x)P (PXx# ⫽ non-compound; scannable as type B1) Table 2.104. Distribution of type C verses containing true compounds, quasi compounds, and non-compounds Table 2.105. Distribution of type D verses containing true compounds, quasi compounds, and non-compounds Table 2.106. Distribution of type E verses containing true compounds, quasi compounds, and non-compounds Table 3.1. Implementation versus suspension of resolution Table 3.2. Determining susceptibility to suspension of resolution by the three subrules Table 3.3. Implementation versus suspension of resolution in Beowulf Table 3.4. Distribution of Px . . . Psx, Px . . . PS, Px . . . PSx, Px . . . P#px, Px . . . P#P, and Px . . . P#Px

xi 141 142 142 142 144 150 151 155 155 157 168

169 170 171 177 179 180 183

185

191 192 192 193 203 205 210 218

MOSP-Prelims.qxd

xii

5/11/04

2:51 PM

Page xii

Tables

Table 3.5. Distribution of Px . . . Psx, Px . . . PS (type A2), Px . . . PSx (subtype D*2a), Px . . . P#px, Px . . . P#P (type A2), and Px . . . P#Px (subtype D*2a) according to the number of syllables in the first drop Table 3.6. Varying frequency of resolution on the first and second lifts relative to metrical types and their major realisation variants Table 3.7. Varying frequency of resolution on the first and second lifts relative to metrical types and their major realisation variants in Beowulf Table 3.8. Distribution of resolution on the second lift of type A2 in relation to the morpholexical status of the words appearing in the second lift and drop Table 3.9. Distribution of resolution on the second lift of type A2b in Beowulf in relation to the morpholexical status of the words appearing in the second lift and drop Table 3.10. Overall distribution of single and double alliteration in the Heliand and Beowulf Table 3.11. Distribution of alliterative patterns in the a-verse in the Heliand and Beowulf Table 3.12. Proportion of single and double alliteration in the minimal variants of major metrical types in the a-verse Table 3.13. Proportion of single and double alliteration of the minimal variants of major metrical types in the a-verse in Beowulf Table 3.14. Distribution pattern of anacrustic verses in the Heliand and Beowulf Table 3.15. Distribution of anacrustic verses in the a-verse in the Heliand and Beowulf Table 3.16. Alliterative patterning of the major variants of type A1 with and without anacrusis Table 3.17. Alliterative patterning of the major variants of type D with and without anacrusis Table 3.18. Alliterative patterning of the major variants of type D* with and without anacrusis Table 3.19. Alliterative patterning of the major variants of type E with and without anacrusis Table 3.20. Alliterative patterning of the unresolved and resolved variants of type A1 Table 3.21. Alliterative patterning of the unresolved and resolved variants of type B1 Table 3.22. Alliterative patterning of the unresolved and resolved variants of type C Table 3.23. Alliterative patterning of the unresolved and resolved variants of type D Table 3.24. Alliterative patterning of the unresolved and resolved variants of type D* Table 3.25. Alliterative patterning of the unresolved and resolved variants of type E Table 3.26. Alliterative patterning of the unresolved and resolved variants of type A1 in Beowulf

218 224 225

233

234 236 236 237 238 239 239 240 241 242 242 244 245 246 246 247 247 249

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page xiii

Tables Table 3.27. Alliterative patterning of the unresolved and resolved variants of type B1 in Beowulf Table 3.28. Alliterative patterning of the unresolved and resolved variants of type C in Beowulf Table 3.29. Alliterative patterning of the unresolved and resolved variants of types D1, D2a, and D2b in Beowulf Table 3.30. Alliterative patterning of the unresolved and resolved variants of subtypes D*1, D*2a, and D*2b in Beowulf Table 3.31. Alliterative patterning of the unresolved and resolved variants of type E in Beowulf Table 3.32. Scansion of two-word concatenations as metrical configurations Table 3.33. Alliterative patterns of normal (non-heavy) a-verses Table 3.34. Alliterative patterns of normal (non-heavy) b-verses Table 3.35. Distribution of alliterative patterns in relation to three constituent types Table 3.36. Distribution of alliterative patterns involving the identical CVTable 3.37. Distribution of alliterative patterns involving the identical CV- (revised) Table 3.38. Distribution of alliterative patterns involving the identical CV- in Beowulf Table 3.39. Distribution of vowel alliteration involving the identical vowels Table 3.40. Distribution of vowel alliteration involving the identical vowels in Beowulf Table 4.1. Distribution of types A1 and A1A1 according to the number of syllables in the first drop Table 4.2. Distribution of type A1A1 according to the number of syllables in the first and second drops Table 4.3. Distribution of the numbers of syllables used as anacrusis in type A1A1 in comparison with type A1 Table 4.4. The varying number of anacrustic syllables in type A1 (or of the initial part before the first lift of type A3A1), compared with that of type A1 in the a-verse and the b-verse Table 4.5. The varying number of syllables before the lift in type A3, compared with that before the first lift of type A3A1 (or with that of anacrusis of type A1) Table 4.6. Pairing of anacrustic type A1 (or type A3A1) b-verses with a-verses according to the distinction between hypermetric and normal verses Table 4.7. Pairing of anacrustic type A1 b-verses with a-verses according to the distinction between hypermetric and normal verses Table 4.8. The varying number of anacrustic syllables in type A1 (or of the initial part before the first lift of type B3A1), compared with that of type A1 in the b-verse and that of type B3 Table 4.9. Pairing of anacrustic type A1 (or type B3A1) b-verses with a-verses according to the distinction between hypermetric and normal verses Table 4.10. The varying number of anacrustic syllables in type A1 (or of the initial part before the first lift of type CA1), compared with that of type A1 in the b-verse

xiii 250 250 251 252 253 257 261 261 271 289 290 291 291 292 298 298 299

308

309 310 310

311 312

313

MOSP-Prelims.qxd

xiv

5/11/04

2:51 PM

Page xiv

Tables

Table 4.11. The varying number of anacrustic syllables in type A1 (or of the initial part before the first lift of type CA1), compared with that of the first drop of type C in the b-verse Table 4.12. Pairing of anacrustic type A1 (or type CA1) b-verses with a-verses according to the distinction between hypermetric and normal verses Table 4.13. Composition of hypermetric a-verses/b-verses with respect to the identity of constituent normal verses, and their occurrences in number Table 4.14. The inventory of well-formed compositions of hypermetric verses with respect to the identity of constituent normal verses Table 4.15. The minimal clusters of hypermetric verses with the candidate verse nb as a point of reference Table 4.16. Distribution of two major hypermetric verse types Table 4.17. Distribution of two major hypermetric verse types in Beowulf Table 4.18. Distribution of type A1A1 according to the number of syllables in the first and second drops in Beowulf, compared with that in the Heliand Table 4.19. Distribution of type A3A1 according to the number of verse-initial unstressed syllables in Beowulf, compared with that in the Heliand Table A1.1. Distribution of Iudeono and Iudeo in combination with the word forms P and Px Table A2.1. List of major variants of type A1 in the Old Saxon Genesis Table A2.2. Distribution of major variants of type C in the Old Saxon Genesis Table A2.3. Distribution of major variants of type D in the Old Saxon Genesis Table A2.4. Distribution of major variants of type D* (subtypes D*1, D*2a, and D*2b) in the Old Saxon Genesis Table A2.5. Proportion of subtypes D*1, D*2a, D*2b to subtypes D1, D2a, D2b, respectively, in the Old Saxon Genesis Table A2.6. Distribution of major variants of type E in the Old Saxon Genesis Table A2.7. Distribution of types A1, A2, D, D*, and E with or without anacrusis in the Old Saxon Genesis Table A2.8. Distribution of major variants of types A1, A2, D, D*, and E, with and without anacrusis in the Old Saxon Genesis Table A2.9. The varying number of anacrustic syllables in the normal verse in the Old Saxon Genesis Table A2.10. The varying number of anacrustic syllables in the normal verse Table A2.11. Distribution of alliterative patterns involving the identical CV- in the Old Saxon Genesis Table A2.12. Distribution of vowel alliteration involving the identical vowels in the Old Saxon Genesis

313

314

314 315 322 324 324

327

328 349 354 355 357 358 358 359 360 361 362 362 364 364

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page xv

Preface The Metre of Old Saxon Poetry: The Remaking of Alliterative Tradition is a comprehensive study of Old Saxon metre both on its synchronic and diachronic dimensions, with emphasis on the metre of the Heliand – an alliterative epic of the Gospel story and the most extensive work of Old Germanic poetry with nearly six thousand lines. Following my monograph on the metre of Beowulf (Suzuki 1996a), this volume constitutes the second part of my on-going project on Old Germanic metre. As with my previous book, this one is grounded on the premise that the metre constitutes an open-ended system of rules and representations, constructed largely on a prototype basis of categorisation, and invariably susceptible to restructuring and reconfiguration due to its inherent structural ambiguity, imperfection, and indeterminacy, particularly on the periphery of the system. For more on general theoretical underpinnings and metatheoretical orientations, the reader may profit from referring to Preface and Chapter 1 of my previous work. Following the overviews offered in Chapter 1 on the literary, metrical, linguistic, and practical bases for this study, the next three central chapters provide an in-depth, comprehensive, and explanatory account of the Heliand metre in the remaking, examined from complementary perspectives in unity – synchronic and diachronic, formal and functional, descriptive and interpretive, extrapersonal from the outside and intrapersonal from the inside. Chapter 2 is concerned with the fundamental issues of Old Saxon metre, namely, those of identifying a set of metrical types, characterising their internal structures and realisation variants, determining their distributional constraints, and exploring their paradigmatic and syntagmatic organisation in the system. Chapter 3 deals with resolution and alliteration, two major metrical devices for maximising prominence of the verse. Chapter 4 is devoted to the composition of the hypermetric verse and line. Finally, Chapter 5 provides conclusions and some thoughts on their implications outside the metre, followed by two appendixes, on the scansion of foreign names (Appendix 1) and on the metre of the Old Saxon Genesis (Appendix 2). Through a detailed synchronic account made on its own intrinsic terms and a subsequent systematic comparison with the metre of Beowulf, which is assumed to be a representative work of traditional versecraft (section 1.2), I shall substantiate that a wealth of metrical innovations were introduced to the Heliand and then explore the motivation and mechanism of such reorganisations from an intrapersonal perspective of the metrical innovator, the Heliand poet at work. In verse-making, he was challenged by two major traditional forces, linguistic and metrical. On the one hand, in composing Old Saxon verse, he had no choice but to work under radically changed prosodic conditions characteristic of Old Saxon, namely, a weakening of stress and its pervasive consequences on phonological structure (section 1.3). On the other hand, in versifying in acceptable traditional form, he had to follow the central rules and conventions of the inherited versecraft as fully as possible. Without placing

MOSP-Prelims.qxd

xvi

5/11/04

2:51 PM

Page xvi

Preface

himself completely at the mercy of these extrapersonal forces, however, the poet acted at his own will and made the best of the options that he had at his disposal. Specifically, by re-enacting the presumable reasonings and decisions that the poet would have made in specific metrical contexts, consciously or unconsciously, to improve on and remake the inherited metre, I shall bring to light his concerted artistic strive for gradation and harmonisation realised in a wide spectrum of innovations. As the underlying motivation for these manifold transformations will thus be identified an overarching principle of artistic design, which may be characterised as the principle of balance and harmony, and which will then be suggested to be parallel to the one responsible for the highly symmetrical organisation of the epic’s narrative structure, recognised independently in previous literary studies (Chapter 5). In light of the minimal departures from tradition and maximal effects of harmonisation that the poet accomplished in his verse-making, we may infer that the Heliand poet would have had a firm working knowledge of traditional metre, a daring spirit of innovation, and a profound sense of balance and symmetry, in conjunction with an admirably sensitive awareness of linguistic differences between Old English and Old Saxon down to their details and subtleties. The major thrust of this book is a structural, systematic thinking that is pursued vigorously both along the synchronic-descriptive and diachronic-comparative dimensions of the metre, conceptualised in its own immanent terms as an integral whole. While determinedly concerned with fully capturing most conspicuous regularities, I am no less interested throughout this work in uncovering patterned variations and subtle gradations among residues and details seemingly exceptional to such grand generalisations, and in accounting for these minute findings, metre-internally in relation to one another as well as to the whole system, metre-externally with reference to linguistic structure, and ultimately against the background of the poet’s remaking of the metre with greater balance and harmony. The book thus differs, on the one hand, from dominant philological scholarship, in which the Heliand metre has often been depreciated as deviant and degenerate against the canonical model (typically based on the Beowulf metre) and metrical emendations have therewith been made here and there rather gratuitously to render it closer to the standard; and on the other hand, this book distances itself from much of current theorising on the metre whereby a fragment of metrical features, all too often isolated from the whole system, is deployed as evidence for or against particular linguistic claims. This project was supported at its initial stage by the Canon Foundation in Europe (September 1994 through August 1995) and the Humboldt Foundation (September 1995 through February 1996), to which I extend my sincere appreciation. Kansai Gaidai University provided most generous support for my work by appointing me a Fellow at the Intercultural Research Institute (April 2001 through March 2004) and by supplying a subvention for the publication of this book. I am deeply grateful to Sadato Tanimoto (President and Chancellor of Kansai Gaidai University), Eiko Tanimoto (Vice Chancellor of Kansai Gaidai University), and Michiharu Ito (Director of the Intercultural Research Institute) for their generosity and understanding. Earlier versions of sections 1.3 and 2.1.7, section 2.9, and section 2.11, respectively, appeared in part as ‘The Metrical Organization of the Heliand: Gradation and Harmonization’, Interdisciplinary Journal for Germanic Linguistics and Semiotic Analysis 6 (2001), 11–39, ‘The Metrical Reorganization of Type E in the Heliand’,

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page xvii

Preface

xvii

American Journal of Germanic Linguistics and Literatures 12 (2000), 281–90, and ‘Anacrusis in the Meter of Heliand’, Interdigitations: Essays for Irmengard Rauch, ed. Gerald F. Carr, Wayne Harbert, and Lihua Zhang, 189–99 (New York: Peter Lang, 1999). Thoroughly rewritten, these are presented here in radically different form. I wish to thank the editors and publishers concerned for giving permission to use the original materials for this book. I am especially indebted to Irmengard Rauch for inviting me to give a talk at the 2000 Berkeley Germanic Linguistics Roundtable, promptly publishing the presentation in her journal (‘The Metrical Organization of the Heliand: Gradation and Harmonization’; see above), and strongly encouraging me to carry through this project. Hirakata October 2003

Seiichi Suzuki

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page xviii

Symbols and Abbreviations* a acc. adj. Angl. Beo C F fem. Gen gen. Gmc. Go. masc. neut. nom. OE OFris. OHG OS P p P/p PGmc. pl. pres. pret. R S s S/s sg. V WGmc. X x X/x

alliterating lift accusative adjective Anglian Beowulf consonant foot feminine Old Saxon Genesis genitive Germanic Gothic masculine neuter nominative Old English Old Frisian Old High German Old Saxon primary-stressed long syllable primary-stressed short syllable P or p Proto-Germanic plural present preterite sonorant (resonant) secondary-stressed long syllable secondary-stressed short syllable S or s singular vowel West Germanic unstressed long syllable unstressed short syllable or unstressed syllable in general; non-alliterating lift in contrast to ‘a’ (alliterating lift) X or x

* For use of some symbols, see section 1.4 in the text.

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page xix

Symbols and abbreviations x... -xx# ␮ ␴ * ⬎ ⬍ 傻 $ # () [] // <> / \ ⫻

one or more unstressed syllables disyllabic ending mora syllable reconstructed or non-occurring form (placed before the form in question, as in *⫻ ⫻ / /); expanded metrical type (placed after the type letter, as in D*) becomes derives from ranks higher than syllable boundary word boundary morpheme boundary foot boundaries; optional elements enclosed syntactic boundaries, unless otherwise specified phonological representation graphemic representation lift heavy drop normal drop

xix

MOSP-Prelims.qxd

5/11/04

2:51 PM

Page xx

. . . der Heliandvers ist eine bewust und meisterhaft gehandhabte form. auch in seinen exuberanzen ist so viel plan und regel, dass man nicht von formlosigkeit reden kann. (Heusler 1920: 47)

Understanding of formal techniques can only result from detailed and sympathetic examination of formal data. (Lehmann 1953: 8)

MOSP-01.qxd

5/11/04

2:36 PM

Page 1

1 Bases of Old Saxon metre: an introduction 1.1. The literary basis: the Heliand Old Saxon poetry is known to us in two works, the Heliand (5983 lines; Behaghel and Taeger 1984) and the Genesis (337 lines; Behaghel and Taeger 1984; Doane 1991), the first being preserved in nearly full form, the other only in portions. In view of the extensive corpus size as well as the virtually self-contained status as a single work, I am primarily concerned with the Heliand in this book, relegating therewith the treatment of the Genesis to an appendix (Appendix 2). The text of the Heliand survives in two major manuscripts, M and C, and three fragments, P, V, and S, as described briefly below (Behaghel and Taeger 1984: xv–xix; Taeger 1985): (1) Extant documents of the Heliand 1. M: Codex Germanicus Monacensis (cgm) 25, Bayerische Staatsbibliothek, Munich; mid ninth century; contains lines 85 to 2198a, 2256 to 2514a, 2576 to 3414a, 3491 to 3950, 4017b to 4674, 4740b to 5275a, 5969 to 5980; Sievers 1878 2. C: Cotton Caligula A.VII, British Library, London; second half of the tenth century; contains lines 1 to 5968, with a few small gaps; Sievers 1878; Priebsch 1925 3. P: Prague fragment, R 56/2537, Deutsches Historisches Museum, Berlin; mid ninth century; contains lines 958b to 1006a; Sievers 1935: 3–4 4. V: Vatican fragment, included in Codex Palatinus Latinus (Cod. Pal. Lat.) 1447, Vatican Library, Rome; mid ninth century; contains lines 1279 to 1358a; Sievers 1935: 5–7 5. S: Straubing fragment, cgm 8840, Bayerische Staatsbibliothek, Munich; ninth century; contains lines 351b to 360a, 368b to 384a, 393b to 400a, 492 to 582b, 675 to 683a, 693 to 706, 715b to 722a; Bischoff 1979; Behaghel and Taeger 1984: 211–16 Lacking the ending in any of the preserved manuscripts, the Heliand cannot be recovered in its full form, although the missing ending would have been a matter of several leafs. Our text is the standard edition (Behaghel and Taeger 1984), based primarily on M, the manuscript preferred to C because of its earlier dating and its more consistent Saxon form. In fact, the C manuscript seems far removed from the other four both in space and time (cf. Gantert 1998: 286): this manuscript was

MOSP-01.qxd

5/11/04

2

2:36 PM

Page 2

1. Bases of Old Saxon metre

produced in England by an Anglo-Saxon (Priebsch 1925); and it is the only manuscript dated to the tenth century (and to the second half of it at that), thus separated from the others by the span of nearly a century. For translation, readers may wish to refer to Scott (1966) and Murphy (1992) in English, and to Stapel (1953) and Genzmer (1955) in German. The Heliand is an alliterative epic of the life of Christ which was composed in all probability during the reign of Louis the Pious (Emperor 814–40) in line with his monastic reforms – whether actually commissioned by the emperor himself or simply so claimed for legitimation – and thus intended for use primarily in a clerical milieu.1 While largely based on a Latin version of Tatian’s Diatessaron or Gospel harmony (cf. Weringha 1965; Petersen 1994; Gantert 1998: 39–42), this epic poem is not a superficial rendering or paraphrasing of a gospel story in Old Saxon. Rather, the work constitutes a masterful synthesis of the two fundamentally distinct cultural traditions, the pagan Germanic on the one hand and the Christian Latin on the other. Without deviating from the theological and dogmatic orthodoxy in the least, the poem reworks the Gospels by adapting to the traditional Germanic cultural world as fully as seems possible. The poet put to practice accommodating methods at all levels of composition, ranging from micro-linguistic units such as lexis and phraseology to macro-narrative devises such as stage setting, description of scenes, sequencing of events, and demarcation of narrative units (fitts or songs): he redefined and rearticulated Christian concepts with reference to the inherited system of Germanic cultural values and the corresponding stock of words and phrases; he transplanted and re-presented original scenes in the natural and cultural environments supposedly familiar to the Saxons; he amplified or simplified individual scenes as deemed appropriate to the audience’s socio-cultural background, and went so far as to rearrange their sequential ordering; and he reorganised narrative material in terms of fitts and provided comments and explanations, presumably to facilitate understanding of the audience.2 Of greater significance still is the poet’s recasting and reconstitution, rather than simple discarding and destroying, of major elements of pagan Germanic belief, in harmony with the Christian doctrine, as shown revealingly in Murphy (1989; 1992; see also Gantert 1998: 107–20). The poet domesticated omnipotent ‘Fate’ in traditional ideology as subordinate to God; and he subtly represented God in more accessible and visually appealing terms by reimaging and reconfiguring certain prominent attributes of Woden (such as runes and accompanying birds) as associated with God. All these and other specific recastings and accommodations may be appreciated as accomplishments of the poet’s unwavering pastoral and artistic efforts to bring the two confronted cultural traditions to a harmonious unity without reducing one to the

1

2

Two Latin documents, Praefatio in librum antiquum lingua Saxonica conscriptum and Versus de poeta et interprete huius codicis (Behaghel and Taeger 1984: 1–4), for all their difficulties in interpretation, have long been accepted as prefaces to the Heliand, and provided a point of departure for exploring the socio-cultural background of its composition. For a recent attempt at contextualising the poem in the concrete cultural-historical setting, see Gantert (1998: 261–94). For exemplification and analysis of such accommodating techniques and other distinct methods of presentation characteristic of the Heliand poet, see Sowinski (1985), Murphy (1989; 1992), Gantert (1998), and Cathey (1999), to name recent notable contributions.

MOSP-01.qxd

5/11/04

2:36 PM

Page 3

1.1. The literary basis

3

other forcefully and relentlessly. Thus, underlying the whole composition of the Heliand we may recognise a grand principle of design, which may be generalised as the principle of balance and harmony. This principle has a purely formal dimension as well in the structuring of the Heliand: the work, in its entirety as well as in its parts, is organised on a numerical and geometrical basis, as argued with varying degrees of persuasiveness by Rathofer (1962) and Murphy (1992) among others. Of especial significance is a symmetrical arrangement of fitts or songs constitutive of the poem, these narrative units themselves embodying a structurally balanced internal organisation (Sowinski 1985: 250–3). Song 38 ‘The Transfiguration on the Mount’ occupies the centre of the epic, around which pairs of songs with parallel content are placed symmetrically at an equal distance away from that centre (Murphy 1992; cf. Rathofer 1962; see further Chapter 5 below). While specific interpretations advanced to these structures, notably Rathofer’s (1962) theological account in terms of number symbolism, may be controversial (cf. Taeger 1970), the highly symmetrical organisation in itself would seem hardly in doubt. According to a typology of Old Germanic alliterative verse, the Heliand is conventionally characterised as a book epic, as opposed to the class of oral lay or song. The distinction primarily rests on two parameters, one concerned with the original medium for composition as well as the mode of presentation, the other with the nature of discourse realised in the work. As a book, the Heliand was composed as a written work and read to the audience or to oneself, rather than secondarily written down after it had been orally produced and circulated as such for a certain period of time. As an epic, the work constitutes a narrative of considerable length and scope, sequentially organised as a story of a heroic figure, that is, a complex of events and actions involving him and dynamically unfolding on the temporal and topological dimensions, embedded in the amplified and detailed background, and bounded with a beginning and an ending. The genre of epic thus stands in contrast to other categories of poems such as ritual, magical, gnomic, eulogistic, commemorative, and lyric, which tend to be brief, episodic, and fragmented, largely limited to a single scene. These two literary genres (lay and book epic) have diachronic implications, as they are assumed to occupy distinct places in an evolutionary trajectory of literary development, with oral lays appearing earlier and book epics emerging later. Added to these two primary parameters are further sets of conceptual distinctions which give rise to an enriched variety of poetic genres (as enumerated in the preceding paragraph) besides the two major categories at issue. Involved as auxiliary parameters are oppositions such as pagan versus Christian (concerning thematics and authorship), secular versus religious (thematics), higher versus lower (thematics in terms of cultural domains and intended audience/readership in terms of social stratification). It should be noted that these parameters are largely interdependent with each other given the socio-historical circumstances characteristic of the north-western Germanic world in the early Middle Ages. For example, the property of written composition implies a Christian authorship and an elite audience/readership. The distinction between lay and book epic has a cluster of significant correlates in style, syntax, and metre, as shown by Heusler (1920; 1943: 198–200). Of paramount importance among the five distinguishing features that Heusler identified as constitutive of book epics would seem that of style, namely, the predominance of the ‘Hakenstil’ (‘hook style’) or the ‘Bogenstil’ (‘bow style’). This notion is predicated

MOSP-01.qxd

5/11/04

4

2:36 PM

Page 4

1. Bases of Old Saxon metre

on the interface of metre and syntax, concerning as it does the relation between metrical line and sentence. The bow style is characterised by a particular mismatch between these two metrical and syntactic entities, whereby a sentence break occurs in the middle of a metrical line exactly at a point between the a-verse and the b-verse (on these and other metrical notions, see section 1.2 below). In other words, a new sentence begins with a b-verse, rather than coinciding with the beginning of a line. The poet thus put a sentence break in the middle of an ongoing line, at a place before the head stave (the governor of the alliteration of a line, that is, the first lift of the b-verse; see section 1.2 below), and thereby deliberately disrupted in syntactic terms the continuity of a metrically integrated alliterative line. Yet the line could not have come to a full stop with the sentence break inserted in between, because it would have been incomplete without the head stave and its metrical dependents being realised. Such a sense of incompletion would in turn have worked as a driving force forward and given rise to an anticipation of a further sentence to follow, thereby guaranteeing the continuity of the poem. In this way, as observed by Heusler (1920: 9–10), Hofmann (1986: 470), and others, the bow style contributed to a composition of an ever extendible discourse by holding the reader/audience in suspense at a critical point in which the sense of completion would otherwise have prevailed. The Heliand is distinguished from other book epics by its uniquely high incidence and density of the bow style, as Heusler (1920: 14–15) originally pointed out, and Hofmann (1991: 202–4) confirmed through re-examination. This may be interpreted as a stylistic basis for composing an exceptionally long work which the Heliand is, and also as a reflection that this work was produced at a later stage of book epic tradition (cf. Heusler 1920; 1943: 200; Bostock 1976: 177). The transition from the genre of lay to that of book epic marked a radical change in early Germanic verse history. Responsible for this innovation were no doubt Anglo-Saxon clerics, notably those in Northumbria, in the late seventh through eighth centuries, as is generally assumed (e.g., Heusler 1920: 1; Lehmann 1953: 17; 1956: 125; Boor 1979: 57). Presumably inspired by Latin Christian poets in the patristic period, such as Juvencus (Evangeliorum libri IV) and Sedulius (Carmen paschale), who wrote biblical epics in classical hexametric verse (cf. Heusler 1920: 3; Kartschoke 1975: 30–125), the Anglo-Saxon monks set out to produce a number of Christian epic poems on the basis of traditional Germanic alliterative metre. In the course of such experimentations and adaptations, several features of the Latin epics were incorporated into the Germanic book epics, notably the amplification of the background including the historical stage-setting and the hero’s past achievements (Lehmann 1956: 166–9). No comparable work in the vernacular, however, is found outside of Anglo-Saxon England, with the exception of Old Saxony. From this region come our two extant Old Saxon alliterative book epics, the Heliand and the Genesis. Northern Germany was the area in which Anglo-Saxon missionary work was carried out most intensively along with Frisia, as witnessed eloquently by the presence of several Old English loanwords that were absent elsewhere, such as OS godspell ‘gospel’ and OHG (Fulda) gotspel derived from OE godspell (cf. Green 1998: 346–9). This limitation in distribution of book epics to the areas closely associated with the Anglo-Saxon mission would suggest that the genre of book epic in itself was foreign to the Continent, and that it was an insular innovation which was subsequently brought about by AngloSaxons (cf. Heusler 1920; Bostock 1976: 302).

MOSP-01.qxd

5/11/04

2:36 PM

Page 5

1.1. The literary basis

5

The prevalence of the bow style in the Heliand to the extent unknown to even Old English book epics, as observed above, makes further testimony that the work presupposed the highly developed Anglo-Saxon tradition of book epics; it would be hardly conceivable that the elaborated bow style of the Heliand came into being on Saxon soil on its own without stimulating contact with the insular tradition. Furthermore, the preservation on the Continent of a few alliterative lays, secular or Christian, namely the Hildebrandslied, the Muspilli, and the Wessobrunn Prayer, seems to have been associated with Anglo-Saxon activity: these poems contain a number of features, linguistic and palaeographic, that may be ascribed to AngloSaxon influences (cf. Boor 1979: 49–54; Bostock 1976). This may as well indicate that particular interests shown in alliterative verse, even in its short, non-epic form, would have been inspired by Anglo-Saxon missionary activities. It would accordingly seem a short step to assuming that native alliterative tradition was much in decline on the Continent and that it would hardly have lent itself to being written down as something worthwhile without Anglo-Saxon involvement. Yet composing a biblical epic in alliterative form was adopted as a viable strategy for preaching not least because the native alliterative tradition was presumably still vigorous enough on the Continent to be amenable to appropriation and manipulation necessary for a successful proselytising. Otherwise the author would have had to choose another mode of versification (as probably did Otfrid von Weissenburg a generation or so later) or come up with a prose rendition. We thus have to assume that the Saxons did not lose their indigenous alliterative tradition completely yet, which the Heliand poet seized upon as a viable medium for creation of an alliterative book epic (cf. Boor 1979: 68; Zanni 1980: 144; Haubrichs 1985: 221; Hofmann 1986: 455; 1991: 40–1). Thus appropriated as a device for aural decoration, the metre was probably instrumental in making the work more impressive and inviting therewith the audience to a deeper appreciation of the verbal message conveyed, much as did illumination or visual ornamentation in manuscripts. By making lesser use of traditional diction largely characterised by the wealth of poetic compounds and the dearth of function words, and by adopting instead more colloquial, prose-like style (cf. Lehmann 1956: 12, 109), however, the Heliand poet may have conceded to the weakened alliterative tradition on native soil or refrained from a full commitment to this traditional mode of verbal representation. Considered from a wider perspective, rather than confined to the purely formal dimension of versification, the poet’s adoption of alliterative verse may be regarded as integral to his accommodation of the Germanic oral poetic tradition at large, with the resulting Germanisation and Saxonisation of the Gospel, real or apparent, intended or unintended, as mentioned above. The appropriation of traditional versemaking practice may be understood as a strategy that the poet designed to maximise the acceptability of his work among largely illiterate, recent converts in Saxony (Gantert 1998: 37–149). In this light, as Gantert (1998) shows persuasively, we make sense of the poet’s abundant use of devices highly characteristic of oral poetry, such as formulas and formulaic expressions (cf. Zanni 1980; Gantert 1998: 72–7), and thematic structures (cf. Gantert 1998: 77–84), although the epic as a whole constituted a written piece of work. More important, by forging a link with the form of representation heavily loaded with pagan Germanic cultural meanings – that is, by encoding the wholly new ideology of the Gospel story in the medium traditionally

MOSP-01.qxd

5/11/04

6

2:36 PM

Page 6

1. Bases of Old Saxon metre

used for conveying a body of inherited cultural knowledge, and thus having it expressed in terms sounding more impressive, appealing, and trustworthy than in plain prose, the Heliand poet would have aimed at adding to the dignity, beauty, and prestige of the message, as well as enhancing its accessibility and memorability on the part of the audience, who still depended largely on aural mode of reception and information processing (cf. Hofmann 1986: 466; Gantert 1998: 121–49, 265–72). Returning to the metre: because we have at our disposal not a single example of oral (or oral-derived) poem that would have been produced on Saxon soil, it would be scarcely possible to determine the extent to which the Heliand metre cultivated the native alliterative tradition and precisely in what ways it benefited from insular influences. Furthermore, even if we were fortunate enough to have access to a single or two such native lays, we would not be much better off: it would be inappropriate to make a simple comparison between them and the Heliand given the enormous difference in corpus size. Many verse forms found in the Heliand might be absent in those hypothetically preserved lays simply for pure accident rather than structural reasons. Of greater importance still, book epics differ from lays fundamentally both in quantity and quality, as observed above: they are much more extended in length and organised in more complex manners; a significant discontinuity accordingly marks the two genres in evolutionary terms as well. For this reason, it would seem conceivable that the long-established metre characteristic of lays would have proved to be less than fully suitable as a medium for inscribing book epics and thus been subjected to elaboration and reorganisation in no small measure. Taking the above observations into consideration, we may turn to Old English Beowulf (Klaeber 1950) as an indispensable dataset for comparative and diachronic investigations of the Heliand metre. While overlaid with a Christian world-view and correspondingly recast in Christian terms here and there, this poem in its essence constitutes a pagan heroic epic in which a warrior leader of traditional Germanic society features centrally. In view of such an underlying thematics, the work would have been deep-rooted in pagan oral tradition, drawing heavily on orally inherited heroic lays, and may thus be characterised as oral-derived at least in part (cf. Suzuki 1996a: 4). By contrast, taken as an integrated whole the work as it was recorded in the extant manuscript is obviously a written product of ecclesiastical society, and in this sense it scarcely differs from book epics of Christian content. Given such inherent ambivalence of the poem as regards its basis of composition, and also given its central concerns with traditional secular values, Beowulf as it comes down to us may contain mixed properties, those derived from lays on the one hand, and those characteristic of book epics on the other. Furthermore, there is substantive reason, primarily metrical, for assuming that Beowulf constitutes the earliest known Old English epic, presumably composed in the eighth century (cf. Fulk 1992). Largely building on the type of work represented by Beowulf, Old English biblical epics would subsequently have developed some features or others of their own. We may thus characterise Beowulf as an intermediary between pagan oral lay and Christian book epic in typological and literary-evolutionary terms (cf. Heusler 1920), and postulate Beowulf as a precursor of the Heliand, a biblical heroic book epic. The profound sympathy and respect of the Heliand poet toward the pagan Germanic cultural tradition as remarked above brings the two poems still closer to each other. This does not mean in the least, of course, that the Heliand poet actually used

MOSP-01.qxd

5/11/04

2:36 PM

Page 7

1.2. The metrical basis

7

Beowulf as his direct model; nor does it imply that the Heliand author did not refer to Old English biblical epic poems whatsoever. Rather, given the limited range of material available to us, we may be justified on methodological grounds in assuming that Beowulf embodies in condensed form a necessary set of traditional conventions of alliterative epic verse that the Heliand poet would have built on for his own composition. Accordingly, the metre of Beowulf will figure prominently as a body of data to be compared with the Heliand counterpart, and will be interpreted as having provided an empirical basis and artistic inspiration for many metrical innovations that the Heliand poet would have accomplished. In other words, the Beowulf metre would have represented the substance of a metrical knowledge that the Heliand poet inherited and made innovative use of for his own work. For our investigations that follow, the metre of Beowulf may thus be identified with the prototype of the Old English alliterative tradition that ran through the composition of Old English book epics. In the light of the Heliand poet’s deep sense of balance and harmony as we find permeating the poem at its various levels of composition, we may naturally expect that he would also have organised the metre with comparable care, skill, and sophistication. It would seem unlikely that the poet would have passively reproduced the traditional metre at hand, indigenous, imported, or in mixed form, with all its attendant structural imperfections. Rather, he would have remade and reorganised the metre as deemed most appropriate to his artistic concerns and aesthetic sensibilities. This book will thus explore, from a synchronic-descriptive perspective, the metrical organisation of the Heliand in its detail and fullness, and then elucidate, from a diachronic-comparative perspective, the poet’s conformity to and remaking of the traditional metre as an integral part of his project of creating the Saxon Gospel.

1.2. The metrical basis: the Old English alliterative tradition This section provides an overview of Old English metre with particular reference to Beowulf, a unique heroic epic with dominantly non-biblical, secular contents, by recapitulating on its major structural properties as presented in Suzuki (1996a: 371–8). The following recapitulation will also serve as an introduction to the descriptive and theoretical framework of this study. The metre is conceptualised as an open-ended system of rules and representations that defines metricality on the basis of prototype in gradient terms in accordance with a complex of preference conditions or parameters (Suzuki 1996a: 10). Accordingly, metrical categories and constructs are characterised most clearly by reference to the properties of their central members; less typical instances, on the other hand, are determined in gradient terms in the light of their varying degrees of similarity and dissimilarity to their prototype. Rather than rigidly excluding violations to the rules on categorical grounds, the metrical system thus allows for a range of deviations from the norm as graded variations on the prototype, and leaves some boundary areas with ambiguity and indeterminacy in categorisation. Old English metre is grounded on the organisation and arrangement of the following three units on the vertical and horizontal dimensions: the line, the verse,

MOSP-01.qxd

5/11/04

2:36 PM

Page 8

8

1. Bases of Old Saxon metre

and the position. On the vertical dimension, these three metrical units are integrated into a hierarchy – the metrical hierarchy – in which a subordinate unit serves as an immediate constituent of a directly superordinate counterpart: the line, the highest unit in the hierarchy, consists of verses, and the verse, the next lower unit, comprises in its turn metrical positions, the lowest unit in the metre. While the line thus constitutes the highest unit in the metrical hierarchy, there is a certain indication that metrical generalisations may go beyond the line on limited occasions: the hypermetric line (see the following paragraph), the markedly long variant of the line, usually occurs in a group, rather than singly, and therefore we can hardly miss that an interlinear relation has some bearing on the metre. Yet such a clustering of hypermetric lines is varied in extension without following any demonstrable formal restriction. Accordingly, we may still maintain the thesis that the line is the highest metrical unit formally definable. Each unit has two realisation variants that stand in binary opposition, one member characterised as marked, the other as unmarked: the lines fall into normal (unmarked) and hypermetric (marked) lines; the verses are dichotomised into a-verses (marked) and b-verses (unmarked); and the positions are divided into lifts (strong positions; marked) and drops (weak positions; unmarked), and the drops are further divided into normal (or light; unmarked) and heavy (marked) drops. A higher unit is formed through integration of opposed variants of a next lower unit: thus, the line, whether normal or hypermetric, is composed by combining an a-verse (a normal a-verse and a hypermetric a-verse, respectively) and a b-verse (a normal b-verse and a hypermetric b-verse, respectively); and the verse is created by arranging lifts and drops as prescribed in the following paragraph. Thus, the metrical units and their variants are organised in the following hierarchy, as represented in Table 1.1. As regards the organisation of the verse, the normal one, whether the a-verse or b-verse, consists of four metrical positions (the principle of four positions per verse). The four positions constitutive of the normal verse are varied in terms of metrical strength or prominence. Two of the four constituent positions count as relatively strong, thereby serving as lifts, while the remaining two constitute relatively weak positions, that is, drops. By contrast, the hypermetric verse usually contains six positions, regardless of the a-verse/b-verse distinction. However, because the hypermetric verse reduces to a well-defined combination of two normal verses, I shall be concerned exclusively with the composition of the normal verse in the following overview. On the basis of differing concatenations of lifts and drops, the following verse forms are recognised as distinct metrical types, that is, abstract metrical structures underlying a host of concrete verses composed with particular language materials (see Table 1.2). Table 1.1. The organisation of metrical units in Old English metre unit

variant

line

normal

verse

a-verse

position

lift

hypermetric b-verse drop

lift

a-verse drop

lift

b-verse drop

lift

drop

MOSP-01.qxd

5/11/04

2:36 PM

Page 9

1.2. The metrical basis

9

Table 1.2. The inventory of metrical types and subtypes in Old English metre type/subtype A1 subtype A1s A2a A2b A2ab A3 B1 B3 C D1 subtype D*1 D2a subtype D*2a D2b subtype D*2b E

metrical structure /
/
/\/
/
/\ /\/\

/

/
/


/
//
//

//\
//
\ /\
/

example

syllable concatenation

Beo 11a gomban gyldan Beo 1682a gromheort guma Beo 1033a scurheard sceþðan Beo 65a wiges weorðmynd Beo 1719a breosthord blodreow Beo 22a þæt hine on ylde Beo 6b syððan æ¯rest wearð Beo 1728a Hwilum he on lufan Beo 11b þæt wæs god cyning Beo 345b min æ¯rende Beo 1532a yrre oretta Beo 31a leof landfruma Beo 326a rondas regnhearde Beo 1307a har hilderinc Beo 938a leoda landgeweorc Beo 8b weorðmyndum þah

Px#Px PS#px PS#Px Px#PS PS#PS xxxxPx xxPx#P xxxxpx xxP#px P#PXx Px#PXx P#Psx Px#PSx P#PxS Px#PxS PSx#P

Key: for metrical structure: /  lift; \  heavy drop;
 drop; for syllable concatenation: P  primary-stressed long syllable; p  primary-stressed short syllable; S  secondarystressed long syllable; s  secondary-stressed short syllable; x  unstressed short syllable; X  unstressed long syllable; #  word boundary (see further Symbols and Abbreviations above, and section 1.4 below)

Table 1.3. The system of metrical types in Old English metre class

A B C D E

category 1: basic

2: increased

3: reduced

A1 B1 C D1 E

A2a, A2b, A2ab

A3 B3

D2a, D2b

The twelve metrical types listed above (subtypes excepted) are integrated into a highly structured unitary system, the metrical paradigm as may be designated, represented in Table 1.3. On the basis of the quantity and quality of component metrical positions, the metrical types fall into two binary-opposed categories: the basic (unmarked) and the derived (marked). The derived category is further divided into two subcategories, the increased and the reduced. The basic category includes the metrical types that comprise two lifts and two normal drops, namely, types A1, B1,

MOSP-01.qxd

5/11/04

10

2:36 PM

Page 10

1. Bases of Old Saxon metre

C, D1, and E. We should note that type E, despite the presence of a heavy drop, is characterised as a basic type in the absence of structural opposition between normal and heavy drops; the configurations /
\ / or /

/ are lacking in the system. The increased category is characterised by the presence of a heavy drop in the place of a normal counterpart. This category consists of types A2a, A2b, A2ab, D2a, and D2b. The reduced category is distinguished by the occurrence of a single lift and three drops, and embodied by type A3 and also by an extremely rare configuration, type B3. Furthermore, there are several notable configurations which are subsumable under some of the established metrical types: subtypes A1s, D*1, D*2a, and D*2b. These are characterised as non-prototypical marked realisation variants of types A1, D1, D2a, and D2b, respectively. While involved in the dual binary oppositions on the paradigmatic axis, the metrical positions of the same value (the lift, the normal drop, or the heavy drop) are subject to fine-scale differentiations of strength based on their location in the verse: given two like positions, the one standing nearer the beginning of a verse counts as relatively stronger. The four positions in the verse (counted from the beginning, designated in the following representation as 1 through 4) are accordingly scaled in terms of metrical strength on the syntagmatic axis, as represented below: (2) Linearity-based Metrical Strength Scale 1 2 3 4  weaker Postulated solely on metrical grounds, the metrical position, along with the other two metrical units, is a purely metrical notion abstracted away from its linguistic manifestations. It is at the level of linguistic-metrical association (Heusler’s (1956: §40) Versfüllung or sprachliche Füllung), an interface between metre and language, in which verse composition takes place. It is accordingly with this level that I am primarily concerned in exploring the mechanism through which an individual metrical position is realised by specific language material (Suzuki 1996a: 2–3). According to the prototypical mode of linguistic-metrical association, the lift and, to a lesser extent, the heavy drop, are occupied by a long stressed syllable (P or S), whereas the normal drop is filled by an unstressed syllable or a sequence of unstressed syllables (x or x . . .). Among the complex of constraints and conventions that further regulate linguistic-metrical association, three devices are of central importance: alliteration, resolution, and anacrusis. These devices all concern relatively strong positions (lifts and heavy drops), particularly lifts. In essence, alliteration requires that the initial consonant of the first lift in each constituent verse of a line be necessarily the same, with optional operation on the second lift of the a-verse: e.g., Beo 7 feasceaft funden; / he þæs frofre gebad. Resolution constitutes a special mode of linguistic-metrical association whereby a lift or a heavy drop is occupied by a disyllabic sequence of a short stressed syllable and a following (long or short) unstressed syllable (pX/x or sX/x), rather than by a single long stressed syllable (P or S) as is usually the case: e.g., Beo 5 monegum mæ¯gþum (pxx#Px; type A1 /
/
) / meodosetla ofteah (pxSx#xP; type E / \
/). By anacrusis, the first lift of the a-verse (and very rarely of the b-verse as well) is prefixed with an unstressed

MOSP-01.qxd

5/11/04

2:36 PM

Page 11

1.3. The phonological basis

11

syllable or two; these extra syllables do not scan as constituting an independent metrical position and are thus immaterial in metrical type categorisation: e.g., Beo 1108a ahæfen of horde (xpx#xPx; type A1 /
/
).

1.3. The phonological basis: a weakening of stress The distinct metrical practice of the Heliand that deviates from Old English metre has often been attributed to linguistic factors, in particular a weakening of stress. Most recently, Russom (1998), following Lehmann’s (1956) groundwork, reiterates the thesis that a weakening of stress was primarily responsible for the differentiation of the Heliand metre from other cognate metres. Hofmann (1991: 35) raises an objection to such linguistic reductionism: he calls into question the allegedly determining role that the weakening of stress played in (re)shaping a metrical system, and claims instead that the Heliand poet simply worked under a tradition different from the one underlying Old English poetry. Both views, if carried to their extremes, would seem highly suspicious, as they would embody absolute linguistic determinism and pure literary conventionalism, respectively: on the one hand, the Heliand poet would have worked at the mercy of linguistic givens in all details; on the other hand, the poet would have composed at will absolutely free from any linguistic constraints. In actuality, however, the two apparently opposed views mentioned in the previous paragraph differ only in points of their emphasis, and they do seem mutually compatible: Lehmann and Russom are mainly concerned with the diachrony of the Heliand metre, more specifically with the linguistic basis that made possible the metrical innovation of the Heliand (see Russom 1998: 170 n. 167); by contrast, Hofmann is interested in the synchrony of the unique Heliand metre, conceptualised as a poetic convention in its own right. Through a fruitful synthesis of the above two indispensable perspectives, this book will provide an integrated account for the linguistic and metrical motivation for and mechanism of the remaking of traditional metre that gave rise to the distinct metrical convention of the Heliand. In this section, I shall explore what has been singled out as by far the most significant linguistic factor for inducing change in the metre of Old Saxon poetry, namely, a weakening of (primary) stress. This prosodic change is often postulated for Old Saxon on the evidence of two phonological processes (Sievers 1893: §105.1, 3; Lehmann 1956: 109–10; Rauch 1992: 207; Russom 1998: 137): (i) restoration of syncopated vowels; (ii) svarabhakti vowels. Yet as is often characteristic of earlier scholarship, the claim was made largely in impressionistic terms. It will therefore be necessary to argue for the weakening of stress in Old Saxon on a more substantial basis. In what follows, I shall propose first that the restoration of syncopated vowels was made possible in Old Saxon because a necessary condition for syncope, namely Defooting in word-final position, was no longer in effect there. Second, svarabhakti vowels are interpreted as a consequence of a reduction in size of primary-stressed syllables. Finally, I shall adduce the development of West Germanic Gemination in Old Saxon as further evidence for the weakened stress in the language.

MOSP-01.qxd

5/11/04

2:36 PM

12

Page 12

1. Bases of Old Saxon metre

1.3.1. Restoration of syncopated vowels In West Germanic, unstressed short vowels were syncopated after a long syllable. In Old Saxon, however, those syncopated vowels were restored in large measure (Holthausen 1921: §§137–40; Gallée 1993: §138): (3) OS (Heliand) diubal–diubules (gen. sg.); OE (Beowulf) deofol–deofles ‘devil’ OS (Heliand) engil–engilos (nom. pl.); OE (Genesis A) engel–englas ‘angel’ OS (Heliand) hêlag–hêlagon (dat. sg. masc.); OE (Genesis A) halig–halgan ‘holy’ The syncopation/apocopation in West Germanic, particularly High Vowel Deletion (see below), may be explained by assuming that the affected vowel was vulnerable to loss by virtue of its minimal prominence; and the minimal prominence may in turn be ascribed to the status of the vowel in question being unprotected in prosodic terms: it stands in syllable-final position and is not organised into a foot, a supersyllabic unit next higher up in the prosodic hierarchy (cf. Suzuki 1995c). We may accordingly postulate the following two rules that were responsible for the loss of unstressed short vowels, /i/ and /u/ in particular, in medial and/or final position (cf. Suzuki 1996a: 244): (4) Defooting F → ø / F __# Key: F  foot (5) High Vowel Deletion V → ø / __$ g [high] Condition: V is unfooted Defooting requires that a derived (non-lexical) foot be deconstructed in word-final position when preceded by another foot, as exemplified below (Suzuki 1996a: 139): (6) OE (mur)nend vs. (mur)(nen)de ‘mourning’ Key: ( )  footed Subsequently, the unfooted vowel, when located in syllable-final position, was subject to loss due to the minimal prominence involved (High Vowel Deletion). Thus, the whole process of vowel loss (as illustrated in (7) below) may ultimately be attributed to a strong dominance of primary stress over unstressed syllables: the forceful primary stress precludes the following, word-final foot from maintaining the status of an independent foot (Defooting). (7) (en)(gilos)  [by Defooting] (en)gilos  [by High Vowel Deletion] (en)glos The restoration of syncopated vowels in Old Saxon, whatever its initial motivation (e.g., analogy aiming at transparency of the paradigm), would suggest that its full

MOSP-01.qxd

5/11/04

2:36 PM

Page 13

1.3. The phonological basis

13

(or near-full) implementation had phonological support. This would imply that the necessary condition for implementation of vowel loss did not obtain in Old Saxon: Defooting was no longer in force in the language. We may accordingly conclude that primary stress did not have the strong dominating power that it had possessed earlier in West Germanic and still had in Old English. In the absence of the strong dominance of primary stress and due to the loss of Defooting as one of its consequences, the restored vowel was no longer subject to loss: the restoration, presumably motivated by analogy in its initial stage, was thus licensed in phonological terms, and consequently the restored vowel was largely re-established as a legitimate entity. In Old Saxon, then, the word engilos, for example, received the foot organisation represented in (8a) as opposed to (8b), that is, the defooted structure obtained for Old English until operation of the syncopation: (8) a. OS (en)(gilos) b. OE (en)gilos 1.3.2. Svarabhakti vowels A second phenomenon that has been put forward as evidence for the weakened stress in Old Saxon concerns svarabhakti vowels. This process effects vowel epenthesis between /r/ or /l/ and a consonant after a short stem vowel in manners that are highly variable within and across manuscripts (Holthausen 1921: §144; Rauch 1992: 207; Gallée 1993: §134). For our purposes here, however, we need not be concerned with all details of the intra- and inter-manuscript variation of this process. Such variation may be viewed as an instance of diachronic synchrony (to use Rauch’s useful term), that is, ongoing variation and change within the Old Saxon language era (Rauch 1992: xxvi, xxxv). This change may be described in its essential form as in (9), and exemplified in (10) below: (9) Vowel epenthesis ø — V / VR__C Key: R  /r/ or /l/ (10) berht  bereht (C), berht (M) ‘bright’ 1750 ferh  ferah (C, M) ‘life’ 4156 sorgon  sorogon (C), sorgon (M) ‘sorrow’ 2617 By making a simple reference to the occurrence of svarabhakti vowels in Old Saxon as above, traditional scholarship did not explain explicitly why such a phonological process should count as evidence for the weakening of stress. Their implicit reasoning, however, would readily lend itself to a reconstruction. Since weakenings of posttonic unstressed syllables, notably syncope and apocope, are generally attributed to the preceding forceful stress, so one might reason, the converse processes, vowel epenthesis in particular, should be regarded as reliable indications of the weakening of stress. Plausible as the above reasoning may sound, a simple appeal to svarabhakti vowels can hardly substitute for a well-articulated argument for the weakening of stress in Old Saxon. A comparable change is known to have affected Old English, too: a vowel

MOSP-01.qxd

5/11/04

14

2:36 PM

Page 14

1. Bases of Old Saxon metre

is inserted between /r/ or /l/ and a non-homorganic consonant, as exemplified in (11) below (Luick 1914–21: §316; Brunner 1965: §164; Campbell 1977: §§360–1). This change seems largely a characteristic of Northumbrian, as it is observed in early Old English texts (Brunner 1965: §2 Anm. 5; Campbell 1977: §§6–7), written in Northumbrian for the most part, such as Franks Casket (FC), Leiden Riddle (LR), the earliest manuscripts of Bede’s Historia Ecclesiastica (BH), Liber Vitae Dunelmensis (LV), as well as in late Northumbrian texts including Lindisfarne Gospels (Li), Rituale Ecclesiae Dunelmensis (Rit.), and Rushworth Gospels (Ru.2). (11) Berecht ‘bright; personal name’ (LV); -berect (BH); burug ‘city’ (Li., Ru.2); -burug (BG); ðerih ‘through’ (LR); worohton ‘worked’ (Li.); wylif ‘she-wolf’ (FC) Since Old English is generally held to be characterised by powerfully stressed syllables (see below), one might be tempted to conclude in view of the Old English data as in (11) that the traditional piece of evidence for weakened stress in Old Saxon hardly carries much weight: vowel epenthesis apparently took place in a language of forceful stress as well. To object that the vowel epenthesis largely concerned Northumbrian dialects, not West Saxon, would not constitute a very convincing counterargument, as Old English poetry would not have been composed exclusively in West Saxon, given the many non-West-Saxon features it exhibits (cf. Campbell 1977: §18). Confronted with the two seemingly similar processes in Old Saxon and Old English, then, we may wish to compare them in depth, particularly by placing them properly in the wider contexts involving other phonological changes, so that we may understand them in more precise terms of phonological history. Before characterising the respective epentheses according to their specific historical contexts, however, we need a fuller description of the Old Saxon vowel epenthesis, as the conventional account that we find in handbooks is far from substantial and illuminating. On closer examination, the process is found subject to a number of restrictions on operation. First, the consonants that follow /r/ or /l/ in forming clusters have to be heterorganic. Accordingly, with rare exceptions, in none of the following words does vowel epenthesis occur, as the consonant clusters involved are all homorganic (i.e., in dental/alveolar position): (12) ald ‘old’; altari ‘altar’; bald ‘brave’; barn ‘child’; beldian ‘strengthen’; bord ‘shield’; derni ‘hidden’; diurlîk ‘precious’; gidorste ‘dared’; eldi ‘people’; erda ‘earth’; erl ‘earl’; fard ‘going’; feld ‘field’; fern ‘inferno’; folda ‘earth’; ford ‘forth’; forn ‘formerly’; furdor ‘farther’; gard ‘field’; geld ‘reward’; geldan ‘reward’; gern ‘eager’; girnian ‘desire’; githuld ‘patience’; gnornon ‘lament’; gold ‘gold’; gornon ‘mourn’; guldin ‘golden’; haldan ‘hold’; halt ‘lame’; hard ‘hard’; hild ‘battle’; hold ‘gracious’; hord ‘hord’; horn ‘horn’; horsk ‘wise’; huldi ‘grace’; hurnid- ‘horned-(ship)’; kald ‘cold’; korn ‘corn’; malsk ‘proud’; meldon ‘report’; mildi ‘generous’; mord ‘murder’; mornian ‘be sad’; nord ‘northwards’; ord ‘point’; porta ‘portal’; sâlda ‘beatitude’; salt ‘salt’; seld‘seldom’; skaldan ‘push’; skard ‘scarred’; skild ‘shield’; skuld ‘debt’; skuldig ‘guilt’; smultro ‘gently’; spildian ‘kill’; spurnan ‘tread’; swart ‘black’; sweltan

MOSP-01.qxd

5/11/04

2:36 PM

Page 15

1.3. The phonological basis

15

‘die’; swerd ‘sword’; tolna ‘toll’; torn ‘anger’; thiorna ‘maiden’; thorn ‘thorn’; thurst ‘thirst’; wald ‘forest’; waldan ‘wield’; waldand ‘ruler’; ward ‘guardian’; awardian ‘harm’; wardon ‘protect’; werd (once werode) ‘worth’; wernian ‘deny’; wirdig ‘worthy’; wirs ‘worse’; wirsista (once wirist-) ‘worst’; word (rarely werod) ‘word’; wurd ‘fate’ Second, the cluster -rC is far more susceptible to vowel epenthesis than -lC. In fact, as far as the cluster -lC is concerned, vowel epenthesis seems to involve only a few examples such as bifelhan ‘order’, balouues/baluuues ‘evil’, and gelouuo ‘yellow’ (Gallée 1993: §134), while the remaining vast majority of words with -lC, as listed below, are totally untouched by the process: (13) belgan ‘be angry’; elkor ‘otherwise’; felgian ‘inflict’; folgon ‘follow’; folk ‘folk’; folmos ‘hands’; galgo ‘gallows’; galm ‘loud voice’; galpon ‘yelp’; gelp ‘scorn’; halba ‘half’; half ‘half’; halm ‘thin stalk’; helpan ‘help’; holm ‘hill’; hwelbian ‘cover’; hwelp ‘puppy’; melm ‘dust’; palma ‘palm’; qualm ‘death’; quelman ‘kill’; salbon ‘anoint’; self ‘self’; selmo ‘bed’; skalk ‘servant’; sulwian ‘sully’; tulgo ‘very’; wolkan ‘cloud’; wulf ‘wolf’ We may also recall here that the rare occurrence of vowel epenthesis on the homorganic cluster is limited to -rC, as shown in (12) above. Third, the vowel epenthesis varies considerably in its implementation according to the nature of the consonants that follow /r/. Specifically, fricatives (/h/, /g/, /f/) are most favourable to the vowel insertion, followed by a nasal (/m/); least conducive, by contrast, are stops (/k/, /p/).3 We may generalise that a wider aperture in the oral cavity is more favourable to vowel epenthesis. Such a patterned variation based as it is on manners of articulation is illustrated in (14) below, where relevant lexical items are listed, and the items printed in bold face are immune to the vowel epenthesis, and those underlined are seldom affected by the process: (14) -rh: berht ‘bright’; ferh ‘life’; ferht ‘wise’; forht ‘fear’; forhta ‘fear’; giwurht ‘deed’; thurh ‘through’; torht ‘bright’; warhta ‘worked’ ; wurhtio ‘worker’; -rg: berg ‘mountain’; burg ‘city’; fergon ‘demand’; gibirgi ‘mountains’; hwergin ‘somewhere’; morgan ‘morning’; sorga ‘care’; sorgon ‘worry’; warg ‘outlaw’; giwargean ‘punish’ -rf: arbedi ‘trouble’; derbi ‘evil’; erbiuuard ‘heir’; hwarbon ‘wander’; hwarf ‘mass of people’; hwerban ‘go’; sterban ‘die’; swerban ‘wipe’; tharbon ‘lack’; tharf ‘need’; thurban ‘need to’; thurft ‘need’; thurftig ‘needy’ -rm: arm ‘arm; poor’; armôdi ‘poverty’; armon ‘become poor’; barm ‘bosom’; farm ‘assault’; formo ‘the first’; karm ‘sigh’; storm ‘storm’; warm ‘warm’; warmian ‘warm’; wurm ‘worm’ -rk: ferkal ‘bolt’; giswerk ‘darkness’; giwerk ‘work’; giwirki ‘work’; karkari ‘jail’; marka ‘border’; gimarkon ‘mark’; mirki ‘dark’; ork ‘jug’; stark ‘strong’; 3

Gallée (1993: §134) states that vowel epenthesis may also involve -ns by referring to finistri ‘darkness’ (OHG finstrî). Yet this may have been a consequence of analogy to finistar (OHG finistar). If so, this instance may be attributed to the first process discussed in section 1.3.1 above.

MOSP-01.qxd

5/11/04

2:36 PM

Page 16

16

1. Bases of Old Saxon metre sterkian ‘strengthen’; (gi)swerkan ‘darken’; werk ‘work’; giwerkon ‘do’; wirkian ‘work’ -rp: skarp ‘sharp’; werpan ‘throw’

The foregoing re-examination of the Old Saxon vowel epenthesis might strengthen the impression that it is a close parallel to the Old English counterpart. Based on the above descriptions of the vowel epentheses in Old Saxon and Old English considered on their own, we now turn to determine their distinct characteristics from a broader diachronic perspective by placing them firmly in context against the background of phonological history. Of particular relevance to our concern are the following two phonological changes in pre-Old English: (i) breaking; (ii) smoothing. Breaking is a diphthongisation that occurred in pre-Old English as a whole. Front vowels /æ/, /æ¯/, /e/, /e/, /i/, and /i/ were thereby diphthongised to /æa/, /æ¯a/, /eo/, /eo/, /io/, and /io/, respectively, when followed by -rC, -lC, -h, or -w (Luick 1914–21: §§133–53; Brunner 1965: §§83–6; Campbell 1977: §§139–53; Suzuki 1994): (15) heard ‘hard’; healf ‘half’; meaht ‘might’; heorte ‘heart’; eolh ‘elk’; feoh ‘cattle’; biorht ‘bright’; neah ‘near’; neor ‘nearer’; lioht ‘light’ For the sake of exposition, we may ignore various contextual restrictions on the operation of breaking and their explanations, as discussed in Suzuki (1994). It should also be noted that moraic value of the vowels affected by breaking is unchanged, so that the diphthongs resulting from short vowels (/æa/, /eo/, /io/) remain monomoraic, equivalent to their inputs in quantity (Suzuki 1982).4 Smoothing, by contrast, is a monophthongisation limited to Anglian (that is, Mercian and Northumbrian dialects), whereby diphthongs, short and long, and of whatever origin, became monophthongs before back consonants /k/, /g/, /x/ with or without intervening /r/ or /l/ (Brunner 1965: §§119–20; Campbell 1977: §§222–33): (16) Angl. mæht ‘might’ (cf. WS meaht), Angl. merg ‘marrow’ (cf. WS mearg), Angl. feh ‘cattle’ (cf. WS feoh), Angl. werc ‘work’ (cf. WS weorc), Angl. gesihð ‘sight’ (cf. WS gesiehþ), Angl. birhtu ‘brightness’ (cf. WS bierhtu)

4

Closely associated with breaking is retraction, by which /æ/ became a back vowel /a/ before /w/ and /l/, rather than being diphthongised to /æa/; further, the long counterpart /æ¯/ also was retracted to /a/ before /w/. As I remarked elsewhere (Suzuki 1994: 67, 85), retraction may be characterised as a stronger version of breaking, whereby the feature [back] came to be fully incorporated into the whole segmental domain of the preceding vowel at the expense of the original value [back], as represented below: Breaking:

V g [back]



V 29 [back] [back]

Retraction:

V g [back]



V g [back]

MOSP-01.qxd

5/11/04

2:36 PM

Page 17

1.3. The phonological basis

17

It should be noted that numerous instances of diphthongs originated from breaking were also affected by smoothing. As a consequence, the diphthongs that had arisen by breaking remained preserved only before -rC and -lC, where C is non-back in place of articulation, as in heard ‘hard’ and eorl ‘earl’. These three processes – vowel epenthesis, breaking, and smoothing – would appear unrelated to or even contradictory to each other, as they involved divergent operations, vowel insertion, diphthongisation, and monophthongisation, respectively. However, Howell (1991) extensively argued with varying degrees of persuasiveness that they are all based on the same phonological motivation: they constitute simply different strategies for dealing with a common phonological problem, namely the phonotactically unstable and marked clusters of liquid  consonant (-RC). While Howell is demonstrably correct in his characterisations of breaking and vowel epenthesis as alternative solutions to the same problem, we cannot accept his analogous claim that smoothing is also functionally equivalent to these two changes. Before criticising Howell’s treatment of smoothing, however, it may be appropriate at this point to elaborate on Howell’s thesis and show that breaking and vowel epenthesis have the common underlying cause of liquid reduction or a weakening of liquids in prefinal position in the coda. First on breaking. Drawing on the claim that Old English breaking constituted a weakening of relatively sonorous segments (including liquids) in the coda as advanced in Jones (1989: 33–58, 141–66), Howell (1991), and Lutz (1991: 173–9), I offered a principled account of the mechanism of breaking that was involved in a complex of parametric variations in implementation (Suzuki 1994). Without going into detail here, breaking started as a syllable-based weakening of the relatively sonorous consonants /r/, /l/, /w/, /h/ that were located in weak position in the syllable, namely prefinal position in the coda; the positional variants of these consonants were accordingly reduced in consonantality and conversely became more vowel-like. Subsequently, the initial portion of such weakened postvocalic consonants was reconfigured as part of the preceding vowels through articulatory and perceptual readjustment (Suzuki 1994: 68–9; see also further below). Having shown that breaking was occasioned by liquid reduction and subsequent vocalic change, we now turn to vowel epenthesis and claim by elaborating on Howell that this is another solution to the same problem of liquid reduction. The functional equivalence of vowel epenthesis to breaking may be made plausible by the clustering of the three circumstantial pieces of evidence as follows (cf. Howell 1991: 60–9). First, vowel epenthesis shares much of its conditioning environment with breaking: the same sequence VrC/VlC is crucially involved in both processes. Second, breaking and vowel epenthesis took place close to each other in chronological terms: both are generally assumed to have occurred in late pre-Old English. Third, and most significant, breaking and vowel epenthesis seem mutually exclusive not only in pre-Old English but in cross-linguistic terms, as demonstrated by Howell (1991: 61–5): where breaking is implemented, there occurs no vowel epenthesis, and vice versa; put another way, breaking and vowel epenthesis do not occur simultaneously (Howell 1991: 65). More concretely, there is no form like *beoriht that is affected by both breaking and vowel epenthesis, in contrast to the attested forms beorht ‘bright’ and bereht. Given the similarity in conditioning and mutual exclusiveness in occurrence, we may be justified in refining Howell’s account and concluding that breaking and

MOSP-01.qxd

5/11/04

18

2:36 PM

Page 18

1. Bases of Old Saxon metre

vowel epenthesis constitute alternative solutions for dealing with the weakened liquids in the sequence VrC/VlC. Since liquids are more sonorous than any other consonants that may occur in the same context at issue, it is hardly surprising that only liquid  consonant clusters are involved in vowel epenthesis. The higher degrees of sonority involved in liquids than obstruents may account for the absence of vowel epenthesis after the latter segments, particularly /s/. Furthermore, the greater sonority of /r/ than of /l/ may be held responsible for /r/’s stronger inducing power of vowel epenthesis particularly in Old Saxon. Despite the enhanced sonority, these reduced liquids remained closely associated with non-reduced variants as their allophones, and their more vowel-like properties came to be extracted as belonging to separate segments by reanalysis; single segments were accordingly split to two parts on the horizontal dimension. Two alternative ways of reanalysis were available that differed in the direction of splitting: on the one hand, the weakened (vocalised) part of the liquid was realigned leftward to the preceding vowel (breaking); on the other, the corresponding portion was placed to the right (vowel epenthesis). As for smoothing, Howell (1991: 68) regards it as a negation of breaking and therefore characterises it as a strengthening (as opposed to a weakening embodied by breaking) of liquids. To be sure, smoothing extensively obliterates the effects of breaking by levelling out the second elements of broken (i.e., diphthongised) vowels, but the primary conditioning environment is the following back obstruent (/k/, /g/, /x/), whereas the existence of a liquid /r/ or /l/ is purely optional. Of particular importance is the involvement of /k/ and /g/ in smoothing, obstruents which had no bearing on breaking. In this light, I would claim that smoothing came into operation through a different motivation and mechanism; that it was thus neutral to the opposition of liquid weakening and strengthening in terms of causation; and accordingly that the apparent negating effect of breaking would have been only coincidental. More specifically, smoothing in my view was simply a transference of the feature [back] that was originally a constituent of the second element of a diphthong, whereby this feature was dissociated from the diphthong in question and subsequently reconfigured as an integrated part of the following back consonant. The reconfiguration was probably a consequence of a temporal readjustment in articulation: a delaying of the articulatory gesture for the feature [back] so that it came to coincide with the articulation of the following back consonant. We may represent the process schematically as follows: (17)

V C g0 g [back][back] [back]



V C g g [back][back]

V C C V C C  g g g g g g0 [back][back] R [back] [back] R [back] That smoothing took place even when a liquid stood between the diphthong and the back consonant may be explained by assuming that with further reduction the liquids involved in the context at issue were so weakened in consonantal articulation that

MOSP-01.qxd

5/11/04

2:36 PM

Page 19

1.3. The phonological basis

19

they came to be treated as no different from vowels. In other words, the sequence vowel  liquid  back consonant would have been hardly distinguishable in articulatory terms from the sequence vowel  back consonant. Interpreted this way, smoothing partly built on the same phonetic factor that was responsible for breaking, namely liquid reduction in prefinal position in the coda. Rather than constituting a liquid reinforcement as opposed to breaking, smoothing thus fed on liquid reduction in part as regards its exact manner of implementation. After operation of smoothing, however, the liquids in the resulting sequence VRC remained susceptible to the same weakening process that originally motivated breaking, as they continued to occupy the same position in the coda. Yet the effects of breaking had just been obliterated by smoothing, and accordingly two different strategies presented themselves in turn as an alternative to breaking. First, the weakened liquids remained as such, immune to further appreciable modification. A second strategy was vowel epenthesis, which constituted another solution for implementing liquid reduction without effecting vowel change (diphthongisation), as we have seen above. The first solution was adopted in Mercian, in which the resultant sequences of monophthong  liquid  back consonant were retained. By contrast, the second strategy was partly taken in Northumbrian: the sequences monophthong  liquid  back consonant were subject to vowel epenthesis, as illustrated in (11) above. In summary, the three sound changes – breaking, smoothing, and vowel epenthesis – all were associated with liquid reduction in their own different ways: breaking and vowel epenthesis were mutually exclusive reactions to liquid reduction, while occurrence of smoothing before the clusters of liquid  back consonant, rather than before a back consonant standing alone, was made possible by liquid reduction; and most importantly, the three processes constituted a series of chain reactions: breaking  smoothing  vowel epenthesis (cf. Campbell 1977: §361). The implementation of these changes, however, varied from dialect to dialect, as indicated in Table 1.4: breaking was of the widest scope, covering Old English in its entirety, smoothing applied to Mercian and Northumbrian, and of the narrowest applicability was vowel epenthesis, limited as it was to Northumbrian. Placed in the context of Old English phonological history with its dialectal variability, then, vowel epenthesis should best be characterised as an auxiliary solution to liquid reduction, contingent on precedent changes; it was limited geographically, and implemented relatively late. The process depended on the prior operation of Table 1.4. Distribution of breaking, smoothing, and vowel epenthesis in Old English dialects and Old Saxon breaking

smoothing

vowel epenthesis

West Saxon Mercian Northumbrian

  

  

  

Old Saxon







Key: ‘’  occurred; ‘’  did not occur

MOSP-01.qxd

5/11/04

20

2:36 PM

Page 20

1. Bases of Old Saxon metre

the two changes: breaking, which constituted the primary strategy for dealing with liquid reduction for pre-Old English as a whole; and smoothing, which happened to eliminate the effects of breaking and thereupon opened way to application of vowel epenthesis. In Old Saxon, by contrast, vowel epenthesis constituted a primary reaction to liquid reduction in the absence of its structural opposite, breaking. Thus, confronted with the same phonological problem of liquid reduction, Old English and Old Saxon reacted divergently by adopting breaking and vowel epenthesis, respectively, as their primary strategy of solution. Despite superficial similarity, vowel epenthesis accordingly had a distinct role to play in the phonological development of these two languages. While vowel epenthesis occurred as a minor readjustment process contingent on breaking and smoothing in Old English, it was implemented as a primary reaction in its own right in Old Saxon. Functionally equivalent in the two languages are therefore Old English breaking and Old Saxon vowel epenthesis, both being primary solutions for liquid reduction. Liquids became more sonorous (vowel-like) in Old Saxon as well as in Old English when standing in prefinal position in the coda (-VRC), a position that is most compatible with non-syllabic segments with maximal sonority. Subsequently, such maximally sonorous variants of liquids were decomposed by reanalysis as clusters of vowel and R. The two languages, however, took different strategies in the way of reanalysis. In Old Saxon, the most vowel-like variants of liquids in question were restructured as RV by vowel epenthesis, whereby the maximally sonorous portion of R was separated out as a distinct vowel following the remainder of R. By contrast, in Old English, the resulting clusters constituted a mirror image of the Old Saxon counterparts, namely VR. That is, the maximally sonorous, vocalic part of R was reanalysed as a vowel on its own that came to stand before the consonantal part of R, which remained as such. These two divergent reanalyses are represented as follows: (18) Old Saxon vowel epenthesis VR° C  VRVC (dissociation from the preceding stressed syllable) Old English breaking VR° C  VVRC (integration to the preceding stressed syllable) Key: R°  reduced variant of a liquid (R) What is needed then is to compare the distinct primary strategies of liquid reduction implemented in the two closely related languages with a view to uncovering their underlying difference in linguistic structure, particularly in the organisation of prosodic phonology, on the assumption that the divergence at issue would have been phonologically motivated. Unlike the restoration of syncopated vowels treated in the preceding section, vowel epenthesis in Old Saxon has no demonstrable effect on the existing phonological structure in terms of foot organisation: with or without epenthesis of a vowel, the stem vowel serves as head of a foot, thereby realising a primary stress of a given word. As far as syllable structure is concerned, however, the process has important implications. As shown in (19) below, the stressed syllable [-VRC] becomes reduced in size, resulting in [-V], because the originally syllablefinal consonant C and the prefinal resonant R are now reorganised as the coda and

MOSP-01.qxd

5/11/04

2:36 PM

Page 21

1.3. The phonological basis

21

the onset of the following syllable, which is formed on the basis of the epenthesised vowel: [-V][RVC]. We may accordingly be led to assume on the evidence of the coda reduction that primary stress has lost much of its original attracting power over constituents of the syllable: the original syllable-final and prefinal consonants are detached from it. Thus, we witness here a reduction in size of the primary-stressed syllable, which may in turn be understood as a consequence of a weakening of primary stress. (19) [-VRC]  [-V][RVC] [-VR][CV-]  [-V][RV][CV-] Key: [ ]  syllable boundaries; R  /r/ or /l/ By postulating the weakening of primary stress in Old Saxon, we may provide a unitary account of the whole process. Due to the reduced attracting power of the weakened primary stress, Old Saxon was not nearly as capable of retaining the complex syllable of the form [-VRC] as was Old English. The situation would have been particularly serious when -RC involved greatest complexity in articulation, that is, when the cluster was not homorganic. The articulatory difficulty would have been compounded when the non-homorganic cluster ended without a complete closure of the oral cavity, because the difference between the preceding sonorant and the syllablefinal consonant in terms of oral aperture was less than maximal and the syllable structure involved was accordingly found far from optimal.5 Given the insufficient attracting power of stressed syllables, then, breaking would not have served as a viable solution to liquid reduction for Old Saxon. Instead, vowel epenthesis was implemented for the clusters that counted as maximally difficult to keep intact. Accordingly, vowel epenthesis occurred most extensively on -RC, where C was a non-homorganic fricative. Another possibility of improvement might be the addition of a vowel in final position, [-VR][CV]. The resultant sequence, however, would be less desirable: because the R in the sequence [-VR][CV] is dominated by the preceding stressed syllable, this alternative would require a greater amount of attracting power of primary stress than does the vowel epenthesis, which is compatible with a minimal strength of the stressed syllable. In this light, we may also account for the change of the syllable sequence [-VR][CV-] into [-V][RV][CV-], as in sorgon  sorogon ‘sorrow’ (see (10) above). By contrast, Old English breaking, a diphthongisation of stressed front vowels, may be characterised as a vowel epenthesis before R rather than after it as in Old Saxon. As far as the location of vowel insertion is concerned, Old English breaking thus constitutes a mirror image of the Old Saxon vowel epenthesis: (20) [-VRC]  [-VVRC] *birht  OE biorht ‘bright’ *ferh  OE feorh ‘life’ 5

Since oral aperture (articulation) and sonority (perception) are closely paralleled if not fully coterminous, we may invoke part (c) of the Coda Law (Vennemann 1988: 21) as an explanatory device in this respect. ‘A syllable coda is the more preferred: . . . (c) the more sharply the Consonantal Strength drops from the offset toward the Consonantal Strength of the preceding syllable nucleus.’

MOSP-01.qxd

5/11/04

22

2:36 PM

Page 22

1. Bases of Old Saxon metre

The occurrence of the R in prefinal, rather than final, position, presupposes the syllable structure [-VRC]. Old English thus tolerates a complex coda consisting of two consonants. The retention of such a complex coda may then be ascribed to a relatively strong attracting power of primary stress in Old English. In this way, the implementation of vowel epenthesis before, rather than after, R in Old English is fully in keeping with the assumption held by many metrists that Old English maintained forceful primary stress, as stated explicitly in Lehmann (1956: 94), for example. In Old Saxon, on the other hand, R in the original sequence [-VRC] was not able to remain in prefinal position by virtue of the weakened attracting power of primary stress. In this way, the divergence between Old English and Old Saxon in their strategies for coping with liquid reduction may be attributed to the varying attracting power of stressed syllables in the two languages. 1.3.3. West Germanic Gemination In West Germanic, the sequence -VCjV- underwent gemination, resulting in -VCCjV(West Germanic Gemination). Subsequently, the gemination-inducing segment /j/ was subject to loss, with the exception of Old Saxon, where it was retained (cf. Prokosch 1939: §30; Krahe and Meid 1969a: §84): (21) PGmc. VCjV-  WGmc. -VCCjV-  OS [-VC][CjV-] other dialects [-VCC][jV-]  [-VC][CV-] e.g., PGmc. *bidjan ‘ask’ (cf. Go. bidjan)  OS biddian, OE biddan, OHG bitten, OFris. bidda We may offer a syllable-based account of the loss versus retention of /j/ by reference to syllable contact, much as the mechanism of West Germanic Gemination in itself was originally explained from an intersyllabic perspective in Murray and Vennemann (1983; see also Vennemann 1988; Suzuki 1989; 1996a: 300–3). More specifically, we may account for the divergent treatment of /j/ in Old Saxon on the one hand and other dialects on the other by postulating different strategies for achieving optimal syllable contact under the assumption that the languages concerned had different degrees of primary stress. In Old Saxon, the string [-VC][CjV-], rather than [-VCC][jV-], counted as optimal by virtue of a lesser degree of the primary stress involved, which was incapable of attracting more than a single segment as its dependent. In the languages other than Old Saxon, on the other hand, the sequence -VCCjV-, the immediate output of the gemination (-VCjV-  -VCCjV-), was syllabified [-VCC][jV-] due to a stronger attracting power of primary stress (as formulated in Vennemann 1988: 61). The resultant sequence [-VCC][jV-], however, turned out to be less than optimal in another respect. The segment /j/, standing in onset position as it had before the implementation of gemination ([-VC][jV-]), was more sonorous than the preceding syllable-final segment C, which is an obstruent. Thus, we would obtain the heterosyllabic cluster [-C][j-] in violation of the Syllable Contact Law, which is stated as follows: (22) Syllable Contact Law The preference for a syllabic structure A $ B, where A and B are marginal segments and a and b are the Consonantal Strength values of A and B respectively, increases with the value of b minus a.

MOSP-01.qxd

5/11/04

2:36 PM

Page 23

1.4. The practical basis for scanning Old Saxon verse

23

Corollary: The tendency for a syllabic structure A $ B to change, where A and B are marginal segments and a and b are the Consonantal Strength values of A and B respectively, increases with the value of a minus b. (Murray and Vennemann 1983: 520) In order to improve the less than optimal syllable contact, the syllable-initial /j/ got deleted, thereby bringing about a more harmonic contact, [-C][C-]. On the other hand, the situation in Old Saxon did not incur the kind of violation that affected the other dialects. The syllable sequence [-VC][CjV-] involved in Old Saxon was found tolerable in the light of the Syllable Contact Law much as was the sequence [-VC][CV-] in the other languages. Thus, the assumption of primary stress that is less prominent and dominating in Old Saxon than in the other West Germanic dialects may lead to a better understanding of the divergent treatment of /j/. 1.3.4. Summary The restoration of syncopated vowels in Old Saxon was brought about by the removal of Defooting in word-final position. Svarabhakti vowels developed in Old Saxon as a primary strategy for dealing with liquid reduction in prefinal position in the coda of stressed syllables; the alternative solution by breaking such as was implemented in Old English was ruled out as incompatible with the reduced coda in Old Saxon. Similarly, the retention of the West Germanic Gemination inducer /j/ in Old Saxon was made possible by the simplified coda structure of stressed syllables. The loss of Defooting and the simplification of the coda, which were immediately responsible for the three phonological changes, may in the final analysis be attributed to a reduced dominance of primary stress in Old Saxon. The weakened primary stress was no longer capable of fully dominating the following word-final foot by Defooting, which was thus removed from the phonology; and the weakening of primary stress entailed its weakened attracting power, which in turn gave rise to the simplification of the coda. Thus, we may generalise in terms of van Coetsem’s (1996) useful accent typology that Old Saxon had a lower degree of dominance of accent prominence than its earlier stage, West Germanic, and its sister language, Old English.

1.4. The practical basis for scanning Old Saxon verse: notations, foreign names, and the quað he phrase Drawing on the metrical basis introduced in section 1.2 above, this section offers a practical basis for scanning Old Saxon verse in this book. Empirical motivations for theoretical notions, descriptive categories, and analytical devices will be provided as appropriate in the following sections. A summary of symbols and notations is provided in Symbols and Abbreviations. A given verse is analysed as a concatenation of syllables, which are differentiated according to length and stress. The short syllable ends in a short vowel (-V$; $  syllable boundary), and thus counts as monomoraic (that is, having one mora – a syllablelength counting unit) in quantity. The long syllable, on the other hand, has at least

MOSP-01.qxd

5/11/04

24

2:36 PM

Page 24

1. Bases of Old Saxon metre

one segment following the nucleus (-VV$ or -VC$), and is accordingly characterised as bimoraic (that is, having two moras) in terms of mora counting. The unstressed syllable in word-final position, however, follows a special rule whereby not only -V# but -VV# and -VC# are regarded as short; the long word-final syllable thus requires the presence of an additional segment following -VV or VC (Suzuki 1996b): differentiated accordingly are hêliand ‘saviour’ (PX#) and hêlian ‘save’ (Px#), for example (on the use of the symbols P, X, and x, see below). While syllable length is always distinctive for stressed syllables, it is potentially significant for unstressed counterparts only within the word domain containing a lift. Stress is subject to a two-way distinction in kind (lexical and non-lexical) on the one hand, and a three-way distinction in degree (primary, secondary, and weak) on the other. A lexical stress inheres in a syllable that carries a lexical meaning (a rootor stem-syllable for the most part), while a non-lexical one occurs secondarily in the course of phonological derivation on a syllable lacking in such a lexical meaning. More specifically, a non-lexical stress is typically assigned to a long medial syllable or a medial sequence of a short syllable and another syllable. A primary stress is assigned to the most prominent syllable in a word, that is, the leftmost stressed syllable of the word; any other stressed syllables, whether lexical or non-lexical in origin, are designated as secondary-stressed. The stressed syllables as a whole are opposed to the unstressed (or weak-stressed) ones, which lack prosodic prominence characteristic of stress.6 While the syllable constitutes a stress-bearing unit, the presence and distribution of stress is determined within the domain of a foot, the prosodic unit whose immediate constituents are syllables. Stress thus crucially depends on foot organisation of syllables. The foot consists of a head syllable as its obligatory element and an optional dependent one. The head syllable of the foot is defined as stressed; any other syllables including those unparsed in terms of foot organisation are accordingly unstressed. Two major variants of feet are distinguished as well-formed in Old Saxon as well as other Old Germanic languages (cf. Suzuki 1995c): (i) the monosyllabic foot, which comprises a long syllable (head) as its sole constituent; (ii) the disyllabic foot, which consists of two syllables, a short syllable (head) and another syllable. In analysing a verse in terms of syllable concatenations, the following symbols are used: P  primary-stressed, long syllable; p  primary-stressed, short syllable; S  secondary- and lexical-stressed, long syllable; s  secondary- and lexicalstressed, short syllable; X  long syllable without lexical stress, that is, unstressed or non-lexical-stressed, long syllable; x  short syllable without lexical stress, that is, unstressed or non-lexical-stressed, short syllable, or unstressed syllable in general in disregard of length distinction. Put briefly, syllabification is implemented in the following way (cf. Holthausen 1921: §§64–5; Rauch 1992: 96–8): a single medial consonant belongs to the following syllable as its onset (e.g., hi$mil ‘heaven’; px), while the first element of a medial cluster closes the preceding syllable as its coda (e.g., hel$pan ‘help’; Px). This general

6

Suzuki (1996a: 137–9) provides a formal account of stress assignment in Old English, which will apply to Old Saxon in its essentials as well.

MOSP-01.qxd

5/11/04

2:36 PM

Page 25

1.4. The practical basis for scanning Old Saxon verse

25

rule of syllabification, however, is overridden by word constituent structure based on compounding in such a way that the syllable boundary has to coincide with the word internal break (e.g., thiod$arbedi ‘great misfortune’; cf. thio$da ‘people’; thio$dan ‘ruler’). The syllabic status of standing immediately before a vowel requires reconsideration. Of particular concern to the metre is the cluster  vowel that is preceded by -VCC or -VVC: decisively depending on the syllabicity of , this sequence, when appearing verse-finally, scans as disyllabic (-VCCjV- or -VVCjV-) or trisyllabic (-VCCiV- or -VVCiV-), and is accordingly subject to divergent scansion, lift  drop on the one hand, and lift  drop  drop on the other; by contrast, the same cluster following -VC receives the identical treatment, lift  drop, whether it scans as disyllabic (-VCjV-) or trisyllabic (-VCiV-), by virtue of the metrical equivalence due to resolution on -VCi-. The crucial sequence in question is embodied among others by class 2 weak verbs -ian/-ion, such as mendian ‘rejoice’ and ôkion ‘increase’ (Holthausen 1921: §463; Gallée 1993: §409). Indeed, the verbs of this class have been adduced as the prime case for the syllabic in prevocalic position (Sievers 1893: §105.4). We may accordingly concentrate on these verbs for re-examination in what follows. The alleged syllabicity of in -ian/-ion and analogous present endings in Old Saxon rests on the metrical treatment of the cognate verb class in Old English (cf. Sievers 1893: §79.1). In Old English metre, the syllabicity of is shown most cogently by the occurrence of the verbs at issue in the last three positions of type D1 (/ /

), as in Beo 2805a heah hlifian and Beo 451b leng sorgian. If were scanned as non-syllabic, these and many other similar verses would contain only three metrical positions in violation of the principle of four positions per verse (section 1.2). Further evidence is provided by the verse-final appearance of these verbs in the b-verse after a sequence of unstressed syllables, as in Beo 292b ic eow wisige and Beo 1819b þæt we fundiaþ. Scanning these verses properly as type C (
/ /
) presupposes that is syllabic; otherwise, we would obtain type A3 verses, which are categorically limited to the a-verse, however. The above arguments for the syllabicity of do not extend to Old Saxon, however; rather, there are indications to the contrary. First, in contrast to Old English poetry, there are no examples of type D verses that end in weak class 2 verbs -ia-/-io-, with a few exceptions found unique to the C manuscript (e.g., 4136a uuerod samnion; 594b upp sîthion); the alternate forms -oia-/-aia- are used instead: e.g., 2428a ford folgoiad (M), forth folgoiat (C); 1418b gode thionoian (M), gode thienoian (C). The occurrence of these obviously trisyllabic strings -VCCoia-, -VVCoia-, -VCCaia-, and -VVCaia- as occupants of the last three positions of type D, to the virtual exclusion of -VCCia-, -VVCia-, -VCCio-, and -VVCio-, thus indicates the non-syllabic value of when standing between -VCC or -VVC and a vowel. Second, and as a corollary of the above point, the use of the -ia-/-io- verbs immediately after the first lift is always (again disregarding the few exceptions in the C manuscript as noted above) accompanied by unstressed syllables in verse-initial position, so that the whole verses concerned may scan as type C (e.g., 1950b so huan so thius uuerold endiad (M), so huann so thesu uuerold endiot (C)), precisely as similarly constructed verses in which the disyllabic alternants -o-/-a- appear in

MOSP-01.qxd

5/11/04

26

2:36 PM

Page 26

1. Bases of Old Saxon metre

verse-final position instead (e.g., 4195b imu all thius uuerold folgot). The absence of verses in which the -ia-/-io- variants appear verse-finally and follow the verse-initial lift without any unstressed syllables preceding it (i.e., P#Pia- or P#Pio-) accordingly shows that these configurations count as unmetrical, consisting as they do of only three metrical positions. Hence it follows that the non-occurring forms -VCCia-, -VVCia-, -VCCio-, and -VVCio- are disyllabic, and the identification of -i- as nonsyllabic is accordingly corroborated. Third, the a-verse may end in an -ia-/-io- verb form that serves as a sole alliterative element (5033a lêt ina gesundion), whereas there is no comparable example in the b-verse. Such a discrepancy in distribution may as well indicate that the configuration at issue has to be scanned as type A3 (

/
), and therefore strengthens the proposed view that the under consideration is non-syllabic. Having thus disqualified the prime candidate for syllabic /i/ in prevocalic position, we are left with far less plausible instances, such as class 1 weak verbs, e.g., fullian ‘fill’, to which the same arguments explored above readily apply for the non-syllabic status of . In the absence of evidence to the contrary, whether internal or external to Old Saxon, I shall then generalise the foregoing observation and regard all instances of standing immediately before a vowel as representing non-syllabic /j/, in agreement with Hofmann (1991) and partly in disagreement with Kauffmann (1887: 290 Anm. 1), Sievers (1893: 151), Holthausen (1921: §171), and Gallée (1993: §195). I accordingly read endian ‘end’, nerian ‘save’, tholian ‘endure’, for example, as /en$djan/ (Px), /ner$jan/ (Px), and /þol$jan/ (Px), respectively, rather than /en$di$an/ (Pxx), /ne$ri$an/ (pxx), and /þo$li$an/ (pxx). When one or more unstressed syllables are involved, the notation x . . . is used. The symbol for the unstressed syllable (X or x) should be distinguished from the sign for the normal drop
(see the following page). As a special variant of unstressed syllable sequences, we should recognise disyllabic endings -xx#: e.g., -ono (n-stem genitive plural), -umu (a-stem adjective dative singular), -aro (a- and ô-stem adjective genitive plural). The disyllabic endings are metrically distinguished from other unstressed disyllables -xx# by their incapacity to constitute two separate drops

(cf. Hofmann 1991: 75–6). Thus, as far as the immediately preceding stem-syllable is not affected by resolution and consequently both constituent syllables of the following disyllabic ending are associated with a drop, we register its special status for scansion: for instance, grimmaro ‘grim’ in the genitive plural feminine (4128b grimmaro thioda) and grimmora in the nominative singular neuter comparative (1348a gibidig grimmora thing) are scanned Pxx# and Pxx#, respectively. Similarly, gumono ‘man’ (3075a gumono gêstos) and brosmono ‘crumb’ (3021a brosmono fulle), both n-stem masculine nouns in the genitive plural, are distinguished by scanning them as pxx# and Pxx#, respectively. In addition to syllables, a word boundary, represented as #, may play a crucial role inasmuch as it terminates the word that contains the first lift. We thus register the location of such a verse-internal word boundary. Where necessary, alliterating segments are shown in bold face when citing particular verses. To take the first line of the Heliand for exemplification, verses 1a Manega uuâron and 1b the sia iro môd gespôn are analysed as pxx#Px and xxxxP#xP, respectively.

MOSP-01.qxd

5/11/04

2:36 PM

Page 27

1.4. The practical basis for scanning Old Saxon verse

27

The syllable concatenations thus analysed are then subject to scansion in terms of metrical types, the well-defined configurations of metrical positions that were introduced in section 1.2 above and explored fully in Chapter 2. The metrical positions fall into three kinds: /  lift; \  heavy drop;
 normal drop. As will be shown in detail throughout this book, metrical prominence in the Heliand is largely encoded by quantity-based means oriented to the horizontal axis, as opposed to those based on weight oriented to the vertical axis. In view of this overall characteristic, I exclusively adopt a linear representation of metrical structure, that is, a concatenation of lift (/) and drop (\ or
), rather than a grid-based hierarchical representation that we found most suitable for the Beowulf metre (Suzuki 1996a). Verses 1a and 1b, given in the last paragraph, are scanned as type A1 (/
/
) and type B1 (
/
/), respectively. Many foreign names such as Elias, Lazarus, Maria, and Satanas, to mention a few, pose difficulty in scansion because of their indeterminate prosodic structure, notably vowel length and the number of syllables involved. Uncertainties about their prosodic properties, in the absence of definite, metre-external evidence, often make it extremely difficult to identify in precise terms the metrical types and realisation variants of the verses containing these names, and also to determine whether these verses undergo resolution or not. For example, verse 5607a Than stuod thar ôc Maria may be scanned as type A3 (


/, with Maria identified as Px or pxx; Kauffmann 1887: 350) or type C (
/ /
, with Maria identified as Pxx; Hofmann 1991: 102). Accordingly, all 520 verses in the Heliand in which one or more foreign names appear (as listed in Appendix 1 below) – including straightforward names like Martha and Petrus (both being obviously Px) for the sake of methodological consistency (cf. Hofmann 1991: 98) – are disregarded from consideration in the main text, except for examining alliteration, a metrical phenomenon which is hardly affected by indeterminate vowel length and syllable structure. I therefore classify these verses simply as F (that is, containing foreign names), abstaining thereby from scanning them according to the system of metrical types. In Appendix 1, I shall explore the prosodic structure of notable foreign names (e.g., Erodes, Iudeo) based on the detailed knowledge obtained in the following chapters of the metrical treatment of native language materials. Following Hofmann (1991: 50, 90), Russom (1998: 138), and others, I disregard from scansion the quað he ‘he said’ phrase. According to number, gender, and tense, the phrase has several variants such as quathie, quað siu, quâðun sie, quiðit he, queðat sie. The phrase serves as a marker of direct speech, inserted in the middle of an on-going speech in a vast majority of cases, and placed typically at the end of the a-verse or the b-verse, and less frequently in the verse-initial drop of the b-verse but not of the a-verse. Thus, the phrase never initiates direct speech; it is invariably preceded, and usually also followed, by a portion of the direct speech that it signals. And the phrase never appears at the beginning of a verse. With all these restrictions, the phrase may be regarded as extraneous to the verse proper, particularly when placed at its end because of the strict organisation of the end of the verse that would exclude the presence of two or more unstressed syllables there. In this light, we have to ignore the verse-final quað he phrase as in 418a ‘diuriða sî nu’, quâðun sie, 2104b ‘ic thes uuirðig ne bium’, quað he, and 3442b ‘huat, ik giuuald hebbiu’, quathie. These verses scan as types A1, B1, and C, respectively, by parenthesising quâðun sie,

MOSP-01.qxd

5/11/04

28

2:36 PM

Page 28

1. Bases of Old Saxon metre

quað he, and quathie. By contrast, in view of the most flexible variability of the verse-initial drop in size, we include the phrase as part of the verse-initial drop, as does Hofmann (1991). Accordingly, verse 499b ‘Thu scalt noh’, quað he, ‘cara thiggean’, a type C verse, is scanned as containing five syllables in its first drop (see Hofmann 1991, II: 147).

MOSP-02.qxd

5/11/04

2:38 PM

Page 29

2 Metrical types and positions: levelling and reorganisation

In this chapter, first from a strictly synchronic perspective I shall identify significant metrical types in the Heliand and closely examine their diverse realisations that result from varying associations of drops with language materials. Through comparison of these metrical types and their major variants in the Heliand with those in Old English metre (Beowulf), I shall then bring to light characteristic metrical features of the Heliand that deviate from the Old English metrical tradition, and explore from a diachronic perspective the motivations and mechanisms of these metrical reconfigurations and reorganisations that profoundly affected the Heliand and therewith contributed to forging a new identity of the metre.

2.1. Type A1 (/ ⫻ / ⫻) In this section, I am concerned with the varying ways in which the first drop of type A1 is realised on the surface and with the differences that they effect in verse distribution patterning. In the first, descriptive part of this section (subsections 2.1.1 through 2.1.6), I shall examine distinct distribution patterns of major realisation variants of type A1. Then in the second part devoted to explanatory accounts (subsection 2.1.7), I shall turn to characterise the observed patterns in general terms, determine their structural basis by identifying the parameters responsible for the characteristic distributions in question, and explore the emergence and establishment of new patterns in the Heliand metre through the remaking of traditional versecraft. 2.1.1. The configuration PS#Px The first drop of type A1 may be realised by a lexical-stressed syllable, as it is occupied by the second element of a compound, exemplified below: (1) 1695a unreht ôðrum 3838a sôðspel sagde 3721b lofsang uuarhtun In marked contrast to Beowulf, the verse configuration PS#Px lacks preference for the a-verse with double alliteration (Table 2.1): the a-verse with double alliteration accounts for the proportion significantly lower than the half of the attested

MOSP-02.qxd

5/11/04

30

2:38 PM

Page 30

2. Metrical types and positions

Table 2.1. Distribution of PS#Px (including Psx#Px) a-verse with double alliteration

a-verse with single alliteration

b-verse

8 (29%)

4 (14%)

16 (57%)

a-verse with double alliteration: 28a, 1695a, 2083a, 2110a, 3122a, 3838a, 4712a, 4908a; a-verse with single alliteration: 2037a, 3520a, 4544a, 4756a; b-verse: 8b, 1077b, 1802b, 1807b, 1819b, 1987b, 2056b, 2427b, 2454b, 3131b, 3452b, 3721b, 3745b, 3873b, 4885b, 5945b

Table 2.2. Distribution of PS#Px (excluding Psx#Px) a-verse with double alliteration

a-verse with single alliteration

b-verse

7 (41%)

1 (6%)

9 (53%)

a-verse with double alliteration: 28a, 1695a, 2083a, 2110a, 3122a, 3838a, 4908a; a-verse with single alliteration: 3520a; b-verse: 8b, 1077b, 1802b, 1819b, 2427b, 2454b, 3131b, 3721b, 5945b

examples; and the b-verse is more frequent than the a-verse in general. Noteworthy further is the absence of verses in which the second position is occupied by an independent word (substantive), comparable to verses like Beo 2987a heard swyrd hilted. Even when we disregard those verses in which the second position is filled by a disyllable of the form sx following Hofmann’s differentiation (Psx#Px; see section 2.1.2 below), we find the distribution basically unchanged, as Tables 2.1 and 2.2 illustrate: the incidence of the a-verse with double alliteration never exceeds 50 per cent; and the b-verse is (if slightly) more favoured than the a-verse with double alliteration, and used more frequently than the a-verse in general. We may accordingly conclude that, scarcely differentiated on categorical grounds from the configuration Px#Px (section 2.1.5), the sequence PS#Px fails to constitute a separate metrical type on its own, a type corresponding to type A2a (/ \ / ⫻) in Beowulf, and therefore we have to reject Sievers’s (1893: §114.2b) scansion. 2.1.2. The configurations PSx#Px and PS#xPx Most striking about the metre of the Heliand is the abundance of the sequences PSx#Px and PS#xPx, in which the first components (PSx# and PS#) constitute compound words. The difference between the two configurations is the place of a word boundary #: PSx# forms a trisyllabic compound (as exemplified in (2) immediately below; cf. Kauffmann 1887: 346–8; Sievers 1893: §114.2d; Hofmann 1991, II: 86–90), PS# a disyllabic one (as exemplified in (3) further below; cf. Kauffmann 1887: 346–8; Heusler 1956: §262; Hofmann 1991, II: 62–80). (2) 1056a nîðhugdig fîund 2492a mancunnie mildie 5443b unhiuri fîond (3) 25a godspell that guoda 4353a duomdag the mâreo

MOSP-02.qxd

5/11/04

2:38 PM

Page 31

2.1. Type A1

31

Table 2.3. Distribution of PSx#Px a-verse with double alliteration

a-verse with single alliteration

b-verse

13 (68%)

2 (11%)

4 (21%)

a-verse with double alliteration: 1277a, 1718a, 1828a, 1861a, 1945a, 2492a, 2587a, 2830a, 4093a, 4337a, 4795a, 5007a, 5693a; a-verse with single alliteration: 1056a, 1076a; b-verse: 956b, 1068b, 4003b, 5443b

Table 2.4. Distribution of PS#xPx a-verse with double alliteration

a-verse with single alliteration

b-verse

15 (94%)

1 (6%)

0 (0%)

a-verse with double alliteration: 25a, 1779a, 2955a, 3216a, 3231a, 3696a, 4010a, 4222a, 4464a, 4547a, 4746a, 5058a, 5514a, 5872a, 5878a; a-verse with single alliteration: 4353a

As a comparison with the sequence PS#Px (section 2.1.1) will indicate, the configurations PSx#Px and PS#xPx show a stronger preference for the a-verse with double alliteration than PS#Px. The above tables, however, also bring to light the distinct distribution patterns of PSx#Px and PS#xPx: while the configuration PSx#Px may occasionally be accompanied with single alliteration, the configuration PS#xPx is nearly limited to the a-verse with double alliteration. The string PS#xPx thus exhibits a greater degree of preference for the a-verse with double alliteration. We must therefore acknowledge that the two configurations are irreducible to a single one because of their different metrical treatment. The structural difference between the two configurations under consideration is the place of the unstressed syllable x relative to the word boundary: #x (word-initial) vs. -x# (word-final). Specifically, we may observe that the linear order #x (or the word-initial unstressed syllable) is more favourable to the a-verse with double alliteration than its reverse -x# (or the wordfinal unstressed syllable), a generalisation that will become clearer in the following discussion. The importance of #x in inducing an increase in the a-verse with double alliteration reveals itself in another context. The configuration PS#xPx shows a stronger preference for the a-verse with double alliteration than a minimally distinct string PS#Px (section 2.1.1), which is lacking in #x. The demonstrably high incidence of the a-verse with double alliteration that we find characteristic of PS#xPx in contrast to PS#Px may therefore lead us to assume that the word-initial unstressed syllable #x plays a vital role in affecting distribution pattern, because it is the only constituent that differentiates between the two otherwise identical configurations. As will be shown in due course, the presence of the word-initial unstressed syllable after the lift (#x) is recurrently associated with an increase in preference for the a-verse with double alliteration. Beowulf contains six examples of PS#xPx (or PS#xPS): 308a, 438a, 608a, 780a, 1649a, 1698a (Suzuki 1996a: 74–5): e.g., Beo 438a geolorand to guþe. By contrast, there seems to be no definite instance of PSx#Px in Beowulf (Suzuki 1996a: 76). The few apparent examples of PSx#Px are verses Beo 1128a wælfagne winter, Beo 1198a hordmaðum hæleþa, and Beo 2193a sincmaðþum selra. However, these are all

MOSP-02.qxd

5/11/04

32

2:38 PM

Page 32

2. Metrical types and positions

Table 2.5. Distribution of Psx#Px a-verse with double alliteration

a-verse with single alliteration

b-verse

1 (9%)

3 (27%)

7 (64%)

a-verse with double alliteration: 4712a; a-verse with single alliteration: 2037a, 4544a, 4756a; b-verse: 1807b, 1987b, 2056b, 3452b, 3745b, 3873b, 4885b

amenable to devocalisation and underdotting (winter, hordmaðum, sincmaðþum; Hutcheson 1995: 188; Suzuki 1996a: 404 n. 16), so that they may be scanned most plausibly as type A2a (/ \ / ⫻) or type E (/ \ ⫻ /). It should be observed by comparison that the Heliand has examples of PSx#Px which are most unlikely to be affected by underdotting, such as verse 2492a mancunnie mildie. Thus, while the Heliand seems hardly different from Beowulf as regards the occurrence and the relative frequency of PS#xPx, it is distinguished by the comparable appearance of PSx#Px, which is totally lacking in Beowulf. A further distinguishable configuration is Psx#Px, in which the secondary-stressed syllable is short (cf. Russom 1998: 165 n. 139): •

•

•

(4) 2037a stênfatu sehsi 3745b thingstedi halden 4885b uuîgsaca frummien As Table 2.5 makes evident, the configuration at issue is distinguished from PSx#Px by its marked preference for the b-verse at the expense of the a-verse with double alliteration. The distinct distribution patterns of PSx#Px and Psx#Px as illustrated above show that Psx#Px is similar to PS#Px rather than to PSx#Px, in that the b-verse is most favoured. We might then be led to conclude on the basis of the alleged equivalence of Psx#Px and PS#Px that Psx#Px involves resolution of the second position by virtue of the equation of -sx# and -S#. Alternatively, and more plausibly, however, Psx#Px may be viewed as more like Pxx#Px (Pxx# constituting a non-compound) in that the a-verse with double alliteration is represented most weakly, as shown in section 2.1.3 below. This fact may prompt us to assume that the second position is not affected by resolution; rather, it constitutes an association that is no different from the one with unstressed disyllables. Further, since Pxx#Px closely resembles Px#Px as will be substantiated in section 2.1.3 below, the configuration Psx#Px may be grouped together with Px#Px. It may follow then that the sequences -sx# and -xx# count as metrically equivalent to -x# regardless of accentuation, and the following equivalence relation may accordingly obtain: -sx# ⫽ -xx# ⫽ -x#. The metrical non-equivalence of S# and sx#, however, might strike us as contrary to what would be expected from the phonological identity of the two entities due to the equivalent foot structure characterised in moraic terms: both S and sx constitute a bimoraic foot. The divergent metrical treatment noted above would thus militate against the prosodic equivalence. Such a mismatch between metre and prosodic phonology, however, is far from an arbitrary metrical convention; rather, it has its own grounding in terms of linguistic-metrical association. As we will consider in

MOSP-02.qxd

5/11/04

2:38 PM

Page 33

2.1. Type A1

33

detail in discussing resolution and suspension of resolution in section 3.1 below, only strong positions, i.e., the lift and, to a lesser extent, the heavy drop, have access to the level of the foot; the normal drop, on the other hand, is strictly confined to a level lower down in the prosodic hierarchy for association, namely, the level of the syllable. Given such a mapping relation, it may follow as a matter of course that the disyllable -sx is treated in parallel to -xx in contrast to the monosyllable -S: these two structures count as equivalent insofar as the syllable level is concerned, as they both contain just the same number of syllables. The stressed disyllable -sx is thus treated as indistinguishable from the unstressed counterpart -xx exclusively on the syllable dimension without regard to supersyllabic properties. 2.1.3. The configurations PXx#Px, Pxx#Px, PX#xPx, and Px#xPx The foregoing consideration of the configurations PSx#Px and PS#xPx brings us next to examine the sets of their minimally distinct counterparts, PXx#Px/Pxx#Px on the one hand, and PX#xPx/Px#xPx on the other. To start with, the configuration PXx#Px (Table 2.6) is characterised by the long medial syllable of a non-compound (PXx#), rather than a compound (PSx#). Examples are as follows: (5) 106a uualdandes uuilleon 3438a scînandia sunna 2274b fîundes craftu Table 2.6. Distribution of PXx#Px (PXx# ⫽ non-compound) a-verse with double alliteration

a-verse with single alliteration

b-verse

14 (58%)

4 (17%)

6 (25%)

a-verse with double alliteration: 106a, 117a, 260a, 779a, 2163a, 2694a, 3145a, 3608a, 3777a, 3881a, 4422a, 4494a, 5065a, 5287a; a-verse with single alliteration: 3438a, 3604a, 3767a, 3936a; b-verse: 2274b, 3032b, 3737b, 4758b, 5116b, 5741b

This configuration resembles PSx#Px on the whole, although the frequency of the a-verse with double alliteration is slightly lower and conversely that of the b-verse higher. As shown in section 2.1.6 below, however, the sequence PXx#Px finds its closest parallel in P#xxPx (see below in this section). To be distinguished from PXx#Px is the configuration Pxx#Px (Table 2.7), where the medial syllable is short. Examples are provided below (cf. Russom 1998: 167): (6) 1939a hêlaga helpa 2435a iungaron mîne 316b drohtines engil By virtue of the low incidence of the a-verse with double alliteration, and conversely the high frequency of the b-verse, the configuration Pxx#Px has to be distinguished from PXx#Px. In section 2.1.2 above, we have noted with respect to the pair of strings PS#xPx and PS#Px that the presence of the word-initial unstressed syllable #x correlates to

MOSP-02.qxd

5/11/04

34

2:38 PM

Page 34

2. Metrical types and positions

Table 2.7. Distribution of Pxx#Px (Pxx# ⫽ non-compound) a-verse with double alliteration

a-verse with single alliteration

b-verse

22 (19%)

22 (19%)

69 (61%)

a-verse with double alliteration: 1514a, 1541a, 1608a, 1939a, 1999a, 2042a, 2200a, 2294a, 2580a, 2792a, 2859a, 2995a, 3155a, 3890a, 4015a, 4510a, 4780a, 4940a, 5164a, 5205a, 5508a, 5544a; a-verse with single alliteration: 140a, 325a, 936a, 1000a, 1591a, 1895a, 2188a, 2435a, 2443a, 3039a, 3042a, 3060a, 3074a, 3165a, 3205a, 3214a, 3302a, 3391a, 3824a, 3902a, 5691a, 5842a; b-verse: 50b, 175b, 316b, 360b, 384b, 418b, 816b, 1030b, 1071b, 1106b, 1335b, 1366b, 1518b, 1727b, 1762b, 1781b, 1917b, 1989b, 2061b, 2139b, 2279b, 2625b, 2791b, 2794b, 2910b, 2951b, 2996b, 3005b, 3052b, 3120b, 3151b, 3208b, 3267b, 3371b, 3406b, 3416b, 3454b, 3516b, 3611b, 3710b, 3930b, 4106b, 4167b, 4202b, 4449b, 4478b, 4521b, 4531b, 4546b, 4601b, 4633b, 4646b, 4656b, 4716b, 4856b, 4859b, 4860b, 4969b, 5141b, 5222b, 5348b, 5510b, 5548b, 5565b, 5657b, 5693b, 5782b, 5904b, 5981b

Table 2.8. Distribution of PXx#xPx a-verse with double alliteration

a-verse with single alliteration

b-verse

13 (72%)

1 (6%)

4 (22%)

a-verse with double alliteration: 523a, 715a, 1187a, 1872a, 2752a, 3275a, 3820a, 4261a, 4612a, 4763a, 5128a, 5225a, 5347a; a-verse with single alliteration: 5862a; b-verse: 2364b, 3358b, 5514b, 5672b

Table 2.9. Distribution of Pxx#xPx (excluding Pxxx#Px) a-verse with double alliteration

a-verse with single alliteration

b-verse

19 (73%)

1 (4%)

6 (23%)

a-verse with double alliteration: 203a, 1077a, 1204a, 1366a, 3174a, 3357a, 3586a, 3664a, 3752a, 3904a, 3958a, 4383a, 4985a, 5028a, 5416a, 5651a, 5669a, 5805a, 5843a; a-verse with single alliteration: 3485a; b-verse: 4b, 2069b, 2165b, 4250b, 5465b, 5471b

the increase in frequency of the a-verse with double alliteration. The question arising then is whether the same generalisation applies to other configurations. At stake in the present context are the following two minimal pairs: (i) PXx#Px and PXx#xPx (Table 2.8); (ii) Pxx#Px and Pxx#xPx (Table 2.9). As the above tables show, the configurations expanded with #x occur with greater frequency in the a-verse with double alliteration than those without, in conformity with the generalisation proposed above. In similar fashion, the string PSx#xPx (Table 2.10) is distinguished from PSx#Px (section 2.1.2) by a higher incidence of the a-verse with double alliteration, as indicated in Table 2.10. As may be recalled, the word-initial unstressed syllable #x occasions a higher incidence of the a-verse with double alliteration than the word-final counterpart -x#: the configuration PS#xPx exhibits a greater preference for the a-verse with double alliteration than PSx#Px (section 2.1.2). One may expect an analogous pattern to

MOSP-02.qxd

5/11/04

2:38 PM

Page 35

2.1. Type A1

35

Table 2.10. Distribution of PSx#xPx a-verse with double alliteration

a-verse with single alliteration

b-verse

2 (100%)

0 (0%)

0 (0%)

a-verse with double alliteration: 1674a, 2724a

Table 2.11. Distribution of PX#xPx a-verse with double alliteration

a-verse with single alliteration

b-verse

7 (70%)

1 (10%)

2 (20%)

a-verse with double alliteration: 45a, 3556a, 3597a, 3643a, 3695a, 4608a, 5490a; a-verse with single alliteration: 4032a; b-verse: 1216b, 2437b

hold good for the pair without lexical stress on the second syllable, namely, PX#xPx (as exemplified in (7) below) and PXx#Px (see (5) above). (7) 45a uuîdost giuualdan 4032a hêleand the gôdo 1216b fîund biuurpun As predicted, the string -X#x induces a greater incidence of the a-verse with double alliteration than -Xx#. Yet it does not show the maximal preference that is characteristic of the stressed counterpart -S#x. On the whole, then, the configuration PX#xPx seems more like PXx#Px in that the proportion of the b-verse is higher than the a-verse with single alliteration. The opposition between word-initial and word-final unstressed syllables that we have seen at work in determining verse distribution may lead us to examine whether a similar situation obtains for the pair of disyllabic drops -x#x as in (8) and #xx as in (9) below. That is, we may ask how the location of the word boundary relative to the first syllable in these two strings affects distribution patterning. (8) 1014a selbon gisâhun 2935a hêrro the gôdo 4244b manno sô spâhi (9) 205a barn an giburdeon 3292a môd umbi herte 5289b uuord endi dâdi As it turns out, the two configurations would appear scarcely distinguishable: while commonly differentiated from -xx# (Table 2.7), the two strings -x#x and #xx would seem to be treated as equivalent to each other. On closer scrutiny, however, whatever slight distinction there is to be recognised indicates that the string #xx is associated with a higher incidence of the a-verse with double alliteration and conversely with a lower incidence of the b-verse, a distribution pattern that is in

MOSP-02.qxd

5/11/04

2:38 PM

36

Page 36

2. Metrical types and positions

Table 2.12. Distribution of Px#xPx and P#xxPx

-x#x #xx

a-verse with double alliteration

a-verse with single alliteration

b-verse

470 (56%) 143 (57%)

68 (8%) 28 (11%)

303 (36%) 82 (32%)

Px#xPx (see under Tables 2.13 and 2.14 below) P#xxPx (see under Tables 2.15 and 2.16 below)

keeping with the generalisation noted above that the absence of the word-final unstressed syllable -x# (or the presence of the word boundary immediately after the first lift) is correlated to a higher incidence of the a-verse with double alliteration. On another dimension, the configuration Px#xP is subject to distinction on its own: depending on whether the word-initial unstressed syllable (#x) is occupied by a prefix or an independent word, this configuration falls into two variants, as shown in Tables 2.13 and 2.14. Thus, in regard to the configuration Px#xPx, the use of an independent word rather than a prefix for the word-initial unstressed syllable #x correlates to the marked increase in occurrence of the a-verse with double alliteration, as shown by Russom (1998: 146), a correlation which holds good in Beowulf as well (Suzuki 1996a: 156). Exactly the same generalisation also applies to the differentiation of the configuration P#xPx, as will be substantiated in section 2.1.5 below. On the other hand, the configuration P#xxPx is not involved in the distinction based on the morphological status of #x, that is, whether it is free or bound, as shown in Tables 2.15 and 2.16. The distinction in question thus concerns exclusively the monosyllabic realisation of the first drop of type A1 in word-initial position. Such a limited applicability may prove to be intelligible in view of the extreme rarity of disyllabic prefixes such as aftar-, under-, uuiðer-. The opposition between free and bound forms is thus largely irrelevant to the disyllabic string #xx and longer sequences. Moreover, the concatenation of a monosyllabic prefix and a non-lexical word (function word) is ruled out independently on morphological grounds (further on this point, see section 2.1.7.1 below). In sharp contrast to the Heliand stands Beowulf, which categorically excludes the configuration PXx#Px (Sievers 1885: 453; Cable 1974: 48; Fulk 1992: §264; Suzuki 1996a: 92). Beowulf disallows any stressed syllables, whether lexically or derivatively assigned, from occupying the first drop of type A1 (see section 2.1.2 above). 2.1.4. The configuration Pxx#Px As will be discussed in depth in section 2.2.2 below, a unique feature of the metre of the Heliand is that the verse-final drop may be realised by a disyllabic ending, -umu, for example (for an introductory note on disyllabic endings, see section 1.4 above). The configuration we are concerned with in this section is characterised by the association of the first drop with one of these disyllabic endings, represented as -xx# in the following discussion. Examples are as follows: (10) 209a uuîsaro uuordo 5671a dôdero manno 3248b gôdumu thegne

MOSP-02.qxd

5/11/04

2:38 PM

Page 37

2.1. Type A1

37

Table 2.13. Distribution of Px#xPx (#x ⫽ prefix) a-verse with double alliteration

a-verse with single alliteration

b-verse

198 (44%)

42 (10%)

211 (47%)

a-verse with double alliteration: 22a, 43a, 91a, 101a, 112a, 202a, 262a, 315a, 337a, 539a, 568a, 576a, 584a, 633a, 662a, 665a, 677a, 695a, 767a, 818a, 863a, 1014a, 1016a, 1052a, 1206a, 1238a, 1248a, 1250a, 1251a, 1280a, 1327a, 1371a, 1380a, 1399a, 1406a, 1409a, 1435a, 1437a, 1469a, 1470a, 1560a, 1569a, 1586a, 1615a, 1618a, 1645a, 1659a, 1673a, 1677a, 1741a, 1764a, 1864a, 1869a, 1877a, 1882a, 1938a, 1954a, 1987a, 2016a, 2076a, 2092a, 2098a, 2113a, 2166a, 2178a, 2258a, 2305a, 2311a, 2327a, 2350a, 2457a, 2475a, 2493a, 2505a, 2513a, 2519a, 2521a, 2534a, 2558a, 2569a, 2588a, 2600a, 2606a, 2626a, 2665a, 2670a, 2747a, 2759a, 2781a, 2795a, 2807a, 2809a, 2841a, 2923a, 2963a, 2972a, 2977a, 3014a, 3056a, 3064a, 3191a, 3215a, 3222a, 3252a, 3293a, 3301a, 3316a, 3329a, 3353a, 3434a, 3469a, 3528a, 3581a, 3588a, 3589a, 3636a, 3689a, 3699a, 3755a, 3773a, 3775a, 3835a, 3864a, 3920a, 4035a, 4036a, 4058a, 4073a, 4075a, 4111a, 4116a, 4120a, 4131a, 4133a, 4166a, 4193a, 4227a, 4258a, 4268a, 4386a, 4399a, 4410a, 4438a, 4462a, 4631a, 4669a, 4671a, 4681a, 4711a, 4767a, 4824a, 4854a, 4858a, 4891a, 4932a, 4934a, 4955a, 4959a, 5023a, 5034a, 5069a, 5070a, 5182a, 5184a, 5186a, 5255a, 5279a, 5288a, 5360a, 5370a, 5391a, 5434a, 5463a, 5489a, 5492a, 5527a, 5529a, 5545a, 5571a, 5579a, 5580a, 5626a, 5635a, 5639a, 5653a, 5680a, 5692a, 5697a, 5727a, 5737a, 5740a, 5754a, 5860a, 5870a, 5903a, 5912a, 5934a, 5943a; a-verse with single alliteration: 163a, 173a, 306a, 724a, 733a, 1018a, 1113a, 1245a, 1408a, 1464a, 1527a, 1692a, 1699a, 1824a, 1963a, 2425a, 2490a, 2725a, 2887a, 3200a, 3209a, 3225a, 3799a, 3855a, 3932a, 4191a, 4192a, 4246a, 4377a, 4489a, 4604a, 5000a, 5035a, 5097a, 5282a, 5331a, 5352a, 5399a, 5603a, 5615a, 5810a, 5949a; b-verse: 55b, 56b, 57b, 84b, 89b, 126b, 187b, 226b, 227b, 274b, 285b, 342b, 344b, 353b, 387b, 422b, 434b, 440b, 471b, 497b, 509b, 511b, 553b, 564b, 588b, 634b, 666b, 688b, 717b, 730b, 739b, 756b, 870b, 884b, 908b, 940b, 952b, 953b, 958b, 1008b, 1020b, 1074b, 1171b, 1184b, 1194b, 1212b, 1267b, 1326b, 1339b, 1344b, 1358b, 1413b, 1415b, 1419b, 1434b, 1440b, 1445b, 1479b, 1482b, 1501b, 1517b, 1520b, 1535b, 1602b, 1616b, 1621b, 1651b, 1656b, 1680b, 1710b, 1712b, 1721b, 1736b, 1748b, 1770b, 1810b, 1817b, 1829b, 1884b, 1893b, 1895b, 1905b, 1928b, 1934b, 2006b, 2011b, 2022b, 2067b, 2100b, 2108b, 2136b, 2196b, 2216b, 2331b, 2352b, 2371b, 2392b, 2403b, 2416b, 2441b, 2536b, 2553b, 2563b, 2636b, 2671b, 2680b, 2686b, 2701b, 2734b, 2744b, 2795b, 2829b, 2837b, 2871b, 2878b, 2879b, 3015b, 3073b, 3091b, 3167b, 3188b, 3240b, 3245b, 3260b, 3268b, 3285b, 3321b, 3412b, 3425b, 3429b, 3455b, 3524b, 3532b, 3575b, 3587b, 3594b, 3626b, 3627b, 3648b, 3735b, 3803b, 3805b, 3811b, 3844b, 3852b, 3856b, 3864b, 3876b, 3900b, 3997b, 4044b, 4057b, 4115b, 4153b, 4154b, 4170b, 4334b, 4339b, 4363b, 4373b, 4384b, 4447b, 4565b, 4603b, 4624b, 4655b, 4679b, 4693b, 4773b, 4852b, 4876b, 4893b, 4917b, 4930b, 4944b, 5039b, 5073b, 5108b, 5113b, 5115b, 5153b, 5181b, 5198b, 5244b, 5247b, 5295b, 5305b, 5308b, 5321b, 5337b, 5355b, 5357b, 5373b, 5401b, 5403b, 5414b, 5433b, 5458b, 5496b, 5569b, 5581b, 5597b, 5654b, 5666b, 5682b, 5708b, 5746b, 5750b, 5754b, 5834b, 5909b

Despite the disyllabicity of the first drop commonly involved, the configurations Pxx#Px (Table 2.7) and Pxx#Px (Table 2.17) behave differently: the a-verse with double alliteration is underrepresented in Pxx#Px, whereas it is most favoured along with the b-verse in Pxx#Px. These configurations have to be distinguished accordingly in metrical terms, in contradiction to Russom’s (1998: 145 n. 50) scansion of Pxx#Px as equivalent to Pxx#Px. The major difference between the two concerns the morphological status of the disyllables involved. The disyllables in Pxx#Px constitute single endings, while those in Pxx#Px serve as exponents of separate morphemes,

MOSP-02.qxd

5/11/04

38

2:38 PM

Page 38

2. Metrical types and positions

Table 2.14. Distribution of Px#xPx (#x ⫽ independent word) a-verse with double alliteration

a-verse with single alliteration

b-verse

272 (70%)

26 (7%)

92 (24%)

a-verse with double alliteration: 11a, 100a, 142a, 171a, 184a, 186a, 190a, 192a, 220a, 265a, 267a, 280a, 295a, 305a, 391a, 439a, 497a, 506a, 508a, 615a, 636a, 650a, 664a, 678a, 692a, 704a, 722a, 761a, 771a, 773a, 790a, 799a, 854a, 887a, 925a, 950a, 953a, 959a, 1021a, 1037a, 1124a, 1147a, 1172a, 1232a, 1286a, 1289a, 1294a, 1360a, 1369a, 1493a, 1498a, 1510a, 1519a, 1543a, 1550a, 1577a, 1578a, 1604a, 1719a, 1723a, 1731a, 1744a, 1765a, 1778a, 1787a, 1853a, 1902a, 1924a, 1959a, 1985a, 1988a, 2008a, 2013a, 2021a, 2023a, 2039a, 2066a, 2077a, 2082a, 2129a, 2150a, 2207a, 2225a, 2228a, 2267a, 2273a, 2341a, 2385a, 2405a, 2463a, 2524a, 2579a, 2633a, 2669a, 2684a, 2685a, 2700a, 2721a, 2740a, 2767a, 2802a, 2817a, 2820a, 2958a, 3012a, 3023a, 3034a, 3080a, 3090a, 3093a, 3177a, 3178a, 3211a, 3236a, 3247a, 3264a, 3291a, 3308a, 3312a, 3339a, 3340a, 3361a, 3377a, 3378a, 3381a, 3382a, 3384a, 3403a, 3418a, 3423a, 3430a, 3435a, 3453a, 3460a, 3462a, 3464a, 3509a, 3510a, 3531a, 3567a, 3568a, 3610a, 3641a, 3642a, 3645a, 3658a, 3673a, 3680a, 3683a, 3700a, 3712a, 3725a, 3726a, 3782a, 3793a, 3865a, 3868a, 3878a, 3888a, 3909a, 3947a, 3961a, 3975a, 3991a, 4051a, 4077a, 4096a, 4108a, 4121a, 4125a, 4138a, 4150a, 4163a, 4233a, 4240a, 4282a, 4306a, 4327a, 4336a, 4397a, 4424a, 4429a, 4474a, 4475a, 4476a, 4482a, 4484a, 4502a, 4504a, 4518a, 4528a, 4566a, 4581a, 4651a, 4670a, 4673a, 4725a, 4752a, 4776a, 4790a, 4798a, 4849a, 4857a, 4889a, 4901a, 4918a, 4928a, 4947a, 4952a, 4988a, 4991a, 5003a, 5056a, 5062a, 5067a, 5079a, 5082a, 5105a, 5108a, 5111a, 5146a, 5161a, 5200a, 5215a, 5263a, 5266a, 5294a, 5306a, 5309a, 5324a, 5326a, 5330a, 5361a, 5381a, 5385a, 5418a, 5419a, 5421a, 5423a, 5438a, 5450a, 5456a, 5480a, 5485a, 5513a, 5518a, 5519a, 5526a, 5535a, 5537a, 5542a, 5562a, 5592a, 5594a, 5633a, 5674a, 5684a, 5686a, 5689a, 5695a, 5718a, 5728a, 5751a, 5759a, 5769a, 5792a, 5845a, 5847a, 5868a, 5876a, 5884a, 5950a; a-verse with single alliteration: 215a, 540a, 927a, 1019a, 1073a, 1459a, 1539a, 1588a, 1878a, 1941a, 2105a, 2230a, 2535a, 2844a, 2868a, 2935a, 3207a, 3258a, 3891a, 3963a, 4004a, 4784a, 4982a, 5249a, 5351a, 5920’a; b-verse: 34b, 51b, 86b, 148b, 207b, 223b, 401b, 498b, 659b, 741b, 857b, 986b, 1025b, 1116b, 1137b, 1221b, 1232b, 1521b, 1607b, 1642b, 1725b, 1844b, 1918b, 2020b, 2126b, 2146b, 2169b, 2170b, 2303b, 2361b, 2363b, 2366b, 2391b, 2404b, 2423b, 2453b, 2510b, 2532b, 2546b, 2550b, 2644b, 2659b, 2661b, 2828b, 2917b, 2954b, 3154b, 3301b, 3431b, 3482b, 3622b, 3647b, 3718b, 3894b, 3961b, 3968b, 3973b, 4189b, 4197b, 4219b, 4244b, 4279b, 4296b, 4300b, 4311b, 4472b, 4594b, 4643b, 4651b, 4666b, 4704b, 4804b, 4809b, 4813b, 4916b, 4977b, 5049b, 5185b, 5235b, 5439b, 5448b, 5545b, 5554b, 5563b, 5566b, 5642b, 5652b, 5690b, 5797b, 5802b, 5861b, 5878b

Table 2.15. Distribution of P#xxPx (#x ⫽ prefix) a-verse with double alliteration

a-verse with single alliteration

b-verse

5 (62.5%)

0 (0%)

3 (37.5%)

a-verse with double alliteration: 29a, 1423a, 2322a, 5165a, 5278a; b-verse: 638b, 3795b, 3859b

derivational and inflectional, respectively. The configuration Pxx#Px is furthermore distinguishable from Px#Px, as shown in section 2.1.5 below. As with the pair of configurations PS#Px and PS#xPx treated in section 2.1.2 above, the addition of the word-initial unstressed syllable #x to Pxx#Px, as exemplified in (11) below, results in a significant increase in the a-verse with double alliteration, as indicated in Table 2.18.

MOSP-02.qxd

5/11/04

2:38 PM

Page 39

2.1. Type A1

39

Table 2.16. Distribution of P#xxPx (#x ⫽ independent word) a-verse with double alliteration

a-verse with single alliteration

b-verse

138 (56%)

28 (11%)

79 (32%)

a-verse with double alliteration: 135a, 175a, 181a, 205a, 242a, 298a, 348a, 366a, 390a, 393a, 466a, 478a, 489a, 502a, 529a, 547a, 644a, 681a, 740a, 793a, 829a, 983a, 1009a, 1099a, 1179a, 1195a, 1330a, 1376a, 1610a, 1627a, 1639a, 1644a, 1650a, 1654a, 1665a, 1715a, 1729a, 1750a, 1761a, 1814a, 1923a, 1926a, 2052a, 2065a, 2118a, 2137a, 2160a, 2205a, 2217a, 2245a, 2260a, 2339a, 2369a, 2501a, 2525a, 2609a, 2632a, 2643a, 2710a, 2834a, 2857a, 2916a, 2941a, 2950a, 2953a, 2959a, 3018a, 3118a, 3233a, 3265a, 3271a, 3281a, 3282a, 3428a, 3448a, 3487a, 3576a, 3583a, 3598a, 3629a, 3631a, 3688a, 3713a, 3880a, 3896a, 3924a, 3927a, 3938a, 3978a, 3988a, 4007a, 4020a, 4021a, 4059a, 4072a, 4091a, 4168a, 4213a, 4276a, 4318a, 4369a, 4380a, 4496a, 4618a, 4630a, 4637a, 4641a, 4674a, 4726a, 4750a, 4830a, 4871a, 4872a, 4919a, 4995a, 5011a, 5022a, 5044a, 5048a, 5100a, 5104a, 5191a, 5394a, 5436a, 5496a, 5586a, 5624a, 5627a, 5659a, 5676a, 5705a, 5738a, 5785a, 5817a, 5839a, 5867a, 5908a, 5913a; a-verse with single alliteration: 41a, 322a, 451a, 617a, 729a, 1197a, 1425a, 1438a, 1664a, 1705a, 2099a, 2152a, 2167a, 2478a, 2480a, 3274a, 3292a, 3518a, 3701a, 3871a, 4080a, 4292a, 4307a, 4403a, 4479a, 4509a, 4517a, 4640a; b-verse: 92b, 159b, 199b, 200b, 348b, 371b, 396b, 493b, 515b, 589b, 598b, 658b, 675b, 828b, 1142b, 1224b, 1247b, 1291b, 1347b, 1477b, 1522b, 1572b, 1792b, 1874b, 1967b, 2034b, 2116b, 2145b, 2157b, 2176b, 2243b, 2252b, 2286b, 2306b, 2355b, 2460b, 2531b, 2584b, 2741b, 2765b, 2923b, 3004b, 3281b, 3419b, 3434b, 3712b, 3774b, 3842b, 3870b, 3890b, 3941b, 3946b, 3992b, 4059b, 4118b, 4281b, 4305b, 4477b, 4560b, 4617b, 4619b, 4664b, 4778b, 4782b, 4878b, 4960b, 4961b, 4963b, 4994b, 5120b, 5169b, 5289b, 5319b, 5437b, 5538b, 5645b, 5777b, 5844b, 5881b

Table 2.17. Distribution of Pxx#Px a-verse with double alliteration

a-verse with single alliteration

b-verse

49 (44%)

15 (14%)

47 (42%)

a-verse with double alliteration: 209a, 816a, 832a, 852a, 892a, 907a, 955a, 1100a, 1252a, 1262a, 1559a, 1587a, 1720a, 1724a, 1734a, 1743a, 1808a, 2407a, 2466a, 2489a, 2617a, 2662a, 2673a, 2703a, 2769a, 2814a, 2843a, 2968a, 2978a, 3173a, 3229a, 3246a, 3456a, 3467a, 3574a, 3956a, 4097a, 4118a, 4209a, 4243a, 4274a, 4501a, 4690a, 4742a, 4828a, 5440a, 5478a, 5483a, 5944a; a-verse with single alliteration: 758a, 1243a, 1806a, 3021a, 3159a, 3374a, 3910a, 4487a, 4562a, 4789a, 5251a, 5312a, 5564a, 5671a, 5889a; b-verse: 75b, 439b, 612b, 613b, 735b, 1149b, 1193b, 1220b, 1564b, 1566b, 1573b, 1575b, 1624b, 1697b, 1849b, 1900b, 1999b, 2091b, 2135b, 2173b, 2222b, 2319b, 2356b, 2455b, 2605b, 2613b, 2691b, 2757b, 2785b, 3048b, 3248b, 3475b, 3478b, 3540b, 3743b, 3744b, 4025b, 4128b, 4451b, 4493b, 4576b, 4689b, 4774b, 5327b, 5429b, 5499b, 5956b

(11) 572a sprâkono sô spâhi 1355a lêðoro gilêstio The disyllabic ending may be separated from the root syllable P by an additional syllable x or X. These two configurations Pxxx#Px and PXxx#Px are exemplified in (12) and (13), respectively: (12) Pxxx#Px 4217a craftigaro kunnio 24b hêlagaro stemnun

MOSP-02.qxd

5/11/04

40

2:38 PM

Page 40

2. Metrical types and positions

Table 2.18. Distribution of Pxx#xPx a-verse with double alliteration

a-verse with single alliteration

b-verse

6 (86%)

0 (0%)

1 (14%)

a-verse with double alliteration: 572a, 886a, 1355a, 2147a, 2681a, 2719a; b-verse: 1717b

Table 2.19. Distribution of Pxxx#Px a-verse with double alliteration

a-verse with single alliteration

b-verse

1 (13%)

3 (38%)

4 (50%)

a-verse with double alliteration: 4217a; a-verse with single alliteration: 1691a, 1966a, 5969a; b-verse: 24b, 491b, 1363b, 5857b

Table 2.20. Distribution of PXxx#Px a-verse with double alliteration

a-verse with single alliteration

b-verse

3 (75%)

1 (25%)

0 (0%)

a-verse with double alliteration: 3149a, 4385a, 5687a; a-verse with single alliteration: 50a

(13) PXxx#Px 3149a libbiendero liobost 5687a uuôpiandero uuîbo The two configurations diverge in distribution: while Pxxx#Px is favoured in the b-verse (Table 2.19), PXxx#Px occurs largely in the a-verse with double alliteration (Table 2.20). 2.1.5. The configurations Px#Px, PX#Px, P#xPx, and P#xx . . . Px The sequence Px#Px has its first drop occupied by a word-final syllable alone, as illustrated below: (14) 27a drohtin diurie 2349a manno barnun 4288b liohtes skîne The distribution of Px#Px closely resembles that of Pxx#Px (section 2.1.3): while the b-verse accounts for the majority, the a-verse with single and double alliteration is divided nearly evenly (cf. Russom 1998: 145). On closer inspection, the sequence Table 2.21. Distribution of Px#Px a-verse with double alliteration

a-verse with single alliteration

b-verse

228 (18%)

213 (17%)

795 (64%)

For verse lists, see under Tables 2.22 and 2.23

MOSP-02.qxd

5/11/04

2:38 PM

Page 41

2.1. Type A1

41

Table 2.22. Distribution of PX#Px a-verse with double alliteration

a-verse with single alliteration

b-verse

7 (28%)

3 (12%)

15 (60%)

a-verse with double alliteration: 358a, 682a, 1040a, 1209a, 1767a, 3096a, 3647a; a-verse with single alliteration: 522a, 1791a, 2109a; b-verse: 1256b, 1598b, 2051b, 2059b, 2544b, 2697b, 2817b, 3254b, 3558b, 3578b, 3717b, 3790b, 4702b, 4861b, 5264b

Px#Px proves to be divided into two classes on the basis of the syllable length of the second position (Table 2.22 and Table 2.23). More specifically, when the first drop is occupied by a long syllable (-CC#) as in (15), the incidence of the a-verse with double alliteration sharply increases at the cost of the a-verse with single alliteration, as indicated in Table 2.22.1 (15) 358a uualdand uuelda 1791a uualdand biddien 5264b Uuîgand frumidun In Beowulf, by contrast, the configurations PX#Px and Px#Px are treated without appreciable distinction, as indicated in Table 2.24. As with the pairs PX#xPx and PXx#Px (section 2.1.3), the opposition between word-initial and word-final unstressed syllables (or the place of the word boundary relative to the unstressed syllable standing immediately before the second lift) seems of vital significance in effecting varying distribution patterns. Thus distinguished from Px#Px is the sequence P#xPx (such as exemplified in (16) below) by the greater frequency of the a-verse with double alliteration (Russom 1998: 145), as shown in Table 2.25. (16) 196a barn an burgun 712a sân antkenda 1240b up gebrâhti In parallel to the configuration Px#xPx treated in section 2.1.3 above, the configuration P#xPx varies largely in distribution depending on whether the first drop (#x) is occupied by a prefix or an independent word, as indicated in Tables 2.26 and 2.27. More specifically, there obtains a trading-off relation between the a-verse with double alliteration and the b-verse in the following way: when the position is filled by a prefix, the incidence of the b-verse prevails; by contrast, when it is filled by an independent word, the a-verse with double alliteration dominates at the cost of the b-verse. Even when the first drop #x is realised by a prefix, however, the configuration P#xPx (Table 2.26) exhibits a greater preference for the a-verse with double alliteration than Px#Px (Table 2.23). This difference constitutes another manifestation of the recurrent pattern whereby the word-initial unstressed syllable #x correlates to a higher frequency of the a-verse with double alliteration than the word-final 1

On the mechanism whereby -x# and -X# came to be differentiated in the Heliand, see section 2.1.7.2 below.

MOSP-02.qxd

5/11/04

42

2:38 PM

Page 42

2. Metrical types and positions

Table 2.23. Distribution of Px#Px (excluding PX#Px) a-verse with double alliteration

a-verse with single alliteration

b-verse

221 (18%)

210 (17%)

780 (64%)

a-verse with double alliteration: 27a, 137a, 178a, 180a, 187a, 214a, 241a, 271a, 319a, 380a, 406a, 409a, 428a, 445a, 486a, 492a, 501a, 525a, 545a, 569a, 652a, 656a, 673a, 737a, 746a, 825a, 835a, 841a, 934a, 968a, 996a, 997a, 1033a, 1039a, 1069a, 1078a, 1083a, 1160a, 1215a, 1227a, 1231a, 1247a, 1265a, 1362a, 1374a, 1378a, 1390a, 1417a, 1447a, 1503a, 1506a, 1532a, 1594a, 1738a, 1777a, 1784a, 1809a, 1832a, 1843a, 1852a, 1856a, 1873a, 1879a, 1892a, 1910a, 1933a, 1974a, 2054a, 2064a, 2093a, 2104a, 2175a, 2193a, 2246a, 2272a, 2280a, 2295a, 2338a, 2376a, 2399a, 2418a, 2422a, 2439a, 2481a, 2494a, 2516a, 2544a, 2552a, 2657a, 2663a, 2668a, 2672a, 2688a, 2799a, 2833a, 2874a, 2875a, 2879a, 2885a, 2928a, 2934a, 3071a, 3075a, 3104a, 3114a, 3131a, 3148a, 3219a, 3223a, 3256a, 3279a, 3317a, 3321a, 3331a, 3390a, 3458a, 3477a, 3480a, 3544a, 3546a, 3570a, 3591a, 3605a, 3624a, 3627a, 3648a, 3651a, 3656a, 3676a, 3715a, 3770a, 3784a, 3795a, 3816a, 3830a, 3833a, 3851a, 3854a, 3867a, 3869a, 3875a, 3887a, 3898a, 3939a, 3959a, 3983a, 4000a, 4029a, 4042a, 4083a, 4113a, 4117a, 4143a, 4260a, 4302a, 4305a, 4405a, 4428a, 4434a, 4457a, 4473a, 4498a, 4519a, 4540a, 4543a, 4580a, 4613a, 4653a, 4660a, 4707a, 4738a, 4747a, 4796a, 4844a, 4860a, 4863a, 4865a, 4874a, 4895a, 4998a, 5008a, 5015a, 5046a, 5055a, 5061a, 5075a, 5080a, 5120a, 5130a, 5136a, 5147a, 5151a, 5156a, 5177a, 5180a, 5183a, 5206a, 5227a, 5233a, 5239a, 5454a, 5472a, 5482a, 5494a, 5500a, 5501a, 5523a, 5536a, 5538a, 5556a, 5578a, 5582a, 5682a, 5764a, 5789a, 5793a, 5840a, 5851a, 5855a, 5902a, 5945a; a-verse with single alliteration: 1a, 46a, 114a, 122a, 134a, 170a, 182a, 206a, 217a, 219a, 269a, 287a, 291a, 373a, 383a, 425a, 431a, 470a, 481a, 487a, 499a, 561a, 641a, 645a, 708a, 710a, 787a, 802a, 808a, 844a, 879a, 906a, 917a, 932a, 956a, 1022a, 1028a, 1044a, 1054a, 1122a, 1165a, 1173a, 1198a, 1230a, 1257a, 1296a, 1401a, 1487a, 1704a, 1706a, 1759a, 1802a, 1812a, 1816a, 1845a, 1883a, 1967a, 1982a, 1983a, 1986a, 2015a, 2018a, 2028a, 2069a, 2109a, 2136a, 2198a, 2318a, 2320a, 2345a, 2349a, 2380a, 2384a, 2452a, 2454a, 2473a, 2482a, 2499a, 2565a, 2639a, 2686a, 2729a, 2758a, 2775a, 2828a, 2840a, 2854a, 2869a, 2895a, 2901a, 2920a, 2954a, 2960a, 3020a, 3026a, 3077a, 3098a, 3101a, 3105a, 3162a, 3190a, 3192a, 3197a, 3202a, 3276a, 3284a, 3286a, 3287a, 3307a, 3346a, 3347a, 3369a, 3372a, 3379a, 3406a, 3452a, 3517a, 3522a, 3538a, 3552a, 3565a, 3585a, 3596a, 3738a, 3763a, 3768a, 3798a, 3808a, 3810a, 3892a, 3914a, 3953a, 3986a, 3987a, 4151a, 4171a, 4218a, 4223a, 4230a, 4241a, 4262a, 4283a, 4358a, 4455a, 4481a, 4490a, 4542a, 4550a, 4556a, 4564a, 4567a, 4570a, 4579a, 4584a, 4602a, 4619a, 4691a, 4727a, 4732a, 4739a, 4768a, 4788a, 4794a, 4827a, 4842a, 4846a, 4869a, 4892a, 4907a, 4983a, 4984a, 5013a, 5014a, 5016a, 5045a, 5072a, 5142a, 5144a, 5149a, 5189a, 5204a, 5219a, 5232a, 5237a, 5277a, 5304a, 5407a, 5412a, 5444a, 5445a, 5446a, 5533a, 5553a, 5581a, 5666a, 5668a, 5694a, 5714a, 5717a, 5771a, 5778a, 5799a, 5800a, 5825a, 5826a, 5841a, 5893a, 5920a, 5936a, 5959a, 5971a; b-verse: 6b, 7b, 13b, 17b, 32b, 37b, 40b, 44b, 54b, 63b, 67b, 80b, 93b, 99b, 101b, 108b, 109b, 110b, 124b, 125b, 136b, 138b, 144b, 146b, 149b, 157b, 160b, 167b, 177b, 181b, 183b, 195b, 204b, 217b, 234b, 244b, 247b, 270b, 289b, 290b, 309b, 311b, 333b, 338b, 339b, 345b, 351b, 364b, 366b, 370b, 377b, 379b, 388b, 389b, 398b, 402b, 404b, 413b, 415b, 417b, 419b, 420b, 429b, 430b, 432b, 436b, 441b, 443b, 448b, 450b, 462b, 482b, 490b, 495b, 496b, 508b, 518b, 520b, 524b, 542b, 543b, 548b, 573b, 579b, 585b, 591b, 606b, 609b, 610b, 618b, 625b, 629b, 631b, 632b, 635b, 656b, 667b, 671b, 683b, 698b, 703b, 706b, 709b, 732b, 736b, 742b, 743b, 744b, 748b, 750b, 768b, 789b, 814b, 823b, 839b, 843b, 844b, 846b, 849b, 853b, 854b, 858b, 867b, 868b, 869b, 872b, 875b, 877b, 880b, 885b, 903b, 904b, 912b, 917b, 918b, 931b, 947b, 961b, 963b, 966b, 971b, 980b, 987b, 1006b, 1010b, 1021b, 1027b, 1038b, 1053b, 1055b, 1056b, 1058b, 1060b, 1076b, 1079b, 1088b, 1093b, 1110b, 1111b, 1118b, 1123b, 1128b, 1129b, 1132b, 1133b, 1138b, 1139b, 1140b, 1147b, 1155b, 1157b, 1162b, 1168b, 1176b, 1177b, 1181b,

MOSP-02.qxd

5/11/04

2:38 PM

Page 43

2.1. Type A1 1210b, 1214b, 1226b, 1235b, 1236b, 1249b, 1253b, 1254b, 1255b, 1268b, 1278b, 1279b, 1282b, 1284b, 1285b, 1286b, 1290b, 1292b, 1295b, 1299b, 1333b, 1334b, 1346b, 1351b, 1361b, 1372b, 1382b, 1383b, 1384b, 1386b, 1391b, 1398b, 1400b, 1405b, 1410b, 1416b, 1421b, 1457b, 1470b, 1471b, 1474b, 1486b, 1507b, 1511b, 1529b, 1533b, 1536b, 1545b, 1567b, 1580b, 1600b, 1601b, 1603b, 1606b, 1609b, 1610b, 1612b, 1628b, 1629b, 1630b, 1640b, 1641b, 1657b, 1690b, 1702b, 1714b, 1718b, 1747b, 1751b, 1752b, 1785b, 1800b, 1808b, 1811b, 1831b, 1834b, 1838b, 1866b, 1899b, 1908b, 1960b, 1962b, 1964b, 1965b, 1974b, 1975b, 1979b, 1993b, 1994b, 1998b, 2031b, 2032b, 2035b, 2040b, 2042b, 2068b, 2074b, 2080b, 2089b, 2096b, 2128b, 2130b, 2141b, 2142b, 2143b, 2181b, 2229b, 2232b, 2242b, 2255b, 2262b, 2263b, 2266b, 2270b, 2287b, 2297b, 2299b, 2308b, 2313b, 2330b, 2359b, 2367b, 2373b, 2374b, 2380b, 2382b, 2387b, 2388b, 2389b, 2396b, 2400b, 2412b, 2413b, 2420b, 2426b, 2430b, 2431b, 2433b, 2449b, 2450b, 2461b, 2462b, 2479b, 2483b, 2484b, 2502b, 2515b, 2527b, 2528b, 2537b, 2542b, 2543b, 2547b, 2548b, 2549b, 2566b, 2573b, 2576b, 2577b, 2578b, 2595b, 2603b, 2609b, 2611b, 2616b, 2622b, 2629b, 2630b, 2646b, 2651b, 2683b, 2690b, 2719b, 2722b, 2732b, 2736b, 2740b, 2742b, 2743b, 2745b, 2755b, 2761b, 2764b, 2772b, 2775b, 2782b, 2793b, 2798b, 2806b, 2816b, 2845b, 2877b, 2880b, 2891b, 2892b, 2898b, 2899b, 2901b, 2905b, 2907b, 2915b, 2920b, 2924b, 2950b, 2979b, 2997b, 3000b, 3003b, 3010b, 3011b, 3020b, 3024b, 3030b, 3086b, 3088b, 3089b, 3100b, 3104b, 3112b, 3121b, 3126b, 3133b, 3157b, 3169b, 3175b, 3180b, 3181b, 3192b, 3219b, 3242b, 3246b, 3263b, 3266b, 3283b, 3294b, 3303b, 3312b, 3326b, 3337b, 3338b, 3356b, 3359b, 3366b, 3379b, 3397b, 3408b, 3409b, 3410b, 3411b, 3415b, 3427b, 3439b, 3446b, 3447b, 3492b, 3508b, 3511b, 3518b, 3530b, 3534b, 3551b, 3555b, 3580b, 3582b, 3590b, 3614b, 3615b, 3616b, 3623b, 3631b, 3635b, 3641b, 3668b, 3676b, 3678b, 3684b, 3693b, 3698b, 3699b, 3706b, 3711b, 3723b, 3724b, 3729b, 3730b, 3731b, 3746b, 3749b, 3761b, 3766b, 3781b, 3788b, 3791b, 3797b, 3800b, 3801b, 3846b, 3849b, 3882b, 3892b, 3901b, 3909b, 3912b, 3915b, 3919b, 3923b, 3926b, 3949b, 3955b, 3960b, 3966b, 3989b, 4002b, 4017b, 4024b, 4031b, 4037b, 4052b, 4054b, 4065b, 4069b, 4070b, 4072b, 4091b, 4105b, 4125b, 4132b, 4134b, 4136b, 4149b, 4162b, 4165b, 4172b, 4175b, 4179b, 4183b, 4187b, 4201b, 4207b, 4213b, 4243b, 4269b, 4288b, 4290b, 4304b, 4312b, 4335b, 4345b, 4365b, 4367b, 4387b, 4412b, 4442b, 4452b, 4453b, 4458b, 4466b, 4468b, 4471b, 4486b, 4499b, 4517’b, 4522b, 4530b, 4536b, 4551b, 4557b, 4559b, 4561b, 4566b, 4591b, 4600b, 4606b, 4623b, 4627b, 4647b, 4648b, 4657b, 4667b, 4677b, 4682b, 4684b, 4696b, 4697b, 4699b, 4701b, 4714b, 4717b, 4722b, 4725b, 4735b, 4740b, 4745b, 4765b, 4772b, 4776b, 4802b, 4807b, 4816b, 4821b, 4830b, 4833b, 4836b, 4840b, 4843b, 4862b, 4864b, 4870b, 4874b, 4875b, 4897b, 4898b, 4900b, 4905b, 4909b, 4914b, 4921b, 4924b, 4927b, 4936b, 4972b, 4984b, 4988b, 4999b, 5004b, 5005b, 5010b, 5017b, 5018b, 5021b, 5027b, 5029b, 5030b, 5036b, 5050b, 5051b, 5066b, 5083b, 5096b, 5110b, 5135b, 5156b, 5166b, 5171b, 5173b, 5174b, 5187b, 5208b, 5216b, 5226b, 5234b, 5243b, 5249b, 5258b, 5260b, 5262b, 5273b, 5291b, 5298b, 5316b, 5324b, 5328b, 5334b, 5340b, 5341b, 5366b, 5375b, 5386b, 5392b, 5396b, 5450b, 5459b, 5461b, 5467b, 5473b, 5477b, 5487b, 5491b, 5495b, 5497b, 5505b, 5506b, 5507b, 5509b, 5516b, 5522b, 5528b, 5532b, 5536b, 5541b, 5547b, 5549b, 5550b, 5555b, 5564b, 5588b, 5599b, 5607b, 5610b, 5611b, 5628b, 5630b, 5636b, 5641b, 5644b, 5649b, 5651b, 5656b, 5661b, 5662b, 5683b, 5694b, 5699b, 5706b, 5707b, 5709b, 5711b, 5713b, 5716b, 5731b, 5739b, 5745b, 5766b, 5768b, 5779b, 5787b, 5788b, 5799b, 5807b, 5815b, 5820b, 5827b, 5828b, 5830b, 5832b, 5833b, 5836b, 5850b, 5858b, 5859b, 5877b, 5891b, 5894b, 5901b, 5911b, 5935b, 5938b, 5939b, 5941b, 5979b

43 1275b, 1329b, 1389b, 1472b, 1581b, 1636b, 1764b, 1953b, 2029b, 2117b, 2261b, 2338b, 2399b, 2456b, 2540b, 2586b, 2670b, 2751b, 2818b, 2914b, 3016b, 3124b, 3244b, 3352b, 3424b, 3550b, 3632b, 3708b, 3771b, 3899b, 3964b, 4066b, 4155b, 4220b, 4380b, 4497b, 4582b, 4680b, 4734b, 4825b, 4888b, 4950b, 5026b, 5114b, 5233b, 5322b, 5411b, 5504b, 5543b, 5621b, 5675b, 5736b, 5818b, 5871b,

MOSP-02.qxd

5/11/04

2:38 PM

44

Page 44

2. Metrical types and positions

Table 2.24. Distribution of PX#Px and Px#Px in Beowulf

Px#Px PX#Px

a-verse with double alliteration

a-verse with single alliteration

b-verse

115 (13%) 5 (14%)

268 (31%) 9 (26%)

488 (56%) 21 (60%)

Table 2.25. Distribution of P#xPx (#x ⫽ prefix or independent word) a-verse with double alliteration

a-verse with single alliteration

b-verse

141 (49%)

27 (10%)

115 (41%)

For verse lists, see under Tables 2.26 and 2.27

Table 2.26. Distribution of P#xPx (#x ⫽ prefix) a-verse with double alliteration

a-verse with single alliteration

b-verse

38 (25%)

24 (16%)

84 (58%)

a-verse with double alliteration: 128a, 538a, 705a, 1365a, 1436a, 1655a, 1792a, 1907a, 2087a, 2276a, 2302a, 2518a, 2571a, 2645a, 2676a, 2689a, 2976a, 3203a, 3322a, 3599a, 3621a, 3754a, 4056a, 4132a, 4176a, 4278a, 4341a, 4348a, 4485a, 4785a, 4900a, 5311a, 5323a, 5417a, 5528a, 5698a, 5753a, 5760a; a-verse with single alliteration: 191a, 231a, 296a, 456a, 712a, 1152a, 1170a, 1218a, 1253a, 1513a, 1547a, 1825a, 1917a, 2162a, 2169a, 2347a, 2517a, 3333a, 3336a, 3463a, 3628a, 3781a, 4081a, 4890a; b-verse: 154b, 166b, 230b, 282b, 421b, 427b, 478b, 517b, 530b, 574b, 624b, 670b, 680b, 775b, 1082b, 1090b, 1240b, 1246b, 1426b, 1476b, 1497b, 1499b, 1524b, 1528b, 1531b, 1534b, 1592b, 1660b, 1670b, 1768b, 1804b, 1966b, 1969b, 2132b, 2188b, 2194b, 2201b, 2284b, 2458b, 2471b, 2522b, 2749b, 2760b, 2787b, 2941b, 3140b, 3153b, 3309b, 3387b, 3443b, 3459b, 3476b, 3514b, 3600b, 3913b, 4048b, 4092b, 4156b, 4222b, 4253b, 4324b, 4342b, 4662b, 4741b, 4743b, 4749b, 4901b, 5009b, 5025b, 5059b, 5076b, 5131b, 5165b, 5197b, 5248b, 5290b, 5292b, 5301b, 5408b, 5415b, 5574b, 5742b, 5752b, 5816b

Table 2.27. Distribution of P#xPx (#x ⫽ independent word) a-verse with double alliteration

a-verse with single alliteration

b-verse

103 (75%)

3 (2%)

31 (23%)

a-verse with double alliteration: 12a, 37a, 62a, 85a, 131a, 196a, 249a, 288a, 354a, 389a, 435a, 556a, 607a, 608a, 747a, 827a, 836a, 874a, 979a, 981a, 1001a, 1004a, 1035a, 1283a, 1343a, 1373a, 1385a, 1392a, 1478a, 1525a, 1582a, 1593a, 1701a, 1836a, 1857a, 1863a, 1906a, 1936a, 2010a, 2047a, 2053a, 2074a, 2086a, 2127a, 2397a, 2476a, 2559a, 2707a, 2739a, 2750a, 2788a, 2793a, 2861a, 2915a, 2921a, 2944a, 3047a, 3095a, 3128a, 3212a, 3226a, 3288a, 3330a, 3354a, 3413a, 3429a, 3626a, 3681a, 3739a, 3740a, 3762a, 4114a, 4287a, 4367a, 4373a, 4431a, 4514a, 4537a, 4601a, 4782a, 4792a, 4806a, 4815a, 4873a, 4880a, 4884a, 4922a, 5168a, 5300a, 5329a, 5374a, 5470a, 5474a, 5512a, 5515a, 5534a, 5648a, 5656a, 5673a, 5749a, 5863a, 5886a, 5910a; a-verse with single alliteration: 420a, 3405a, 3877a; b-verse: 115b, 213b, 246b, 661b, 685b, 719b, 726b, 759b, 803b, 864b, 888b, 1199b, 1489b, 1562b, 1585b, 1647b, 1858b, 1887b, 2865b, 3298b, 3720b, 3813b, 3853b, 3936b, 4266b, 4584b, 4962b, 5306b, 5435b, 5849b, 5957b

MOSP-02.qxd

5/11/04

2:38 PM

Page 45

2.1. Type A1

45

Table 2.28. Distribution of P#xxPx, P#xxxPx, P#xxxxPx, P#xxxxxPx, and P#xxxxxxPx a-verse with double alliteration a-verse with single alliteration b-verse P#xxPx 143 (57%) P#xxxPx 128 (87%) P#xxxxPx 73 (85%) P#xxxxxPx 20 (91%) P#xxxxxxPx 5 (100%)

28 (11%) 3 (2%) 1 (1%) 0 (0%) 0 (0%)

82 (32%) 16 (11%) 12 (14%) 2 (9%) 0 (0%)

P#xxPx (see under Tables 2.15 and 2.16 above) P#xxxPx a-verse with double alliteration: 66a, 107a, 153a, 194a, 216a, 292a, 302a, 310a, 312a, 320a, 324a, 379a, 382a, 394a, 437a, 449a, 477a, 504a, 527a, 543a, 555a, 592a, 663a, 713a, 741a, 754a, 777a, 783a, 797a, 805a, 822a, 831a, 911a, 919a, 933a, 1048a, 1080a, 1104a, 1136a, 1138a, 1161a, 1169a, 1222a, 1537a, 1576a, 1598a, 1614a, 1754a, 1762a, 1888a, 1943a, 2060a, 2067a, 2091a, 2101a, 2103a, 2124a, 2156a, 2203a, 2226a, 2265a, 2292a, 2335a, 2386a, 2427a, 2442a, 2742a, 2770a, 2784a, 2790a, 2827a, 2873a, 2884a, 2940a, 3009a, 3124a, 3156a, 3160a, 3161a, 3213a, 3705a, 3709a, 3841a, 3861a, 3954a, 4061a, 4063a, 4130a, 4158a, 4159a, 4355a, 4460a, 4521a, 4583a, 4710a, 4714a, 4775a, 4945a, 4967a, 4976a, 5002a, 5098a, 5115a, 5192a, 5350a, 5362a, 5425a, 5453a, 5460a, 5484a, 5567a, 5608a, 5630a, 5632a, 5688a, 5703a, 5756a, 5758a, 5762a, 5766a, 5779a, 5781a, 5808a, 5809a, 5824a, 5828a, 5880a, 5939a; a-verse with single alliteration: 995a, 3025a 3659a; b-verse: 1066b, 1085b, 1399b, 2418b, 2445b, 2836b, 3687b, 3855b, 3981b, 4515b, 4783b, 4873b, 5191b, 5449b, 5670b, 5772b P#xxxxPx a-verse with double alliteration: 26a, 211a, 273a, 407a, 484a, 500a, 582a, 593a, 658a, 707a, 723a, 774a, 786a, 842a, 851a, 962a, 999a, 1012a, 1070a, 1189a, 1402a, 1404a, 1458a, 1548a, 1628a, 1713a, 1782a, 1835a, 1837a, 1932a, 2048a, 2071a, 2182a, 2195a, 2381a, 2467a, 2530a, 2560a, 2568a, 2692a, 2753a, 2761a, 2765a, 2951a, 2970a, 3140a, 3220a, 3343a, 3523a, 3579a, 3640a, 3675a, 3829a, 4229a, 4255a, 4435a, 4569a, 4617a, 4626a, 4649a, 4877a, 4966a, 5365a, 5503a, 5548a, 5569a, 5573a, 5593a, 5677a, 5733a, 5767a, 5775a, 5873a; a-verse with single alliteration: 3960a; b-verse: 46b, 528b, 1599b, 2151b, 2206b, 3138b, 3296b, 3527b, 3609b, 5529b, 5776b, 5812b P#xxxxxPx a-verse with double alliteration: 921a, 928a, 938a, 1672a, 2027a, 2409a, 2491a, 2654a, 2846a, 3239a, 3376a, 3564a, 4100a, 4868a, 4963a, 5159a, 5163a, 5217a, 5393a, 5510a; b-verse: 2694b, 2894b P#xxxxxxPx a-verse with double alliteration: 2656a, 2716a, 4273a, 5378a, 5475a

counterpart -x#, as with the pairs of PS#xPx and PSx#Px (section 2.1.2), and of PX#xPx and PXx#Px (section 2.1.3). As remarked in section 2.1.3 above, the distinction between prefix and independent word concerns exclusively the monosyllable #x. In this light, it seems reasonable to disregard this morphological distinction and treat the configuration P#xPx as an integrated group when comparing it with the minimally longer configuration, P#xxPx. As it turns out, the configuration P#xPx as a whole (Table 2.25) falls short of the configuration P#xxPx (Table 2.12; section 2.1.3) in terms of the incidence of the a-verse with double alliteration. We may plausibly suspect that the presence of an additional unstressed syllable x has to do with such a distinction. We may accordingly wish to consider longer sequences, #xxx, #xxxx, #xxxxx, and the like, in order to ascertain the empirical adequacy of the above account based on the number of unstressed syllables involved. These longer strings favour the a-verse with double alliteration to such an overwhelming extent that it is realised by more than four verses out of every five, as substantiated in Table 2.28. This increase may overall be

MOSP-02.qxd

5/11/04

2:38 PM

Page 46

46

2. Metrical types and positions

interpreted as a natural consequence of the extension of the generalisation that the addition of #x results in extensive occurrences of the a-verse with double alliteration. 2.1.6. The configuration PxSx The second lift of type A1 may be occupied by the second element of a compound, as follows: (17) 785a môdarmâgun 2158a sâliglîco 4177b môdgithâhti This configuration has the first drop realised by a monosyllable with the exception of verses 3692a thea uurdegiskefti and 65b eðiligiburdi, which contain two syllables instead. As regards the distribution pattern of the configuration at issue (Table 2.29), the Heliand exhibits the exact reversal of what is observed in Beowulf: while in Beowulf the a-verse accounts for 63 per cent of all occurrences (Suzuki 1996a: 344–5), the b-verse occupies a comparable percentage in the Heliand, a proportion which is analogous to the pattern of the configuration Px#Px shown in section 2.1.5 above. 2.1.7. The graded patterning of the first drop of type A1 2.1.7.1. A synchronic perspective The foregoing examinations have uncovered the varied verse distribution patterns that are effected by the differing ways the first drop of type A1 verses is realised through association with language material. The major variants of the first drop that we have identified above may be listed below by ways of summary primarily in the increasing order of preference for the a-verse with double alliteration, and secondarily in the decreasing order of preference for the b-verse (Table 2.30). The set of variant configurations of type A1 at issue may roughly be grouped into three classes on the basis of the shared relative distribution patterns, although the categorisation is less than clear-cut in some cases. Of greater significance seems an overall gradient patterning obtained, rather than a precise grouping of individual members, as we shall see in due course. Focusing on the variants of the first drop of type A1, the membership of the three classes is specified as follows: (18) Class 1: sx#; xxx#; x#; xx# Class 2: X#; S#; xx#; #x Class 3: x#x; #xx; Xx#; Sx#; X#x; Xxx#; #xxx . . .; S#x The three-way distinction of the varied forms of the first drop of type A1 in the Heliand (as represented in Table 2.30) may be described in overall terms on the basis of the distinct distribution pattern of the a-verse and the b-verse. Class 1 is characterised by the minority of the a-verse with double alliteration. Class 3 is distinguished by the dominance of the a-verse with double alliteration. Characteristic of class 2 is the absence of such marked avoidance of or preference for the a-verse with double alliteration. Classes 1 and 3 thus constitute polar opposites of each other, with class 2 standing between these two extremes. Class 3 is further involved in fine-grained

MOSP-02.qxd

5/11/04

2:38 PM

Page 47

2.1. Type A1

47

Table 2.29. Distribution of PxSx (including two instances of PxxSx) a-verse with double alliteration

a-verse with single alliteration

b-verse

1 (1%)

32 (35%)

58 (64%)

a-verse with double alliteration: 785a; a-verse with single alliteration: 58a, 328a, 333a, 448a, 580a, 1354a, 1780a, 1811a, 2158a, 2661a, 2801a, 3189a, 3304a, 3437a, 3519a, 3529a, 3554a, 3612a, 3690a, 3692a, 3812a, 4595a, 4718a, 4748a, 4902a, 5066a, 5117a, 5307a, 5333a, 5822a, 5844a, 5901a; b-verse: 48b, 65b, 68b, 69b, 151b, 233b, 838b, 891b, 1034b, 1097b, 1169b, 1298b, 1379b, 1387b, 1430b, 1483b, 1508b, 1525b, 1643b, 1653b, 1755b, 1868b, 1881b, 1925b, 1971b, 2249b, 2641b, 2723b, 2800b, 2835b, 2849b, 2890b, 2992b, 3076b, 3315b, 3318b, 3355b, 3373b, 3462b, 3465b, 3553b, 3866b, 3874b, 4177b, 4436b, 4446b, 4464b, 4488b, 4586b, 4786b, 4987b, 5038b, 5302b, 5687b, 5809b, 5913b, 5946b, 5948b

Table 2.30. List of major variants of type A1 variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

Psx#Px Pxxx#Px Px#Px Pxx#Px

1 (9%) 1 (13%) 221 (18%) 22 (19%)

3 (27%) 3 (38%) 210 (17%) 22 (19%)

7 (64%) 4 (50%) 781 (64%) 69 (62%)

PX#Px PS#Px Pxx#Px P#xPx

7 (28%) 7 (41%) 49 (44%) 141 (49%)

3 (12%) 1 (6%) 15 (14%) 27 (10%)

15 (60%) 9 (53%) 47 (42%) 115 (41%)

Px#xPx P#xxPx PXx#Px PSx#Px PX#xPx PXxx#Px P#xxx . . . Px PS#xPx

470 (56%) 143 (57%) 14 (58%) 13 (68%) 7 (70%) 3 (75%) 226 (87%) 15 (94%)

68 (8%) 28 (11%) 4 (17%) 2 (11%) 1 (10%) 1 (25%) 4 (2%) 1 (6%)

303 (36%) 82 (32%) 6 (25%) 4 (21%) 2 (20%) 0 (0%) 30 (12%) 0 (0%)

differentiation predicated on the varying degrees of preference for the a-verse with double alliteration. What is the basis of the differentiation and grouping of these variants, then? We may set up the following five parameters to account for the varied patterning observed: (i) long versus short syllables; (ii) the presence of an unfooted syllable; (iii) monosyllabic versus disyllabic endings; (iv) the number of syllables; (v) independent words versus prefixes. The parameters all concern various properties of syllables in such ways that a long syllable, the existence of a larger number of syllables, and the presence of a disyllabic (rather than monosyllabic) ending or of an independent word (rather than a prefix) contribute to increased preference for the a-verse with double alliteration

MOSP-02.qxd

5/11/04

48

2:38 PM

Page 48

2. Metrical types and positions

in their different ways. The parameters are binary in value, marked versus unmarked; and whatever configuration is specified as marked along a given parameter has a stronger tendency to occur in the a-verse with double alliteration than does its unmarked counterpart. The five parameters, however, are far from equal in effecting increased frequency of the a-verse with double alliteration; on the contrary, the parameters are context-sensitive, and partially ranked in the effects that they bring about. Specifically, we may note the following contextual restrictions. Unstressed syllables divide into two major domains in relation to the first lift that they follow: (i) non-word-initial (posttonic) syllables, that is, the syllables that belong to the same word as the preceding lifted (i.e., primary-stressed) foot (P or px); (ii) wordinitial syllables, that is, the syllables that are separated from the preceding lifted word by the intervening word boundary (#). The parameters apply to the two domains separately. The non-initial syllables are subject to the three parameters: (i) the presence of a long syllable after the lifted foot; (ii) the presence of an unfooted syllable; (iii) the presence of a disyllabic ending in its prosodically integral shape. Parameter (i) applies to the syllables that immediately follow lifted feet (P or px); parameter (ii) becomes relevant only to the syllables standing after the long syllables that are identified by parameter (i); and parameter (iii) concerns whatever syllables that remain intact by foot organisation. Parameters (i) and (ii) are thus hierarchically ordered in such a way that the second applies to the strings that are marked by the first. It should be noted in parenthesis that the presence versus absence of stress does not figure as a critical conditioning factor for association with the first drop of type A1. Although the lexically assigned stress (S) seems responsible for increased incidence of the a-verse with double alliteration for some configurations, the distinction involved is far from conspicuous, and limited to class-internal gradation (section 2.1.7.2). Moreover, the pair of strings Psx# and Pxx# are insensitive even to such a smallscale differentiation. In concrete terms, while the strings -X#, -Xx#, -S#, and -Sx# are all marked by parameter (i) by virtue of the long syllables after the first lift in distinction from -x#, -xx#, and -sx#, the disyllabic sequences -Xx# and -Sx# are further distinguished as marked in opposition to the monosyllabic ones -X# and -S# by parameter (ii) in the presence of the following unfooted syllables; their greater degrees of markedness are reflected correspondingly in their higher incidence of the a-verse with double alliteration than that of their less marked opposites. By parameter (iii), whereby the unfooted sequence -xx# is assigned a marked value in opposition to -x#, Pxx# (class 2) is differentiated from Px# (class 1), as are PXxx# and PSxx# from PXx# and PSx#, respectively (internal to class 3); these configurations with a disyllabic ending show a correspondingly stronger preference for the a-verse with double alliteration than their unmarked opposites. Parameter (iii) deserves a closer consideration as regards the qualification ‘a disyllabic ending in its prosodically integral shape’. The treatment of Pxxx# (class 1) would appear anomalous at first glance: it is used with a lower frequency in the a-verse with double alliteration than the shorter sequence Pxx# (class 2). Given the scarce significance of the short medial syllable -x- in differentiating between the strings Pxx# and Px#, we would expect an analogous situation to obtain for the pair

MOSP-02.qxd

5/11/04

2:38 PM

Page 49

2.1. Type A1

49

Pxxx# and Pxx#. Taking into consideration the prosodic structure involved, however, we may gain a better understanding of the distinct behaviour. As far as the sequence Pxxx# is concerned, the disyllabic ending involved does not constitute a unit in prosodic terms: the second and third syllables (-xx-) are grouped together to form a foot after the one dominating the root syllable (P), as represented below (for a fuller discussion, see section 2.12.4 below): (19) F

F

g

56

P x

x_ x_

Thus, the disyllabic ending is split apart due to the prosodic organisation in force; the sequence in question accordingly cannot be identified as terminating in a disyllabic ending in the same way as the sequence Pxx#, in which the morphological unity is fully respected by its prosodic organisation in terms of foot constituency, whereby the final two syllables are not dissociated from each other by foot construction. Therefore, in strictly prosodic terms the sequence Pxxx# cannot be paired with Pxx#; rather, it has to be grouped together with the sequences ending in -x#. In view of the prosodic distinction of the disyllabic endings, we may better restrict the term disyllabic ending and the notation -xx# to those verses in which the morphological unity of a disyllabic ending is respected in prosodic terms as well (cf. section 1.4). The question arising then is: which particular configuration ending in -x# is the sequence Pxxx# to be matched with? Specifically, why does the sequence in question differ from PXx# (class 3), which is also double-footed as follows: (20) F F g

g

PXx While a long stressed syllable on the one hand and a sequence of a short stressed syllable and another syllable on the other are treated as equivalent for association with the lift as well as with the heavy drop (resolution; section 3.1), such metrical equivalence does not apply to the normal drop. It is prototypically associated with an unstressed syllable or a sequence thereof (section 1.2). In other words, the normal drop is mapped to language material without regard to its foot organisation. Thus, the word form Psx# (class 1) is not treated parallel to the word form PS# (class 2) with respect to the association with the first drop of type A1, as demonstrated above; rather, purely on the dimension of the syllable, the sequence Psx# is counted as unmarked and correspondingly less prominent than PS# by parameter (i) due to the status of -sbeing a short syllable, although the disyllable -sx# and the monosyllable -S# are bimoraic and equally organised into a foot. By the same token, for the association with the first drop of type A1 the sequence Pxxx# is determined as less prominent than PXx# because of the short syllable (-x-) standing immediately after P, although these two sequences are indistinguishable as regards their foot organisation.

MOSP-02.qxd

5/11/04

50

2:38 PM

Page 50

2. Metrical types and positions

Characterised as the closest parallel to Pxxx# should therefore be the configuration Pxxx#. This configuration, however, does not happen to be used immediately before the second lift, although we find a few instances of the configuration Pxxx#x . . . Px, namely 611a sâligoron undar them gisîðea and 5973b endi hêlegoda sie alle. The divergent treatment of a long syllable and a sequence of a short syllable and another syllable may also be discernible in the contrastive behaviour of verse 972a on the one hand and verses 988a and 3290b on the other. These are instances of the configurations Psxx#Px and PSxx#Px, respectively. (21) 972a thiodgumono bezto 988a diurlîcara dûbun 3290b kindiungumu manne While -gumono, -lîcara, and -iungumu are all footed, the status of the disyllabic endings involved is different. In -lîcara and -iungumu, the disyllabic ending remains undisturbed by foot organisation; by contrast, the foot organisation of -gumono stands in conflict with the morphological unity of the disyllabic ending, whereby the ending -ono is disintegrated by foot structure: the first constituent of the disyllabic ending serves as the right branch of the foot headed by the preceding secondarystressed short syllable -s-, and the second constituent of the ending remains unfooted. Accordingly, while the configuration PSxx# has a disyllabic ending standing in harmony with prosodic organisation, the shorter counterpart Psxx# suffers from a mismatch between morphological and prosodic organisation, as represented below: (22) F F g

g

P S x_x_

F

F

g

56

P s

x_ x_

Terminating thus in a disyllabic ending in its integral shape, the configuration PSxx# is determined as marked in contrast to Psxx# by parameter (iii), and therefore it is distinguished from Psxx# by a stronger preference for the a-verse with double alliteration. More specifically, PSxx# may fall under the upper division of class 3 as does PXxx#, because it should rank higher than PSx# by virtue of the markedness for the disyllabic ending (see Table 2.31 below). On the other hand, the sequence Psxx# should be grouped with Pxxx# (class 1) due to the foot organisation commonly involved. The effects of the above three parameters are cumulative: the larger the number of positive values of parameters, the greater the likelihood of manifesting the a-verse with double alliteration. The cumulative effects in question are illustrated in Table 2.31, in which the posttonic sequences that are internal to the lifted words are compared. Going beyond the domain of medial and word-final syllables, we take into account the syllables that are separated from the first lift by the word boundary. A string provided with a word-initial syllable (#x) is rated as more favourable to the a-verse with double alliteration than the comparable one without. Accordingly, the strings -x#x, -xx#x, -X#x, -xx#x, -S#x, -Xx#x, and -Sx#x are distinguished by greater

MOSP-02.qxd

5/11/04

2:38 PM

Page 51

2.1. Type A1

51

Table 2.31. Cumulative effects of the three parameters (i) long

parameter

(ii) unfooted syllable

class 1

sx# x# xx# xxx# sxx#

class 2

X# xx# S#

⫹

Xx# Sx#

⫹ ⫹

⫹ ⫹

Xxx# Sxx#

⫹ ⫹

⫹ ⫹

class 3

(iii) disyllabic ending

total value 0 0 0 0 0

⫹ ⫹

1 1 1 2 2

⫹ ⫹

3 3

Key: ‘⫹’ ⫽ identified as marked

frequencies of the a-verse with double alliteration from their respective shorter counterparts lacking #x, namely, -x#, -xx#, -X#, -xx#, -S#, -Xx#, and -Sx#. The domain following the word boundary is subject to the two parameters: (i) the number of syllables; (ii) the opposition between prefix and independent word. The parameter of the number of syllables specifies the probability of the a-verse with double alliteration in such a way that the larger the number of syllables found after the word boundary #, the higher the incidence of the a-verse with double alliteration. As an overarching generalisation, the first parameter applies to the entire syllables and syllable sequences that come after the word boundary # without regard to the difference in their morpholexical status. On the other hand, the second parameter, namely the opposition between prefix and independent word, concerns monosyllables to the exclusion of syllable sequences, because prefixes are largely monosyllabic and therefore the opposition loses its relevance for polysyllabic sequences. Thus, the second parameter, sensitive to the morpholexical distinction involved, serves as a special condition on monosyllables in such a way that an independent word is more likely to be found in the a-verse with double alliteration than a prefix. In addition to the gradation patterning of the sequence #x . . . substantiated in Table 2.28 above, the significance of the number of syllables is born out also by the comparison of strings ending in #x with those in #xx, notably (i) -x#xx versus -x#x, (ii) -X#xx versus -X#x, (iii) -xx#xx versus -xx#x, and (iv) -Xx#xx versus -Xx#x.2 The overall picture that we find in Table 2.32 confirms the generalisation: the disyllables

2

In the complete absence of examples of the configuration PSx#xxPx in the corpus, we cannot compare -Sx#xx with -Sx#x.

MOSP-02.qxd

5/11/04

52

2:38 PM

Page 52

2. Metrical types and positions

Table 2.32. Distribution of Px#xxPx, PX#xxPx, Pxx#xxPx, and PXx#xxPx, compared with Px#xPx, PX#xPx, Pxx#xPx, and PXx#xPx, respectively

Px#xxPx Px#xPx PX#xxPx PX#xPx

a-verse with double alliteration

a-verse with single alliteration

b-verse

242 (82%) 470 (56%)

16 (5%) 68 (8%)

39 (13%) 303 (36%)

8 (100%) 7 (70%)

0 (0%) 1 (10%)

0 (0%) 2 (20%)

Pxx#xxPx Pxx#xPx

15 (94%) 19 (73%)

0 (0%) 1 (4%)

1 (6%) 6 (23%)

PXx#xxPx PXx#xPx

10 (100%) 13 (72%)

0 (0%) 1 (6%)

0 (0%) 4 (22%)

Px#xxPx a-verse with double alliteration: 3a, 16a, 33a, 95a, 98a, 127a, 141a, 165a, 174a, 185a, 223a, 239a, 299a, 314a, 356a, 367a, 392a, 397a, 415a, 464a, 474a, 479a, 534a, 552a, 583a, 643a, 646a, 649a, 684a, 688a, 701a, 720a, 778a, 795a, 830a, 845a, 897a, 942a, 964a, 1029a, 1051a, 1067a, 1094a, 1120a, 1143a, 1148a, 1154a, 1164a, 1167a, 1174a, 1177a, 1185a, 1213a, 1237a, 1260a, 1276a, 1284a, 1293a, 1338a, 1341a, 1350a, 1377a, 1386a, 1439a, 1449a, 1452a, 1453a, 1481a, 1483a, 1491a, 1505a, 1509a, 1535a, 1570a, 1571a, 1583a, 1671a, 1694a, 1711a, 1733a, 1737a, 1753a, 1758a, 1768a, 1804a, 1830a, 1862a, 1918a, 1937a, 1942a, 2012a, 2036a, 2043a, 2049a, 2097a, 2106a, 2123a, 2159a, 2172a, 2185a, 2189a, 2249a, 2275a, 2288a, 2304a, 2309a, 2351a, 2375a, 2411a, 2434a, 2440a, 2470a, 2474a, 2549a, 2554a, 2577a, 2584a, 2602a, 2608a, 2612a, 2715a, 2717a, 2848a, 2909a, 2927a, 2966a, 2998a, 3003a, 3007a, 3013a, 3022a, 3050a, 3055a, 3087a, 3089a, 3110a, 3130a, 3151a, 3166a, 3206a, 3218a, 3262a, 3309a, 3313a, 3332a, 3334a, 3362a, 3385a, 3386a, 3468a, 3472a, 3473a, 3481a, 3488a, 3513a, 3530a, 3542a, 3559a, 3566a, 3572a, 3584a, 3601a, 3609a, 3620a, 3650a, 3667a, 3674a, 3697a, 3722a, 3760a, 3817a, 3832a, 3845a, 3858a, 3862a, 3872a, 3906a, 3934a, 3940a, 3945a, 3976a, 4005a, 4023a, 4039a, 4103a, 4188a, 4212a, 4220a, 4224a, 4316a, 4328a, 4340a, 4360a, 4361a, 4505a, 4506a, 4524a, 4575a, 4578a, 4603a, 4614a, 4638a, 4645a, 4675a, 4720a, 4797a, 4943a, 5001a, 5037a, 5041a, 5043a, 5054a, 5078a, 5119a, 5152a, 5187a, 5229a, 5274a, 5280a, 5284a, 5317a, 5318a, 5336a, 5369a, 5372a, 5384a, 5424a, 5430a, 5432a, 5447a, 5596a, 5625a, 5652a, 5701a, 5726a, 5730a, 5786a, 5795a, 5801a, 5888a, 5890a, 5952a; a-verse with single alliteration: 30a, 408a, 1430a, 2328a, 2420a, 2421a, 2488a, 2598a, 2631a, 4164a, 4236a, 4321a, 4349a, 4470a, 4523a, 4705a; b-verse: 224b, 293b, 465b, 586b, 855b, 1023b, 1213b, 1229b, 1696b, 1841b, 1990b, 2019b, 2489b, 2724b, 2861b, 2944b, 2966b, 3206b, 3886b, 3942b, 4064b, 4112b, 4146b, 4184b, 4228b, 4398b, 4517b, 4558b, 4596b, 4750b, 4915b, 5081b, 5140b, 5468b, 5479b, 5723b, 5785b, 5813b, 5937b Px#xPx (see under Table 2.12 above) PX#xxPx a-verse with double alliteration: 1633a, 2241a, 2902a, 4086a, 4503a, 4511a, 4760a, 4793a PX#xPx (see under Table 2.11 above) Pxx#xxPx a-verse with double alliteration: 83a, 410a, 3786a, 3837a, 3970a, 4549a, 4791a, 4835a, 4989a, 5006a, 5211a, 5613a, 5618a, 5657a, 5883a; b-verse: 776b Pxx#xPx (see under Table 2.9 above) PXx#xxPx a-verse with double alliteration: 453a, 462a, 550a, 1589a, 3431a, 3756a, 4357a, 4541a, 4920a, 5471a PXx#xPx (see under Table 2.8 above)

MOSP-02.qxd

5/11/04

2:38 PM

Page 53

2.1. Type A1

53

after the boundary occur with higher frequency in the a-verse with double alliteration than the monosyllables. Having identified the parameters that affect the verse distribution patterning through their varied degrees of preference for the a-verse with double alliteration, we may ask whether a unified principle is at work behind the multiplicity of conditioning factors involved. As discussed above, the parameters all but one concern the quantity of syllable sequences that occupy the first drop: syllable length and the number of syllables. We may therefore be led to suspect that it is a question of relative prominence of the metrical position as it is realised through association with various language material. The long syllable in isolation on the one hand and a long sequence of syllables on the other contribute in their own ways to enhancing prominence of the position involved. Also pertinent to relative prominence seems the opposition between prefix and independent word: since prefixes are incapable of occurring independently of words that they are attached to, they are regarded as less autonomous and hence less prominent than words. Thus, we may generalise the correlation between prominence and distribution pattern as follows: the greater the prominence is with which the position is realised, the likelier it is to occur in the a-verse with double alliteration. The apparently distinct parameters accordingly constitute different manifestations of the same underlying property, namely, prominence. Against the background of the overall graded patterning of the first drop of type A1 that we have identified as characteristic of the Heliand metre, we may draw attention to the following three most outstanding properties that distinguish it from the Beowulf metre. First, in Beowulf the variety of the first drop of type A1 is categorised into the two binary opposed classes, the normal drop (unmarked) and the heavy drop (marked). The heavy drop is sharply differentiated from the normal counterpart by its maximal preference for the a-verse with double alliteration and conversely its minimal preference for the b-verse. These two kinds of drop are categorically distinguished in terms of association with language material: the heavy drop is occupied by a foot, while the normal drop is filled by unfooted material. The presence of a foot thus serves as a primary organising feature in Beowulf. By contrast, the corresponding distinction is graded rather than dichotomised in the Heliand: more than two classes are recognisable that are differentiated by their gradient preferences for the a-verse with double alliteration. Second, while Beowulf does not allow for the sequences -Xx# and -Sx# as realisations of the first drop of type A1, the Heliand legitimises them as metrical. Third, the Heliand not only expands the range of linguistic forms for the first drop of type A1 as shown in Tables 2.30 and 2.33 (see section 2.1.7.2 below), it also reorganises the existing expressions in regard to their verse distribution. At issue here are the strings -sx# and -S#: neither of these shows a maximal preference for the a-verse with double alliteration in the Heliand, in marked contrast to Beowulf (compare Table 2.34, section 2.1.7.2); what is more, among the variant strings listed in Table 2.30, -sx# occurs in the a-verse with double alliteration with the lowest frequency. In this way, the presence of stress does not constitute a highly significant basis of categorisation in the Heliand. 2.1.7.2. A diachronic perspective In this section, we explore from a diachronic perspective the motivation and mechanism of the reorganisation of the metre that gave rise to the fully graded patterning of the first drop of type A1 in the Heliand, brought to light in the preceding section.

MOSP-02.qxd

5/11/04

2:38 PM

54

Page 54

2. Metrical types and positions

Table 2.33. Linguistic realisations of the first drop of type A1, with reference to the medial and word-final syllables Heliand

a-verse (aa)

a-verse (ax)

b-verse

total

class 1

-x# -xx# -sx#

221 (18%) 22 (19%) 1 (9%)

210 (17%) 22 (19%) 3 (27%)

780 (64%) 69 (62%) 7 (64%)

1211 (100%) 113 (100%) 11 (100%)

class 2

-X#

7 (28%)

3 (12%)

15 (60%)

25 (100%)

-S#

7 (41%)

1 (6%)

9 (53%)

17 (100%)

-Xx#

14 (58%)

4 (17%)

6 (25%)

24 (100%)

-Sx# -X#x

13 (68%) 7 (70%)

2 (11%) 1 (10%)

4 (21%) 2 (20%)

19 (100%) 10 (100%)

-S#x

15 (94%)

1 (6%)

0 (0%)

16 (100%)

class 3

Examples: -x#: 2349a manno barnun -xx#: 1939a hêlaga helpa -sx#: 2037a stênfatu sehsi -X#: 358a uualdand uuelda -S#: 3838a sôðspel sagde -Xx#: 106a uualdandes uuilleon -Sx#: 2492a mancunnie mildie -X#x: 45a uuîdost giuualdan -S#x: 25a godspell that guoda

Since the reorganisation characteristic of the Heliand largely concerns medial and word-final syllables, we primarily focus on such strings in the following discussion. The gradation at issue is reproduced in Table 2.33 and may as well be compared with the treatment of the same position in Beowulf, as summarised in Table 2.34. We may account for the emergence and establishment of the gradation pattern in the Heliand by considering other ways of reorganisation that would have been available to the poet. First, the strings -xx#, -sx#, -X#, and -S# could have been treated as equal as they all are footed, and thus distinguished from the unfooted -x# (option 2; see Table 2.35 below). Second, the monosyllabic -x#, -X#, and -S# could have been grouped together in distinction from the disyllabic -xx# and -sx# (option 3; see Table 2.35 below). Third, the second elements of compounds -S# and -sx# could have been differentiated from -x#, -X#, and -xx#, fully in accordance with traditional practice and still motivated by making a distinction between lexical and non-lexical feet (option 4; see Table 2.35 below). And finally, all strings could have been made equivalent without discrimination (option 5; see Table 2.35 below). The Heliand poet thus had at his disposal at least five alternative ways of reorganising the metrical practice that he inherited, as summarised in Table 2.35. Of these five options, option 5 would have been found least appealing from an aesthetic point of view if nothing else, and had to be rejected immediately: it would otherwise have created an unordered state of affairs by lumping together the multiplicity of forms without any differentiation.

MOSP-02.qxd

5/11/04

2:38 PM

Page 55

2.1. Type A1

55

Table 2.34. Linguistic realisations of the first drop of type A1 in Beowulf, with reference to the medial and word-final syllables Beowulf

a-verse (aa) a-verse (ax) b-verse

normal drop, -x# 115 (13%) type A1 (/ ⫻ / ⫻) -X# 5 (14%) -xx# 9 (30%) heavy drop, -sx# type A2a (/ \ / ⫻) -S# –

-Xx# -Sx# -S#x

67 (96%) 83 (88%) 0 (0%) 0 (0%) 6 (100%)

268 (31%) 9 (26%) 4 (13%)

total

488 (56%) 871 (100%) 21 (60%) 35 (100%) 17 (57%) 30 (100%)

0 (0%) 4 (4%)

3 (4%) 7 (7%)

70 (100%) 94 (100%)

0 (0%) 0 (0%) 0 (0%)

0 (0%) 0 (0%) 0 (0%)

0 (0%) 0 (0%) 6 (100%)

Examples: -x#: Beo 66a georne hyrdon -X#: Beo 1814a wigend wæ¯ron -xx#: Beo 105b weardode hwile -sx#: Beo 136a morðbeala mare -S#: Beo 1111a swatfah syrce -S#x: Beo 608a gamolfeax ond guðrof

Table 2.35. Options for reorganisation of the Heliand metre with respect to the association of medial and word-final syllables with the first drop of type A1

option 1 option 2 option 3 option 4 option 5

primary distinguishing feature

group I (unmarked)

group II (marked)

syllable length presence of a foot number of syllables status of a foot –

-x#, -xx#, -sx# -x# -x#, -X#, -S# -x#, -X#, -xx# -x#, -xx#, -sx#, -X#, -S#

-X#, -S# -xx#, -sx#, -X#, -S# -xx#, -sx# -S#, -sx# –

In evaluating the remaining alternative solutions (options 1 through 4), we need to take into consideration other varieties of the first drop of type A1 that are allowed in the traditional metre, namely sequences containing a syllable that is separated from the first lift by the word boundary that terminates the lifted word, that is, the sequences that contain the word-initial syllable (#x). As substantiated in Suzuki (1996a: 149–58), the varied realisations of the first drop of type A1 have distinct distribution patterns. By far the most important for our purposes here are the following generalisations: (23) a. The presence of a syllable separated from the first lift by the word boundary (#x) increases the incidence of the a-verse with double alliteration. b. Conversely, medial and word-final syllables (syllables belonging to the same word as the first lift) show a stronger preference for the b-verse. c. The larger the number of syllables involved, the higher the frequency of the a-verse with double alliteration.

MOSP-02.qxd

5/11/04

56

2:38 PM

Page 56

2. Metrical types and positions

We are reminded here that these three rules are in force in the Heliand as well, as we have seen in section 2.1.7.1 above. Of most significance at this point, the distinction involved is gradient, rather than categorical. Therefore these varied expressions of the first drop of type A1 should be characterised as graded realisations of the same metrical category, that is, the normal drop of type A1 in the second position. In this way, while the first drop of type A1 (/ ⫻ / ⫻) is categorically opposed to the heavy drop of type A2a (/ \ / ⫻) that may occur in the same position (see Table 2.34 above), it is in itself a graded category, involved as it is in gradient distinctions determined by association with varying language materials that are unfooted. And those variants of the normal drop that show a strong preference for the a-verse with double alliteration are correspondingly difficult to distinguish from instances of the heavy drop. A sequence of unstressed syllables that counts as most harmonic with the a-verse with double alliteration in the light of the set of generalisations given in (23) above accordingly occurs on the edge of the two categories, the heavy drop and the normal drop: it is open to the two competing scansions, type A2a and type A1. Type A1 verses with its first drop filled by a relatively long string of unstressed syllables (as exemplified under Table 2.36 below) may therefore be identified as non-central members of type A2a by virtue of category extension (Suzuki 1996a: 158). In other words, the two categories, the heavy drop and the normal drop, may merge at their boundaries. Such a loss of category distinction can be self-subverting, because it may potentially lead to a disintegration of the system. The traditional metre then suffered from a loose organisation in regard to the categorisation of the heavy and normal drops for the second position of type A. In addition to such a general deficiency inherent in the system, there was a specific verse form that would have disrupted the wholesale maintenance of the traditional metrical practice. At issue is the sequence PS#xPx: e.g., Beo 608a gamolfeax ond guðrof; Beo 1649a egeslic for eorlum. As it stood in traditional versecraft, this string would have posed a problem in scansion in light of the general constraint on the association of the heavy and normal drops with language materials: the string -S#x does not constitute an adequate realisation of the heavy drop, because it contains an unfooted syllable (x) in addition to a foot; nor is it scannable as an acceptable instance of the normal drop, because it contains a foot (S). The string -S#x would thus have made it difficult to maintain the traditional dichotomy between the two kinds of drop that was predicated on the presence versus absence of a foot. The marginal but acceptable

Table 2.36. The categorisation of the first drop of types A1 and A2a in the traditional metre normal drop -x#, -xx#,

heavy drop -x#x . . ., #x, #xx, #xxx . . .,

-S#x,

Examples (see also the examples given under Table 2.34 above): -x#x: Beo 14a folce to frofre #x: Beo 13a geong in geardum #xx: Beo 8a weox under wolcnum #xxx: Beo 1859a wesan, þenden ic wealde

-S#, -sx#

MOSP-02.qxd

5/11/04

2:38 PM

Page 57

2.1. Type A1

57

use of -S#x as a realisation of the first drop of type A (A1 or A2a) would therefore have required the traditional system to be reintegrated in one way or another. To recapitulate, with respect to the association with language materials the first drop of type A in the traditional metre is organised by two different principles of categorisation, one based on a graded category (the normal drop), the other based on a classical category that is predicated on the presence versus absence of a foot, the necessary and sufficient condition for defining the category of the heavy drop. Furthermore, the two categories intersect in such a way that they may be hardly distinguishable as regards periphery members of the normal drop. The whole categorisation of the relevant materials may thus be represented as in Table 2.36. The Heliand poet then was confronted with the structural problems to solve. He had to organise the first drop of type A in a more integrated way; more specifically, he had to systematise the two heterogeneous principles of organisation and consequently to harmonise the relation between the two kinds of drop; and he needed to provide a proper scansion of the verse form PS#xPx. Further, he found himself standing in radically changed prosodic conditions: the weakening of primary stress in general and the reintroduction of a foot to an unstressed (non-lexical) short disyllable (-(xx)#), and also to an unstressed long monosyllable (-(X)#), both in wordfinal position, as discussed in section 1.3.1 above. Largely conforming to tradition, but at the same time highly sensitive to need for improvement, the poet was thus challenged to reintegrate as fully, coherently, and systematically as possible the following set of prosodic forms as legitimate realisations of the first drop of type A: (24) -x#, -(xx)#, -(X)#, -x#x . . ., #x . . ., -(S)#x, -(S)#, -(sx)# Key: ( ) ⫽ footed In systematising the set of metrical forms at hand, the poet was also required to follow as closely as he could their varying distribution patterns, in particular their distinctive preferences for the a-verse with double alliteration. Put another way, the poet had to respect the traditional hierarchy whereby various syllable sequences were ordered along the parameter of the preference for the a-verse with double alliteration, as summarised in Tables 2.34 and 2.36 above. In reorganising the traditional metre, the poet would have been guided by the following line of reasoning. The traditional organisation of the first drop is partly redundant and partly ambiguous: the verse form with a maximal preference for the a-verse with double alliteration (type A2a) could be interpreted as a most peripheral member of type A1. As discussed extensively in Suzuki (1996a: 149–58), the normal drop is a prototype-based category with gradient distinctions: the more deviant an instance is from the prototype, that is, the unstressed short syllable in word-final position (-x#), the stronger is its preference for the a-verse with double alliteration. Given such a conceptual overlap and redundancy between the two metrical types (A1 and A2a), a reintegration of the traditional system may lead to a simplification of the dual organisation on which they are based by incorporating the more specific metrical category (the heavy drop) to the more general (the normal drop). As a consequence, the first drop of type A came to be organised fully and consistently by a single category, the normal drop, on a gradient, prototype basis. Accordingly,

MOSP-02.qxd

5/11/04

58

2:38 PM

Page 58

2. Metrical types and positions

type A2a (/ \ / ⫻) ceased to be an independent metrical form; instead, it became just a particular realisation of type A1 (/ ⫻ / ⫻) at large. In addition to a purely logical mode of thinking, some specific factors would also have converged to lead the poet to choose the particular metrical recategorisation (option 1, Table 2.35). First, as the string -xx# became footed (-(xx)#), the presence of a foot was rendered incapable of serving as a sufficient feature of the heavy drop. The reintroduction of a non-lexical foot in word-final position would thus have contributed to the elimination of the heavy drop by reinterpreting it as simply a nontypical member of the normal drop. Second, the existence of the sequence #x . . ., which is characterised by an increased incidence of the a-verse with double alliteration much as is type A2a (see (23a) and (23c) above), would have revealed that the foot did not constitute a necessary condition for the heavy drop. Finally, the presence of the problematic string -(S)#x alongside the regular one -(S)# would have convinced the poet that the foot could not be conceptualised as an exclusive occupant of the position for realising a maximal degree of the a-verse with double alliteration. Accordingly, the poet did not choose the foot as a primary feature for organisation in the ways specified by options 2 and 4 in Table 2.35 above. Use of the number of syllables as a principal way of organisation (option 3; see Table 2.35 above) would have proved to be equally unsatisfactory. As particularly problematic would have stricken the three-way distinction of the unstressed disyllables -xx#, -x#x, and #xx. All these consisted of the same number of syllables, yet their distribution patterns were far from identical: #xx showed the strongest preference for the a-verse with double alliteration, followed by -x#x and -xx# in that order. That the number of syllables was incapable of constituting a primary condition was well recognised in the traditional metre: as generalised in (23a) and (23b) above, the syllables separated from the first lift by the word boundary were sharply distinguished from those immediately following the lift by their stronger preference for the a-verse with double alliteration. The remaining candidate then is syllable length (option 1). Why was syllable length chosen as a major organising feature for the category of the first drop of type A in the Heliand? Were there any positive grounds for the choice, apart from the fact that the other options were fraught with difficulty? As demonstrated in Suzuki (1996a: 165–9), the short syllable in word-final position (-x#) is a prototype of the normal drop. Being the only segment that follows the root syllable (P or the resolvable disyllable px) within the same word domain, -x# may be characterised also as a short syllable standing immediately after the root syllable. Given such a characterisation, the prototypical mode of association with the normal drop is obtained by aligning the position with a short syllable on the left edge (that is, the first constituent of the drop is typically a short syllable). Thus, both -X# and -S# turn out to be diametrically opposed to this prototypical property of the normal drop: they are long syllables, located immediately following the root syllable as well as word-finally, precisely the place that may be occupied by the short syllable, the prototypical occupant of the normal drop; as far as the long syllables -X# and -S# are concerned, the normal drop would accordingly have to be aligned with a long syllable on the left edge. In this way, the establishment of syllable length as a primary organising feature was derived largely from the prototype of the normal drop in the traditional metre, and consequently, class 2 (Table 2.33) came into existence as an independent class.

MOSP-02.qxd

5/11/04

2:38 PM

Page 59

2.1. Type A1

59

Of particular interest in this regard is the newly introduced distinction between -x# and -X#. Earlier in the traditional metre, the two entities were scarcely distinguished, as illustrated in Table 2.34, for good reason: prosodically, they were of the same status, both being unfooted. The normal drop in the traditional metre was occupied by unfooted material without regard to syllable length. Accordingly, -x# and -X# were allowed to fill the normal drop without differentiation (see Table 2.34). By contrast, in the Heliand metre, syllable length figured as a central organising feature, as argued above. With such metrical reorganisation in force, -x# and -X#, undifferentiated as they had been traditionally, had to be treated differently in corresponding fashion. The emergence of -X# as an entity distinct from -x# may thus be explained as a consequence of the establishment of syllable length as a primary organising feature in the reconstituted metre of the Heliand. By contrast, the strings -xx# and -sx# would appear to have greater similarity to -x#: they start with a short syllable immediately following the root syllable, as does the prototype -x#; and they usually (though not always) end in a monomoraic word-final syllable. In other words, the strings -xx# and -sx# would not count as marked in the same sense as -X# and -S#, because they do not begin with a long syllable, the structural opposite of the short syllable, the prototype of the normal drop: for -xx# and -sx#, as for -x#, association with the normal drop would thus be implemented by alignment with a short syllable on the left edge. By virtue of the prototypical association determined in terms of syllable length, then, -xx# and -sx# were grouped together with -x# as members of the unmarked class (if not its central members), and accordingly the membership of class 1 was established as in Table 2.33. The revaluation of -sx# as a non-central variant of -x#, however, constituted a significant departure from traditional practice: earlier -sx#, alongside -S#, had been a member of a marked category in opposition to -x# (see Table 2.34 above); but now it belonged to the same unmarked category as -x#, removed thereupon from -S# (see Table 2.33 above). Such a radical reconfiguration arose as a result of choosing syllable length as a new organising feature in the innovated metre of the Heliand. This kind of break with tradition, however, would have been intolerable had it not been for some measure of compensation available. The much diminished use of poetic compounds ending in -sx# or -S# in the Heliand (cf. Lehmann 1956: 105–6; for a fuller discussion on infrequent occurrences of such compounds in the first two positions of type A1, see section 2.2.1 below), whatever its motivation, would certainly have contributed to keeping the deviation from tradition to a minimum in favour of carrying out the innovation. We next turn to consider the strings containing an additional syllable or syllables that do not belong to the same word as the first lift: -x#x . . ., #x . . ., -(S)#x. How can we distinguish -x#x . . . from -x# on the one hand and -(S)#x from -(S)# on the other? Obviously, syllable length is irrelevant here. At this point, the number of syllables after the word boundary # figures as a secondary feature, a parameter which is supplementary to that of syllable length. Accordingly, -x#x is distinguished from -x# by the presence of an additional syllable #x, as regulated in the traditional metre (23a); by the same token, #x, #xx, #xxx, and so forth are differentiated in ranking according to the number of syllables involved (23c). As observed above, the primary organising feature, syllable length, derives from the short word-final syllable identified as the prototype of the normal drop. This would entail that any syllables that are

MOSP-02.qxd

5/11/04

60

2:38 PM

Page 60

2. Metrical types and positions

separated from the preceding first lift by the word boundary add to degrees of nonprototypicality or markedness. Exactly the same numerical distinction may now extend to the pair of -(S)#x and (S)# in the Heliand metre: differentiated from (S)# by the presence of #x, the string -(S)#x shows a greater propensity to appear in the a-verse with double alliteration. Such a generalisation now applicable to the lexical-stressed syllable (a word-final foot built over the second element of a compound), however, was unavailable to the traditional metre because of the categorical distinction between heavy and normal drops: as pointed out above, the string -S#x defied scansion in traditional versecraft; upon reduction of the heavy drop to the normal, the generalisation became accessible to the Heliand poet, who identified -S#x as a marked variant of -x#x. Thus, the string -S#x came to be differentiated from class 1 by virtue of syllable length (-S#), as well as from class 2 by virtue of the presence of the word-initial unstressed syllable (#x), and consequently was integrated into the whole system as a member of a separate class, in parallel to the sequence -X#x. The elimination of the heavy drop as an independent category for the first drop of type A had further notable consequences. In the traditional metre, the syllables qualified for constituting a heavy drop were unqualified for forming a normal drop or a part thereof, and vice versa; accordingly, a combination of members of the two distinct classes of syllables was totally disallowed. The reorganisation of metrical practice, however, obliterated such a dichotomy in categorisation, and consequently made possible a combination of a long medial syllable (footed) and a short final syllable (unfooted). More specifically, the strings -Sx# and -Xx# became qualified as a metrical realisation of the first drop of type A1. The use of these new strings, however, had to be harmonised with the existing set of realisation forms of the first drop of type A1. Due to the word-final unstressed syllable involved -x#, that is, the presence of x before rather than after the word boundary, the strings -Sx# and -Xx# were distinguished from -S#x and -X#x, respectively, by a weaker preference for the a-verse with double alliteration, much as -xx# was contrasted with -x#x by the same parameter. Thus, in addition to being distinguished from -S# and -X# (class 2) by the presence of an extra syllable (the second parameter), class 3 came to be further differentiated on the basis of the opposition between wordinitial and word-final unstressed syllables or the location of the additional syllable that followed -S and -X in accordance with the traditional constraints, given in (23) above. The class distinction between -Sx#/-Xx# and -S#/-X# thus obtained, however, is not fully paralleled by that involving the pair -xx#/-x#: the strings -Sx# and -Xx# are differentiated from -S# and -X# by a higher incidence of the a-verse with double alliteration, whereas -xx# and -x# are hardly distinguishable in this respect (see Table 2.33). There is good reason for making a distinction between -Sx#/-Xx# and -S#/-X# at the cost of missing a perfect parallel with the pair -xx#/-x#. The strings -Sx# and -Xx#, previously excluded from the system, were acknowledged anew as its legitimate members, in contrast to -S#/-X#, which had been fully metrical in traditional practice as well. Accordingly, sharply differentiated from the traditionally accepted -S#/-X#, the strings -Sx# and -Xx# were incorporated into the system as marked members, and they thus constituted a class of their own (class 3), thereby largely differentiated from class 2. The pair -xx#/-x#, however, failed to be implicated in such a differentiation in membership in conformity with the inherited practice whereby the two strings had

MOSP-02.qxd

5/11/04

2:38 PM

Page 61

2.1. Type A1

61

not been established as sharply distinguished entities. As substantiated in Table 2.34 above, the two strings occur in the b-verse with similar frequency, although -xx# shows a greater preference for the a-verse with double alliteration than -x#. Classes 2 and 3 are subject to class-internal gradation based on the distinction between -S and -X. As indicated in Table 2.33, the strings -S#, -Sx#, and -S#x show a stronger preference for the a-verse with double alliteration than the strings -X#, -Xx#, and -X#x, respectively. We may generalise accordingly that the status of stress, that is, lexical or non-lexical, plays a certain role in determining verse preference patterning: lexical stress induces a greater frequency of the a-verse with double alliteration. As pointed out in section 2.1.7.1 above, however, stress does not seem significant insofar as the relation between -xx# and -sx# (class 1) is concerned. The role of stress is thus limited to the marked classes, classes 2 and 3, to the exclusion of the unmarked class, class 1. Such a limited significance of stress may be understood as a consequence of the recategorisation of the first drop of type A whereby syllable length was established as a primary organising feature at the expense of stress, which had served as a major distinguishing feature so far in the traditional metre, as discussed above. In formal terms, then, we may characterise the decrease in significance of stress as resulting from a reranking of parameters whereby stress came to be subordinated to syllable length through the motivation and mechanism detailed above. In this way, the gradient class distinction of the first drop of type A1 in the Heliand metre as represented in Table 2.33 came to be established by the complex of decisions that were motivated metrically as well as phonologically.3 2.1.8. The configuration PS#px We have yet to consider a further variant of type A1, the so-called subtype A1s (‘s’ for short; Sievers’s type A2k). Characteristic of this configuration is that the second and third positions are always occupied by the second element of a compound and the unresolved short syllable, respectively (cf. Kauffmann 1887: 297; Hofmann 1991, II: 59–60): (25) 792a mancraft mikil 2382a friðubarn godes 5194b mênuuerk frumid Table 2.37. Distribution of PS#px a-verse with double alliteration

a-verse with single alliteration

b-verse

13 (33%)

9 (23%)

17 (44%)

a-verse with double alliteration: 792a, 1220a, 1475a, 1698a, 1703a, 1801a, 1827a, 2604a, 2734a, 2870a, 3392a, 4322a, 5942a; a-verse with single alliteration: 413a, 703a, 759a, 2382a, 4590a, 5366a, 5565a, 5776a, 5836a; b-verse: 410b, 519b, 630b, 876b, 978b, 983b, 1288b, 1367b, 1468b, 1619b, 1842b, 1851b, 3883b, 3918b, 4525b, 5194b, 5281b

3

The reorganisation of the first drop of type A1 in the Heliand as discussed above led in turn to the restructuring of type E. On this topic, see section 2.9 below.

MOSP-02.qxd

5/11/04

2:38 PM

62

Page 62

2. Metrical types and positions

The distribution of this configuration differs from any of the other type A1 variants treated above. Apparently most similar is the configuration PS#Px (Table 2.2), which is more or less evenly distributed in the a-verse and the b-verse. As regards the distribution of single and double alliteration in the a-verse, however, the two configurations are rather distinct: while the configuration PS#Px is strongly favourable to double alliteration, the configuration PS#px is only weakly so. As discussed at length in Suzuki (1996a: 81–9), the configuration PS#px is characterised as a marked variant of type A1 which arises through association of the second lift with a short stressed syllable (mi-, go-, and fru- in the above examples) as well as with the second mora of the immediately preceding (rather than the following as by resolution) syllable (-aft, -barn, and -uuerk in the above examples; for a formal representation, see (26a) below). Thus brought together, the constituents of the two separate syllables provide bimoraic structure necessary for realising the lift. Under normal circumstances, that is, unless resolution applies, the lift is mapped onto the long stressed syllable, which contains two moras (␮) under the syllable node (␴; see (26b) below). When resolution applies, the lift is associated with the two syllable nodes the first of which is monomoraic (see (26c) below). The exclusive presence of -S# in the immediately preceding position in the marked configuration at issue in the metre of Beowulf is accounted for by the fact that only -S# makes available a needed mora standing under a stressed syllable. It should be noted that -sx# cannot meet the condition, as the second mora belongs to an unstressed syllable; nor are -X# and -x# capable of providing a mora because they are unfooted and unstressed. (26) a.

␴

␴

g6

g

b.

L

c.

L

g

g6

␴

␴ ␴

g6

g h

mancraft mikil

␮␮

␮␮

g

g

gg

g g

␮

␮

␮␮ ␮

gg

g

man

mikil

36 L Key: ␴⫽syllable; ␮⫽mora; L⫽lift The configuration PS#px in the Heliand exhibits three outstanding features when compared with the analogous verse form in Beowulf. First, the frequency of the configuration is markedly lower in the Heliand than in Beowulf: the Heliand attests only slightly over the half of Beowulf (thirty-nine in the Heliand versus sixty-seven in Beowulf; only unambiguous examples counted).

MOSP-02.qxd

5/11/04

2:38 PM

Page 63

2.1. Type A1

63

Second, in contrast to Beowulf, the Heliand does not show any instances of related yet exceptional variants, PX#px (Beo 2457b ridend swefað) and Px#px (Beo 1828b hwilum dydon; Hofmann 1991: 70; Suzuki 1996a: 83). In other words, the conformity to the prototypical configuration PS#px is the strictest in the Heliand, disallowing any deviations from it. Third, while the sequence Psx#px is excluded as a variant of subtype A1s in Beowulf (Suzuki 1996a: 86–7) for the structural reason elucidated above, we seem to find a single such example in the Heliand, namely verse 1742a uuînberi uuesan. This verse might then be scanned as subtype A1s and interpreted accordingly as evidence for the poet’s license of using the disyllable -sx# in the second drop of subtype A1s in disregard of the traditional practice to the contrary, thereby calling into question the Heliand poet’s strictest conformity to the prototype of subtype A1s as suggested in the last paragraph. Closer consideration, however, will provide us with an alternative, more plausible scansion: the verse in question constitutes an instance of type E. As discussed in detail in section 2.1.7 above, the string -sx# has the closest parallel to -x# and -xx#, in distinction from -X# and -S#, as a realisation of the first drop of type A1, under which subtype A1s is also subsumed. In view of such a network of structural relations, then, it would seem least feasible to postulate an equivalence between the strings -sx# and -S# in the present context, precisely the assumption underlying the scansion of the sequence Psx#px as subtype A1s. On a more positive side, the proposed scansion of verse 1742a as type E is supported by the distribution pattern of the strings PS#, PX#, and Px# in relation to the two word forms, px and Px, as represented below (‘*’ denotes absence; see also sections 2.1.1, 2.1.2, and 2.1.5 above): (27) PS#px vs. PS#Px *PX#px vs. PX#Px *Psx#px vs. Psx#Px *Px#px vs. Px#Px The complete absence of Px#px would seem most compatible with the corresponding absence of Psx#px insofar as the latter configuration is characterised as a variant of subtype A1s. This interpretation is made further plausible by the lack of even PX#px, the string that would otherwise be expected to behave more like PS#px than would the string Psx#px. Taken altogether, then, the above distribution pattern may be adduced in support of the view that the Heliand metre follows the strictest rule for deriving subtype A1s, which in turn implies that the configuration Psx#px as exemplified by verse 1742a should be scanned as type E, rather than subtype A1s. The scansion of Psx#px as type E seems to be also borne out in traditional metrical practice. Beowulf provides an instance of a similar verse form, namely, Beo 1009b Healfdenes sunu, which is scanned as Pxx#px in strict metrical terms, rather than Psx#px as expected on morphological grounds (on the metrical value of complex proper names in Beowulf, see Suzuki 1996a: 73–4). This verse, too, should be identified as type E, rather than subtype A1s, primarily due to the identity of sunu as a short disyllable, the string most susceptible to resolution (Suzuki 1996a: 415 n. 79; for

MOSP-02.qxd

5/11/04

64

2:38 PM

Page 64

2. Metrical types and positions

related discussion, see also section 3.1.2 below). Thus, one of the strings closest to Psx#, namely Pxx#, fails to serve as a constituent of subtype A1s, and this restriction may therefore strengthen our scansion of the sequence Psx#px as type E. The diminished use of the configuration PS#px in the Heliand may be related to the loss of the minimally distinct configuration PS#Px as an autonomous type, treated in the following section. The reorganisation of the configuration PS#Px as a variant of type A1 and the consequent demise of its categorical distinguishability would then have led to a lesser motivation for composing the configuration PS#px (subtype A1s) as its structural opposite. The weakened structural basis of subtype A1s would in turn have removed the condition that was in effect in Beowulf for an extended use of it by analogy, that is by creating marginal variants PX#px and Px#px. There are three groups of verse that are not only scannable as subtype A1s but amenable to an alternative scansion, as follows: (i) x . . . PS#px; (ii) P#P#px; (iii) PxS#px. The first group (x . . . PS#px) may be scanned as subtype A1s with anacrusis and type B1 alike: (28) 195a suîðo godcund gumo 450a untthat that friðubarn godes 1011b Uuala that eu thes mag frâhmôd hugi Notice that the same phrase friðubarn godes is used in verses 450a (given in (28) above) and 2382a (subtype A1s, given earlier in (25)). As will be fully discussed in section 2.4.2.2 below, these apparently ambiguous verses should best be analysed as type B1. The second group (P#P#px) comprises nine verses, as listed below, which may be scanned as subtype A1s and also as subtype D2a, type D (/ / ⫻ ⫻; section 2.7). To decide on proper scansion, however, would require us to look into syntactic constituency, an issue that will be most appropriate to be treated in relation to alliterative pattern in section 3.2.2 below. At this point, we simply take for granted the result of syntax-based disambiguation as specified in the following list. (29) 2218a 2836a 4062a 2648b 2908b 3364b 3471b 5605b

4

5

dages lioht sehan (subtype A1s) tuê hund samad (subtype A1s) Krist godes sunu (subtype D2a; also 5584a)4 that godes barn sehan (subtype A1s, with anacrusis)5 Skrêd lioht dages (subtype D2a) hriop up thanen (subtype D2a) godes anst manot (subtype A1s) sehan lioht godes (subtype D2a)

One might propose another scansion for verses 4062a and 5584a, which contain a resolvable disyllabic word in medial position (godes): it might be scanned as type E by invoking suspension of resolution for the second position. Yet such a scansion, based on suspension of resolution for a primary-stressed syllable, would seem hardly feasible given the complete absence of the sequence P#px#P in contrast to the resolved configuration P#pxx#P (section 3.1.1.5). Alternatively but less plausibly, verse 2648b might be scanned as type B1. On this scansion, see section 2.4.2.2 below.

MOSP-02.qxd

5/11/04

2:38 PM

Page 65

2.1. Type A1

65

The third group (PxS#px) involves a seemingly idiosyncratic, single verse, with two occurrences: 3261a and 3772a mêðomhord manag.6 In addition to being scannable as subtype A1s with the second lift suspended from resolution and realised as a short stressed syllable (p), these verses might be analysed as an exceptional variant of type E, with the second lift resolved and the second drop standing more prominent than the first, in a way analogous to subtype D2b. Of the two competing interpretations, the scansion as a variant of type E has to be rejected as infeasible in metrical terms: there are no definite examples of PxS#P that may justify postulation of the configuration / ⫻ \ / as a metrically acceptable form (Sievers 1893: §114.7; cf. Hofmann 1991: 135; Russom 1998: 143); indeed, the absence of such a configuration seems to be well-motivated for structural reasons (cf. Suzuki 1996a: 129–30). On the other hand, scanning the verse at issue as subtype A1s as did Sievers (1893: §116.2) would not require any such additional apparatus otherwise unjustifiable. Given the independently motivated interpretive rule that is responsible for the composition of subtype A1s as discussed earlier in this section and more fully in connection with suspension of resolution in sections 3.1.1.1 and 3.1.1.6 below, the configuration PxS#px may yield to a proper scansion. The short stressed syllable (p) is allowed to fill the second lift on its own, because it is immediately preceded by a stressed syllable, which occupies the first drop in conjunction with the preceding unstressed syllable. Striking us as aberrant, however, would be the association of the first drop of type A1 with the syllable sequence -xS#, rather than -Sx#, which is found occasionally (section 2.1.2). By contrast, there are no instances of the configuration PxS#Px, as pointed out by Sievers (1893: §116.2) and Hofmann (1991: 136). In this light, the verse in question would appear nearly unmetrical, whether regarded as an original error or attributed to a corrupt transmission, and might accordingly invite emendation: one might propose as does Russom (1998: 143) mêðmhord manag (without the epenthetic vowel; a common example of subtype A1s) or manag mêðomhord (subtype D2b). These emendations, however, are not compelling. First, as noted by Hofmann (1991: 135–6), verse 1643b mêðomhordes, in which -odefinitely counts for scansion, casts doubt on the plausibility of reading mêðmhord manag. Second, there seem no independent, non-metrical reasons for transposing the word order mêðomhord manag recorded in the manuscripts: the verse is attested consistently without variation in two different places in two different manuscripts. We should therefore better adhere to the manuscript reading with Hofmann (1991: 135). On deeper consideration, the configuration PxS#px, deviant as it is from traditional practice, would be appreciated as fully conforming to the metrical system that the Heliand poet inherited and reorganised, in which the first drop of type A1 was given access to an enriched variety of language materials including the secondary-stressed syllable S, either independently or in combination with one or more unstressed syllables. There would accordingly have been no metre-internal reasons why the first drop of type A1 should not be associated with -xS#, apart from the total absence of such instances in the traditional stock of type A1 verses: the configurations PxS#px and 6

Hofmann (1991: 134) subsumes another verse under this minor group, namely verse 1061a aftar them fiuuartig dagun, on the assumption that fiuuartig constitutes a compound of the form PxS. This analysis, however, does not seem to be tenable on metrical grounds, as will be substantiated in section 2.2.1 below.

MOSP-02.qxd

5/11/04

66

2:38 PM

Page 66

2. Metrical types and positions

PxS#Px would thus have been deemed potentially metrical. Of these two potentially metrical forms, the non-existing configuration PxS#Px is characterised as more marked than the rare but attested configuration PxS#px, much as the configurations PS#Px and x . . . P#Px count as more marked than PS#px and x . . . P#px, respectively, precisely because these contexts normally induce suspension of resolution, and therefore they show greater preference for the association with the short stressed syllable as opposed to the long counterpart (section 3.1.1.6). Given such a markedness relation, it would be hardly surprising that the poet gave priority in his presumably experimental composition to the less marked and more acceptable configuration PxS#px at the expense of the more marked counterpart PxS#Px, rather than vice versa. The configuration PxS#px may thus be regarded as another instance of the Heliand poet’s metrical experiments, highly innovative and even revolutionary in appearance at times, but strictly conforming to the metrical system that he had at his disposal.

2.2. Type A2 (/ ⫻ / \) This section deals with the association of the second drop of type A with varying language materials and explores its consequences for metrical patterning. 2.2.1. The configurations Px . . . PS, Px . . . P#x, and Px . . . PX The second drop of type A may be realised by a lexical-stressed syllable, that is, an independent lexical word or the second element of a compound, as exemplified below:7 (30) 674a gold endi uuîhrôc 1557a gerno thurh godes thanc 4330a hungar hetigrim Table 2.38. Distribution of Px . . . PS and Px . . . P#P a-verse with double alliteration

a-verse with single alliteration

b-verse

76 (99%)

1 (1%)

0 (0%)

a-verse with double alliteration (the underlined verses are Px . . . P#P): 156a, 264a, 554a, 570a, 753a, 848a, 948a, 1024a, 1032a, 1345a, 1393a, 1508a, 1557a, 1638a, 1846a, 1865a, 1891a, 1916a, 1955a, 2005a, 2070a, 2173a, 2204a, 2233a, 2240a, 2268a, 2289a, 2343a, 2354a, 2360a, 2585a, 2655a, 2682a, 2702a, 2712a, 2881a, 2892a, 2931a, 2947a, 3082a, 3106a, 3303a, 3323a, 3450a, 3478a, 3734a, 3747a, 3842a, 3860a, 3897a, 4134a, 4275a, 4330a, 4343a, 4469a, 4492a, 4622a, 4658a, 4688a, 4692a, 4695a, 4731a, 4909a, 4954a, 5060a, 5116a, 5124a, 5137a, 5167a, 5209a, 5210a, 5252a, 5299a, 5321a, 5400a, 5970a; a-verse with single alliteration: 674a

7

Disregarded from our consideration in the following are the configurations Px . . . Psx and Px . . . P#px, which might be equated with the configuration Px . . . PS by invoking resolution on the final position (cf. Sievers 1893: §109). Such a scansion, however, has to be rejected, and these configurations should rather be identified as type D*, subtype D*2a, as will be discussed in detail in section 3.1.3 below.

MOSP-02.qxd

5/11/04

2:38 PM

Page 67

2.2. Type A2

67

The first drop involved in these verses may be realised by a wide variety of language material, as summarised below (cf. Kauffmann 1887: 298–9): (31) -x#: 948a, 1024a, 2173a, 2360a, 2585a, 3082a, 3860a, 3897a, 4330a, 4469a, 4658a, 5124a, 5321a, 5400a, 5970a -S#: 2343a, 3747a, 3842a, 4909a #x: 1032a, 2354a, 3450a -xx#: 1508a, 1955a, 2233a -xx#x: 2268a, 5299a -xx#: 2289a, 2881a -Xx#x: 2005a, 2712a, 5252a -x#x: 1557a, 1638a, 1865a, 1916a, 3106a, 4692a, 4731a -x#xx: 1393a, 4492a, 5060a, 5137a #xx: 156a, 570a, 674a, 753a, 848a, 1345a, 1891a, 2070a, 2204a, 2702a, 3323a, 3734a, 4134a, 4343a, 4688a, 4695a, 5167a, 5210a #xxx: 264a, 554a, 1846a, 2655a, 2947a, 4275a, 4954a -x#xxx: 2892a, 2931a, 3303a, 3478a -xx#xx: 5116a #xxxx: 2240a, 2682a, 4622a -x#xxxx: 5209a Thus, regardless of the quantity and quality of language material that occupies the first drop, the string Px . . . PS shows a strong preference for the a-verse with double alliteration. Of particular importance in the present context are the two variants -x# and -xx#, the strings that have been demonstrated to be minimally favourable to the a-verse with double alliteration when used as the first drop of the verse form / ⫻ / ⫻ (section 2.1.3). The consistency with which the verses containing such strings as their first drop constitute a-verses with double alliteration shows that the presence of a lexical stress on the second drop overrides the effects of the minimal first drop by fully enforcing its preference for the a-verse with double alliteration. The inherent effects of the string PS in bringing about a maximum incidence of the a-verse with double alliteration are also revealed when we focus on varying numbers of unstressed syllables that stand after the first lift by comparing the differing distribution patternings of P#x . . . PS on the one hand and P#x . . . Px on the other. It may be recalled that the string #x . . . exhibits an increasing preference for the a-verse with double alliteration in proportion to the number of syllables involved. Thus, other things being equal, we would expect that the two configurations prove to be comparable in showing graded preferences for the a-verse with double alliteration. As Table 2.39 makes evident, however, the two configurations are distinct in their distribution patterning: the one ending in PS is nearly exceptionless in realising the a-verse with double alliteration fully independent of the number of the preceding unstressed syllables; by contrast, the configuration P#x . . . Px is sensitive to the increasing number of syllables in the first drop. The contrastive patterns of distribution may thus justify us in assuming that P#x . . . PS should be distinguished from P#x . . . Px by virtue of the presence of lexical stress on verse-final position; and

MOSP-02.qxd

5/11/04

68

2:38 PM

Page 68

2. Metrical types and positions

Table 2.39. Distribution patterns of P#x . . . PS and P#x . . . Px

x⫽1 x⫽2 x⫽3 x⫽4 x⫽5

a-verse with double alliteration

a-verse with single alliteration

b-verse

P#x . . . PS P#x . . . Px

P#x . . . PS P#x . . . Px

P#x . . . PS P#x . . . Px

3 (100%) 141 (50%) 17 (95%) 143 (57%) 7 (100%) 128 (87%) 3 (100%) 73 (85%) 0 (0%) 20 (91%)

0 (0%) 1 (5%) 0 (0%) 0 (0%) 0 (0%)

0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%)

27 (10%) 28 (11%) 3 (2%) 1 (1%) 0 (0%)

115 (41%) 82 (32%) 16 (11%) 12 (14%) 2 (9%)

this may provide added support to the conclusion that the configuration Px . . . PS in general is to be metrically distinguished from Px . . . Px on the basis of the presence of a lexical-stressed syllable in verse-final position. Since the sequence Px . . . Px is a variant of type A1 (/ ⫻ / ⫻), the sharply differentiated configuration Px . . . PS has to be assigned a correspondingly distinct metrical value, namely, / ⫻ / \, whereby the second drop is characterised as a heavy drop. This metrical type will be referred to as type A2 in distinction from type A1 throughout the following discussion. Apparently related to the above configuration are a number of verses that contain a non-lexical (function) word in the place of a lexical one at the end of a verse, as exemplified below (Kauffmann 1887: 298): (32) 5104a gihid that he god sî 2109a uualdand frô mîn 4765b drohtin frô mîn Table 2.40. Distribution of Px . . . P#x (#x ⫽ non-lexical word) a-verse with double alliteration

a-verse with single alliteration

b-verse

4 (29%)

1 (7%)

9 (64%)

a-verse with double alliteration: 3832a, 4618a, 5104a, 5191a; a-verse with single alliteration: 2109a; b-verse: 46b, 490b, 971b, 1522b, 2836b, 3281b, 4765b, 4861b, 5017b

The above verses are varied as regards the ways the first drop is realised: (33) -x#: 490b, 971b, 4765b -X#: 2109a, 4861b, 5017b #xx: 4618a, 5104a, 5191a, 1522b, 3281b -x#xx: 3832a #xxx: 2836b #xxxx: 46b Remarkable about this configuration is the high proportion of the b-verse; by contrast, the a-verse occurs only just about once out of every three. Since the

MOSP-02.qxd

5/11/04

2:38 PM

Page 69

2.2. Type A2

69

distinguishing feature of the metrical form / ⫻ / \ is the nearly obligatory occurrence of the a-verse with double alliteration, the absence of this feature, and conversely the high frequency of the b-verse, demonstrate that the configuration under discussion is irreducible to the form / ⫻ / \ (type A2); rather, it constitutes a variant of type A1 (/ ⫻ / ⫻; cf. Heusler 1956: §261). The strong propensity of the verses ending in a function word to occur in the b-verse as we have seen above becomes more pronounced when compared with what would be characterised as the least prominent variety of type A2 verses, the group of verses in which the last two positions are occupied by a quasi compound – a complex word one of the two constituents of which is incapable of occurring on its own, namely either a prefix like un- or a heavy derivational suffix such as -dôm, -lîc (Suzuki 1996a: 68–73; see also sections 2.12.5 and 3.1.3 below). Under this group are subsumed the following thirteen verses: (34) 848a uuord endi uuîsdôm 1638a erlos an unreht 4492a talode im sô treulôs a-verse with double alliteration: 848a, 1024a, 1393a, 1638a, 1846a, 1891a, 2005a, 2892a, 3897a, 4343a, 4492a, 5209a, 5252a With regard to the composition of the first drop, these verses are distributed as follows: (35) -x#: 1024a, 3897a -Xx#: 2005a, 5252a -x#x#: 1638a -x#xx: 1393a, 4492a -x#xxx: 2892a -x#xxxx: 5209a #xx: 848a, 1891a, 4343a #xxx: 1846a As should be clear, the two groups of configurations Px . . . P#x and Px . . . PS (PS ⫽ quasi compound) are highly comparable both in terms of the number of examples attested in the corpus and in terms of the range of variability in the realisations of the first drop. Yet their distribution pattern proves to be divergent: the verse form Px . . . PS listed in (34) invariably appears in the a-verse with double alliteration, whereas the form Px . . . P#x given in (32) and Table 2.40 above occurs in the b-verse overwhelmingly. The contrast is most outstanding when we consider the realisations of the first drop that are commonly observed in the two groups and that are identified as most favourable to the a-verse with double alliteration and to the b-verse, respectively, namely, (-x)#xxx(x) on the one hand and -x# on the other. As it turns out, the verses of the group Px . . . P#x containing the first drop (-x)#xxx(x) (i.e., 46b, 2836b) appear in the b-verse despite the strong preference of such a long syllable sequence for the a-verse with double alliteration when used as the first drop of type A1. By contrast, the verses grouped under Px . . . PS with the first drop -x# (i.e., 1024a, 3897a) are all a-verses with double alliteration contrary

MOSP-02.qxd

5/11/04

70

2:38 PM

Page 70

2. Metrical types and positions

Table 2.41. Distribution of Px#PX a-verse with double alliteration

a-verse with single alliteration

b-verse

6 (18%)

0 (0%)

28 (82%)

a-verse with double alliteration: 271a, 409a, 1083a, 3458a, 4874a, 5075a; b-verse: 270b, 370b, 379b, 404b, 1138b, 1249b, 1334b, 1993b, 2032b, 2074b, 2089b, 2299b, 2577b, 2901b, 2924b, 3157b, 3219b, 3550b, 3892b, 4380b, 4600b, 4606b, 4657b, 4745b, 4909b, 5027b, 5630b, 5739b

to the close association of this string with the b-verse when used as the first drop of type A1. The divergence in metrical pattern observed between P#x and PS brings us then to consider another class of comparable verses, Px#PX, namely those verses ending in a non-compound with a long syllable in final position: e.g., 409a uueroldes uualdand. The distribution of this configuration is shown in Table 2.41 above. The configuration Px#PX is thus sharply distinguished from Px#PS by virtue of its strong propensity for the b-verse. Indeed, it seems hardly different from the configuration Px#Px in its overall distribution pattern, as shown in Table 2.23 above. The foregoing consideration shows that only the configuration Px . . . PS, in distinction from Px . . . P#x and Px . . . PX, is identified as constituting a separate metrical type due to its near obligatory occurrence in the a-verse with double alliteration. The composition of the sequence PS is effected by a compound and a quasi compound alike. This metrical structure might be labelled type A2b (/ ⫻ / \) following the term used for the metre of Beowulf (section 1.2; Suzuki 1996a: 13, 135). However, the Heliand lacks what would constitute its structural opposite, namely what might be called type A2a (/ \ / ⫻), as we have seen in section 2.1.1 above. In view of the lack of opposition, then, we adhere to labelling the configuration / ⫻ / \ simply as type A2. Aside from the terminology, we may naturally pose a fundamental question: what is the structural basis of this asymmetry in the Heliand? Why is the configuration / ⫻ / \ allowed, and the configuration / \ / ⫻ disallowed, as an independent metrical type, rather than vice versa? In order to address the issue raised immediately above, we need to review the relationship between type A1 with a relatively long first drop (Px . . . Px) and type A2a (PS#Px) in Beowulf. In it, the a-verse with double alliteration is strongly preferred when the first drop is realised by more than two syllables regardless of the exact language material used, while shorter syllable strings are varied in their verse distribution depending on their morpholexical status (Suzuki 1996a: 149–58). Thus, the accumulation of unstressed syllables may be regarded as largely equivalent to the presence of a lexical stress in its overall metrical effect: both contribute to a significant enhancement of the incidence of the a-verse with double alliteration. The predominant use in a-verses with double alliteration of the long sequences of unstressed syllables that are far removed from the prototype of the normal drop (i.e., the unstressed wordfinal syllable) is based on category extension by virtue of the relatively great prominence involved: those relatively long materials are treated thereby analogous to lexical-stressed syllables, which qualify for forming the heavy drop on their own (Suzuki 1996a: 168–9). We may accordingly characterise that the heavy drop is

MOSP-02.qxd

5/11/04

2:38 PM

Page 71

2.2. Type A2

71

determined disjunctively by two independent parameters, stress or the number of unstressed syllables. We should note that these two parameters normally do not come into play in conjunction: a sequence of a stressed syllable and unstressed syllables (in whatever order) is not allowed to realise a heavy drop. Hence in Beowulf verses of the form PS#x . . . Px are rarely attested (Suzuki 1996a: 74–5; see also section 2.1.2 above), and those of PxSx . . . Px totally unattested. In other words, the two parameters are mutually exclusive in scope: given language materials fall under the purview of one requirement in defiance of the other. While not wholly inconsistent and contradictory in itself, the coexistence of two mutually exclusive parameters may plausibly be regarded as less than fully integrated in structural terms, and there remains room for a greater degree of systematisation here. Faced with the above point of loose organisation open to potential disintegration, the Heliand poet would have attempted at a fuller integration by privileging one parameter at the sacrifice of the other: by systematisation the poet arrived at the generalisation that syllable quantity, that is, the number of syllables and syllable length, served as a prime parameter for determining the incidence of the a-verse with double alliteration for the first drop of type A1, as detailed in section 2.1.7.2 above. Accordingly, the presence of stress functioned only subordinate if any to the overall requirement of syllable quantity. Thus, a lexical-stressed syllable -S was treated essentially no different from -X, as was the string -sx from the string -xx. Since the number of syllables was of pervasive force, strings consisting of a stressed and unstressed syllables became qualified to realise the position in question. In addition to -S# and -sx# were now made available a whole array of strings, -Sx#, -S#x, -Sx#x, -Sx#xx, and the like. Furthermore, in line with the original generalisation that the word-initial syllable #x induced a higher incidence of the a-verse with double alliteration than the word-final counterpart -x# (see (23a) and (23b) above), the sequence -S#x was distinguished from -Sx# by its higher frequency of the a-verse with double alliteration. A further important consequence of the restructuring of the first drop of type A1 was its wholesale gradient manifestation: in sharp contrast to Beowulf, the presence of stress was no longer capable of delineating the heavy drop in categorical distinction from the normal counterpart by enforcing nearly obligatory realisation of the a-verse with double alliteration. What obtained in the Heliand in its place, however, was a gradient categorisation without clear boundaries which was predicated primarily on the number of unstressed syllables. Thus, the metre of the Heliand does not recognise the heavy drop as a distinct category for the first drop of type A; there is no question of postulating type A2a as a separate type, which constitutes a structural gap in the system. In marked contrast, the second drop of type A was free from the potentially selfsubverting, dual basis for the opposition between heavy and normal drops that we find characteristic of the first drop of type A in Beowulf. This verse-final position was limited to a single syllable for association by virtue of its minimality in metrical strength (section 1.2; Suzuki 1996a: 165–9). Thus, there was no possibility for the number of syllables to come to figure as an effective alternative parameter in determining the nature of the drop in verse-final position. The heavy drop was accordingly identified solely by the presence of stress. In the absence of potential threat of disintegration, the poet would not have recognised a need to interfere with

MOSP-02.qxd

5/11/04

72

2:38 PM

Page 72

2. Metrical types and positions

the traditional, binary distinction between heavy and normal drops that was grounded on the presence of stress. Accordingly, the sequence Px . . . PS was retained as a separate type (type A2) sharply distinguished from Px . . . Px (type A1) much in conformity with tradition. This asymmetry in metrical status between the configurations PS#Px (type A1) and Px#PS (type A2)8 may in turn be held responsible in no small measure for the much differentiated frequency in their occurrence in the Heliand, in sharp contrast with the largely even distribution in Beowulf. The Heliand offers twenty-eight examples of PS#Px (Table 2.1, section 2.1.1), and sixty-six examples of Px . . . PS (Table 2.38); by contrast, the corresponding figures for Beowulf amount to 167 (type A2a) and 132 (type A2b), respectively. To be sure, the use of poetic compounds is generally diminished in the Heliand, yet particularly significant seems the drastic decrease in their use for the first two positions of type A1: given that the corpus of the Heliand is about twice larger than that of Beowulf, the occurrence of compounds in the first two positions (PS#Px) in the Heliand decreased to less than 10 per cent of that in Beowulf, and that for the last two positions (Px#PS) about 25 per cent. Such a disproportionate manner of decrease brought about as a consequence a reversal of distribution in strong favour of the configuration Px . . . PS in the Heliand. Underlying such a radical change in distribution would have been the poet’s hesitation to use the identical prosodic form PS# with divergent metrical value for the two highly similar but categorically differentiated configurations, types A1 and A2: lift ⫹ normal drop for type A1 (PS#Px) on the one hand, and lift ⫹ heavy drop for type A2 (Px#PS) on the other. Confronted with the two conflicting scansions, then, the poet would have worked out a compromise by suppressing use of this ambivalent form PS# for the first two positions of type A1 (/ ⫻) and reserved it largely for the last two positions of type A2 (/ \). His decision would seem well-reasoned: first, the identity of type A2 rested nearly exclusively on the compound form PS# at issue, while the same form constituted only a variant among the wide spectrum of language materials available for the first two positions of type A1; and second, the value lift ⫹ heavy drop was an inherited one with a wealth of traditional verses in support, whereas the alternative value lift ⫹ normal drop was an innovation that lacked such a stock of expressions and had accordingly a potential of contradicting traditional practice. As a further testimony of the poet’s reluctance to use the form PS as the first constituent of type A1, we may draw attention to the extreme rarity of the configuration PS#PS in the Heliand (cf. Russom 1998: 167), which should be a variant of type A2 as well. There are only four such examples in the corpus, as given below. As expected, these verses are all a-verses with double alliteration due to the presence of lexical stress (S) in verse-final position. (36) 2343a lêðlîc lôngeld 3747a unreht ênfald (also 3842a) 4909a diurlic dages lioht

8

The quantity and quality of the second drop for PS#Px, and of the first drop for Px#PS, is immaterial for the following discussion.

MOSP-02.qxd

5/11/04

2:38 PM

Page 73

2.2. Type A2

73

No less significant than the rarity of the configuration PS#PS in itself is that the above examples all have a quasi compound for the first two positions. The Heliand accordingly contains no instance of the configuration PS#PS in which the first S is occupied by the second element of a true compound. In Beowulf, by contrast, there are twenty examples of PS#PS, of which thirteen contain a true compound as their first constituent.9 The predominance of true compounds in the corresponding position in Beowulf may thus prompt us to assume that the distribution pattern in the Heliand would have been far from accidental and accordingly to attribute the total absence of true compounds in the Heliand to the poet’s deliberate avoidance of using the form PS as the first constituent of type A1. Before we move on, we may pause here for a moment to consider briefly how the well-established status of type A2 bears on particular scansions. At stake is the following group of verses that are regarded by Hofmann (1991, II: 36–54) as containing a compound (22 according to his formula): (37) 5149a thrîtig scatto 144b efno tuêntig a-verse with single alliteration: 5149a; b-verse: 144b, 450b, 843b, 963b, 1053b Of particular interest about the above group are the following three points. First, the verses in question all contain numerals ending in -tig. Second, with the exception of verse 144b, the numerals occupy the first two positions. Third, none of the examples manifest double alliteration, and all but one are b-verses. Bringing these features together, we may arrive at a scansion different from Hofmann’s. Although the second element -tig may be identified on morphological grounds as an instance of heavy derivational suffixes such as -lîk, the words involved should better be analysed in metrical terms as indistinguishable from simplex words, rather than as quasi compounds. Accordingly, the verses concerned may be treated as equivalent to Px#Px, in distinction from Px#PS and PS#Px. Furthermore, in keeping with the demonstrably strong preference of Px#Px for the b-verse (section 2.1.5), the proposed characterisation is compatible with the total absence of comparable verses in the a-verse with double alliteration. 2.2.2. The configuration Px . . . Pxx This configuration is distinguished by the presence of a disyllabic ending in the versefinal drop. It is differentiated from the configuration Px . . . Pxx (subtype D*1), in which the last two syllables constitute separate morphological entities. As discussed extensively in section 2.8 below, the configuration Px . . . Pxx occurs in the a-verse with double alliteration predominantly (about eight times out of every ten), and the a-verse with single alliteration is extremely rare, a distribution pattern diametrically opposed to the configuration Px . . . Pxx, as substantiated in Table 2.42 below. Along with the configuration Px . . . Pxx, Kauffmann (1887: 286–7, 291, 293, 294, 295, 303, 306), Sievers (1893: §116.1), and Martin (1907: 58) scan as type A1

9

Verses containing a true compound in the first two positions of type A2b: Beo 193a, 215a, 232a, 330a, 485a, 1246a, 1267a, 1650a, 1719a, 1881a, 2154a, 2339a, 2618a; verses containing a quasi compound in the same positions: Beo 251a, 287a, 641a, 690a, 1424a, 1489a, 3041a.

MOSP-02.qxd

5/11/04

74

2:38 PM

Page 74

2. Metrical types and positions

Table 2.42. Distribution of Px . . . Pxx a-verse with double alliteration

a-verse with single alliteration

b-verse

8 (22%)

15 (42%)

13 (36%)

a-verse with double alliteration: 319a, 1568a, 1617a, 2223a, 3191a, 3880a, 4775a, 5037a; a-verse with single alliteration: 917a, 936a, 1438a, 1459a, 1699a, 1963a, 2490a, 3197a, 3284a, 3701a, 4377a, 4842a, 4984a, 5014a, 5035a; b-verse: 223b, 1712b, 1748b, 1964b, 2433b, 2644b, 2899b, 3803b, 4153b, 4281b, 4689b, 4773b, 5198b

(rather than subtype D*1) the configuration Px . . . Pxx as well in which the penultimate unstressed syllable arose secondarily through restoration of syncopated vowels (see section 1.3.1 above). At issue are verses such as 4371a drohtines engilos and 5905a rîkies drohtines. These scansions have to be rejected, however, as Hofmann (1991: 108–9) argues convincingly (see also Russom 1998: 168): the two unstressed syllables constitute separate positions elsewhere, as in 2084a drôm drohtines and 3350b Godes engilos, both instantiating subtype D1; otherwise we would have to postulate verses with only three metrical positions. Associating the disyllable in question with a single position on the one hand, and with two positions on the other, would amount to arbitrary practice in the extreme. The first drop of the configuration Px . . . Pxx is realised by a variety of language materials as shown below: (38) -x#: 319a, 917a, 3197a, 3284a, 4842a, 4984a, 5014a, 1964b, 2433b, 2899b -xx#: 936a -xx#: 4689b -x#x#: 1459a, 223b, 2644b -x#x-: 1699a, 1963a, 2490a, 3191a, 4377a, 5035a, 1712b, 1748b, 3803b, 4153b, 4773b, 5198b #xx: 1438a, 3701a, 3880a, 4281b -x#xx: 5037a #xxx: 4775a -x#xxxx: 1568a, 1617a -xx#xx: 2223a As it turns out, the disyllabic ending does not favour the a-verse with double alliteration as would be expected from its behaviour as a realisation of the first drop in contrast to the monosyllabic drop. Instead, this configuration seems to show a certain preference for the a-verse with single alliteration: when the first drop is occupied by a word-final syllable (Px#Pxx), the a-verse with single alliteration predominates over the b-verse, which would otherwise be prevalent (see Table 2.21 above, section 2.1.5): one example for the a-verse with double alliteration (10 per cent); six examples for the a-verse with single alliteration (60 per cent); three examples for the b-verse (30 per cent). Further, when the first drop is realised by a disyllable with the second syllable serving as a prefix (Px#xPxx), we find a relatively high frequency of the a-verse with single alliteration, this time at the expense of the a-verse with double alliteration (compare Table 2.13 above, section 2.1.3): one example for the a-verse with double alliteration (8 per cent); five examples for the a-verse with single

MOSP-02.qxd

5/11/04

2:38 PM

Page 75

2.2. Type A2

75

alliteration (42 per cent); six examples for the b-verse (50 per cent). Thus, the configuration Px . . . Pxx cannot entirely be reduced to Px . . . Px, although the two appear more similar to each other than to the configuration Px . . . PS by virtue of their general avoidance of the a-verse with double alliteration. The second drop of type A1 that is realised by a disyllabic ending is accordingly distinguished from the monosyllabic counterpart. The lack of demonstrable preference for the a-verse with double alliteration on the part of the disyllabic ending would appear puzzling in view of the fact that the same ending induces a higher incidence of the a-verse with double alliteration when used as the first drop of type A1 in contrast to the word-final monosyllable (-x#) and disyllable (-xx#; section 2.1.7.1). As it seems, then, the behaviour of the disyllabic ending in relation to the frequency of the a-verse with double alliteration is context-sensitive: verse internally, it correlates to increased use of the a-verse with double alliteration in comparison with the word-final monosyllable and disyllable; verse-finally, by contrast, it fails to show a comparable correlation. Yet the existing context-dependent pattern would make better sense than the logically possible reverse pattern whereby the disyllabic ending favours the a-verse with double alliteration only when used in verse-final position. As demonstrated above, the first drop is subject to a far greater range of variation as regards its realisation than the second counterpart, which is limited to monosyllables except for disyllabic endings. Accordingly, the first drop has a potential for being involved in a finer gradation pattern of preference for the a-verse with double alliteration. By contrast, the second drop shows what may be characterised as a binary opposition in terms of preference: presence versus absence of preference for use of the a-verse with double alliteration. Such a binarity in choice is ascribed to the binary opposition that is in effect for the second drop of type A: the normal drop (A1) versus the heavy drop (A2). Given such a minimal range of variability, it would appear hardly surprising that the disyllabic ending, a third and relatively minor group of material available for the position, fails to be encoded with a distinct value of preference for the a-verse with double alliteration. Since the heavy drop prototypically requires a lexical-stressed syllable for realisation, the disyllabic ending cannot be equated to such a secondary-stressed syllable as regards qualification for association with this position: as a grammatical ending, it does not bear lexical stress. Accordingly, determined as far from equivalent to the sequence Px#PS, the whole configuration Px#Pxx is incapable of inducing a maximal incidence of the a-verse with double alliteration. While sharply differentiated from Px#PS, the configuration Px#Pxx is not fully reducible to the sequence Px#Px. Yet the observed difference in distribution pattern between the configurations Px#Pxx and Px#Px would seem far from arbitrary in view of the differential treatment of the two in other contexts. Although not specifically conducive to use of the a-verse with double alliteration, the configuration Px#Pxx still shows a weaker preference for the b-verse than the sequence Px#Px, a property which makes sense given the relatively greater prominence of -xx# than -x#. Totally unintelligible in structural terms, on the other hand, would be the contrary pattern whereby the sequence Px#Pxx would exhibit a greater propensity for the b-verse than Px#Px.

MOSP-02.qxd

5/11/04

76

2:38 PM

Page 76

2. Metrical types and positions

2.3. Type A3 (⫻ ⫻ / ⫻) This type is characterised by replacement of a lift by a drop in initial position; the verse accordingly consists of one lift and three drops. A whole list of examples is provided below (cf. Kauffmann 1887: 308–12; Hofmann 1991, II: 13–20):

(39) 24a that sea scoldin ahebbean 3780a Sô gifragn ik that thar an themu uuîhe 4074a hêt ina thô lêdien a-verse with single alliteration: 10a, 24a, 34a, 48a, 74a, 77a, 90a, 116a, 123a, 129a, 146a, 148a, 161a, 164a, 176a, 179a, 207a, 212a, 226a, 248a, 303a, 308a, 369a, 396a, 405a, 469a, 498a, 562a, 574a, 585a, 614a, 642a, 660a, 726a, 738a, 812a, 821a, 826a, 834a, 858a, 872a, 873a, 877a, 878a, 880a, 884a, 888a, 894a, 939a, 963a, 971a, 975a, 1030a, 1065a, 1071a, 1075a, 1090a, 1101a, 1139a, 1162a, 1194a, 1221a, 1224a, 1288a, 1298a, 1367a, 1370a, 1389a, 1414a, 1419a, 1426a, 1434a, 1443a, 1461a, 1474a, 1476a, 1490a, 1492a, 1499a, 1507a, 1517a, 1518a, 1528a, 1552a, 1565a, 1566a, 1567a, 1573a, 1579a, 1584a, 1601a, 1605a, 1606a, 1611a, 1612a, 1613a, 1616a, 1621a, 1640a, 1647a, 1649a, 1657a, 1666a, 1678a, 1700a, 1708a, 1739a, 1771a, 1781a, 1786a, 1788a, 1794a, 1831a, 1841a, 1844a, 1867a, 1884a, 1887a, 1894a, 1915a, 1920a, 1922a, 1927a, 1950a, 1956a, 1957a, 1968a, 1971a, 1984a, 2003a, 2006a, 2017a, 2030a, 2045a, 2058a, 2142a, 2148a, 2153a, 2187a, 2194a, 2222a, 2248a, 2253a, 2281a, 2283a, 2293a, 2299a, 2344a, 2364a, 2379a, 2392a, 2401a, 2406a, 2408a, 2415a, 2417a, 2431a, 2438a, 2448a, 2450a, 2471a, 2483a, 2486a, 2496a, 2497a, 2503a, 2506a, 2526a, 2573a, 2578a, 2621a, 2623a, 2627a, 2637a, 2638a, 2649a, 2678a, 2690a, 2691a, 2697a, 2720a, 2726a, 2743a, 2749a, 2763a, 2778a, 2782a, 2835a, 2866a, 2883a, 2898a, 2949a, 2952a, 2969a, 2980a, 2992a, 2997a, 3015a, 3028a, 3057a, 3076a, 3123a, 3157a, 3181a, 3188a, 3193a, 3194a, 3228a, 3243a, 3263a, 3267a, 3268a, 3278a, 3320a, 3341a, 3368a, 3371a, 3394a, 3400a, 3402a, 3420a, 3436a, 3461a, 3465a, 3471a, 3484a, 3516a, 3525a, 3550a, 3551a, 3575a, 3607a, 3613a, 3619a, 3623a, 3652a, 3661a, 3663a, 3670a, 3694a, 3703a, 3706a, 3714a, 3724a, 3728a, 3732a, 3743a, 3758a, 3766a, 3780a, 3788a, 3796a, 3805a, 3815a, 3819a, 3823a, 3827a, 3836a, 3839a, 3840a, 3846a, 3848a, 3849a, 3850a, 3853a, 3857a, 3863a, 3870a, 3876a, 3889a, 3912a, 3921a, 3926a, 3942a, 3950a, 3951a, 3964a, 3974a, 3980a, 3989a, 4008a, 4018a, 4074a, 4094a, 4112a, 4139a, 4155a, 4172a, 4173a, 4183a, 4202a, 4226a, 4280a, 4286a, 4296a, 4308a, 4339a, 4344a, 4346a, 4378a, 4407a, 4409a, 4423a, 4430a, 4444a, 4468a, 4471a, 4486a, 4493a, 4513a, 4520a, 4526a, 4531a, 4538a, 4539a, 4577a, 4596a, 4605a, 4636a, 4644a, 4652a, 4657a, 4667a, 4677a, 4689a, 4696a, 4704a, 4709a, 4723a, 4754a, 4764a, 4769a, 4777a, 4817a, 4825a, 4840a, 4850a, 4851a, 4906a, 4925a, 4933a, 4961a, 5012a, 5017a, 5031a, 5033a, 5036a, 5073a, 5074a, 5085a, 5123a, 5141a, 5154a, 5155a, 5157a, 5194a, 5196a, 5197a, 5198a, 5223a, 5269a, 5295a, 5335a, 5337a, 5342a, 5348a, 5353a, 5377a, 5397a, 5401a, 5411a, 5467a, 5493a, 5495a, 5504a, 5506a, 5520a, 5532a, 5546a, 5547a, 5560a, 5600a, 5602a, 5636a, 5696a, 5724a, 5729a, 5746a, 5761a, 5818a, 5820a, 5835a, 5854a, 5859a, 5865a, 5879a, 5972a

MOSP-02.qxd

5/11/04

2:38 PM

Page 77

2.3. Type A3

77

Table 2.43. Distribution pattern of types A3, B1, and C according to the number of unstressed syllables before the first lift number of syllables

type A3

1 2 3 4 5 6 7 8 9 10 11

0 (0%) 7 (2%) 25 (6%) 134 (34%) 108 (27%) 75 (19%) 32 (8%) 11 (3%) 5 (1%) 0 (0%) 0 (0%)

total

397 (100%)

type B1 139 (4%) 609 (19%) 769 (24%) 661 (21%) 485 (15%) 278 (9%) 111 (4%) 66 (2%) 19 (1%) 4 (less than 1%) 1 (less than 1%) 3142 (100%)

type C 187 (10%) 450 (25%) 415 (23%) 303 (17%) 200 (11%) 115 (6%) 67 (4%) 30 (2%) 11 (1%) 10 (1%) 2 (less than 1%) 1790 (100%)

The number of unstressed syllables appearing before the lift varies between two and nine, with four the commonest followed by five and six, as indicated in the second column of Table 2.43. As shown in Table 2.43, the distribution pattern of type A3 differs significantly from that of types B1 and C, which begin with a drop as does type A3. First, while types B1 and C show the three commonest numbers of syllables to be from two to four, type A3 shifts the central area of distribution higher up to from four to six. Second, type A3 is distinguished by a more compact distribution than types B1 and C. More specifically, in type A3 about 80 per cent of all examples have one of the commonest numbers of syllables; by contrast, in types B1 and C the corresponding figures fall short of two-thirds. Since type A3 is limited to the a-verse, however, we may suspect that the second feature noted above is not genuine. In particular, when we recall that the b-verse exhibits preference for a larger number of syllables in verse-initial position than the a-verse (sections 2.4.1, 2.6.1, and 2.12.3), the diffuse distribution that would appear characteristic of types B1 and C may be due to their occurrence in the b-verse rather than the inherent nature of these two metrical types. Moreover, both types B1 and C occur in the b-verse more frequently. It would accordingly be more instructive to exclude this possible distorting factor and concentrate on the a-verse of types B1 and C for purposes of comparison. The distribution pattern is summarised in Table 2.44. As suspected earlier, we obtain analogously compact distributions for the three types: the three most commonly occurring numbers of syllables account for about 80 per cent of all instances for each type. Of paramount significance still is the validity of the first point we have made above: while the three commonest numbers of syllables range from two to four for types B1 and C with two the commonest, they are from four to six for type A3 with four the commonest. Of further interest, while the focal

MOSP-02.qxd

5/11/04

2:38 PM

Page 78

78

2. Metrical types and positions

Table 2.44. Distribution pattern of types A3, B1, and C in the a-verse according to the number of unstressed syllables before the first lift number of syllables

type A3

type B1

type C

1 2 3 4 5 6 7 8 9

0 (0%) 7 (2%) 25 (6%) 134 (34%) 108 (27%) 75 (19%) 32 (8%) 11 (3%) 5 (1%)

87 (12%) 219 (29%) 209 (28%) 139 (18%) 56 (7%) 19 (3%) 17 (2%) 5 (1%) 1 (less than 1%)

63 (11%) 192 (34%) 167 (29%) 84 (15%) 31 (5%) 18 (3%) 7 (1%) 6 (1%) 0 (0%)

total

397 (100%)

752 (100%)

568 (100%)

area of distribution is shifted upwards along the numerical scale by two, the minimal number of syllables for type A3 is two, rather than one as for the other types, and its occurrence is highly exceptional. The question arising then is: what is the structural basis for the shifted distribution pattern of type A3 in distinction from types B1 and C? Of particular concern is the total absence of a single syllable in type A3. We may account for this property of type A3 by referring to its unique composition: ⫻ ⫻ / ⫻. That is, it begins with a sequence of two drops, rather than one as in types B1 and C. Since each position has to be filled by language material for realisation, the minimal number of unstressed syllables must be two, which are then associated with the two positions involved on a one-to-one basis; otherwise, the resultant verse would contain only three positions, thereby violating the fundamental principle of verse composition, which requires that every verse should have at least four positions. By contrast, because types B1 and C begin with a single drop, the minimal number of syllables may be one. Corresponding to the duality of drops involved in type A3, the average number of syllables appearing before the lift is larger than that for types B1 and C. Thus, the structural form underlying type A3 has to be identified as ⫻ ⫻ / ⫻, rather than ⫻ / ⫻, which would be at odds with the distinct ways the verse-initial part is realised in terms of the number of syllables. A similar distribution pattern is observed in Beowulf (Table 2.45). The minimal number of syllables for the first drop of type A3 is two, which is in itself extremely exceptional. The focal area of distribution ranges from three to five, therewith accounting for more than 90 per cent of all examples. By comparison, types B1 and C are virtually limited to the first three numbers for the realisation of the first drop, again in parallel to their composition in the Heliand. One notable difference between the two works, however, lies in the more diffuse distribution in favour of larger numbers of syllables in the Heliand, a distinctive characteristic of the Old Saxon poem which will be treated in depth in subsequent sections. Having determined the nature of the first two drops, we now turn to consider the final drop as regards its linguistic realisation. In the Heliand there are three examples

MOSP-02.qxd

5/11/04

2:38 PM

Page 79

2.3. Type A3

79

Table 2.45. Distribution pattern of types A3, B1, and C in the a-verse in Beowulf according to the number of unstressed syllables before the first lift number of syllables

type A3

type B1

type C

1 2 3 4 5 6 7

0 (0%) 8 (2.5%) 110 (36%) 131 (42%) 49 (16%) 11 (3.5%) 0 (0%)

55 (18%) 196 (64%) 48 (15.5%) 8 (2.5%) 1 (nearly 0%) 0 (0%) 0 (0%)

89 (17%) 332 (64%) 79 (15%) 21 (4%) 1 (nearly 0%) 0 (0%) 0 (0%)

total

309 (100%)

308 (100%)

522 (100%)

of type A3 verse in which the final weak position is occupied by an independent word, a variant which is totally absent in Beowulf: (40) 3805a endi an thene godes uueg 4226a habde ine thiu smale thiod 5879a al sô it giduan uuas Furthermore, there is a single instance of type A3 in which the corresponding position is filled by the second constituent of a compound word, as follows: (41) 3766a endi siu an that tresurhûs The extreme rarity of the examples makes a marked contrast with Beowulf, which offers as many as sixteen comparable instances: Beo 168a, 219a, 484a, 1496a, 1504a, 1599a, 1652a, 1877a, 1995a, 2101a, 2258a, 2389a, 2466a, 2699a, 2770a, 3147a (Suzuki 1996a: 127). The final drop may be realised by the second element of a quasi compound, as is in Beowulf. Five examples are found in the Heliand as given below, while there are two such examples in Beowulf (Beo 1059a, 2376a). (42) 207a that it elcor sô uuânlîc 207a, 308a, 1101a, 1492a, 3652a The diversified ways in the Heliand of realising the final drop in type A3 involving independent words may be related to the general obscuring of sharper distinctions which are in force in Beowulf. Of particular relevance here would be the fact that independent lexical words are allowed greater freedom to occur in the normal drop in the Heliand than in Beowulf, a significant point that will be noted recurrently below. Yet the weakening of distinctions does not mean loss of distinctions. The occurrence of lexical-stressed syllables in the final drop of type A3 is still kept highly exceptional; what is involved here is rather an introduction of gradient-based

MOSP-02.qxd

5/11/04

80

2:38 PM

Page 80

2. Metrical types and positions

distinction in the Heliand, whereby deviant materials may be used but only occasionally. Finally, the last drop of type A3 may be occupied by a disyllabic ending (-xx#) as exemplified below: (43) 877a 880a 884a 1139a 1221a 1298a 1434a 2471a 5197a

iro selboro euuar selboro euuar selbaro huuô sie scoldin iro selboro thoh sie thar alle be gelîcumu huuilike uuârin allaro that he âdrana talod imu, huô he ôðrana sô it an iuuuaro aldrono

The above configuration xx . . . Pxx is distinguished from the sequence xx . . . Pxx, which constitutes a variant of type C, verse 1229a that sie ûses drohtines, for example. The justification for the distinction, however, is slight when considered in isolation. Since both configurations occur in the a-verse with overwhelming frequency (type A3 without exception, the variant of type C at issue here with rare exceptions; see section 2.6.2 below), the distribution pattern is far from compelling as a justification. The complete absence of the configuration xx . . . Pxx in the b-verse accordingly does not seem to carry much weight. Yet when taking into account the distinguishability of the disyllabic ending -xx# from the disyllabic sequence -xx# in other metrical contexts (sections 2.1.4, 2.2.2, and 2.8), it would appear reasonable to maintain the distinction here as well.

2.4. Type B1 (⫻ / ⫻ /) 2.4.1. The first drop of type B1 The first drop of type B1 is realised by a varying number of unstressed syllables ranging between one and eleven, as shown in Tables 2.46 and 2.47 (cf. Kauffmann 1887: 312–24; Hofmann 1991, II: 94–142, 191–2). Overall, the commonest number of syllables used is three (Tables 2.46 and 2.47). Yet the length of the initial position bears significantly on verse distribution pattern. Specifically, the number of syllables associated with the first drop is related to the distribution pattern of the a-verse versus the b-verse in the ways we may generalise as follows: the larger the number of syllables involved, the higher the incidence of the b-verse; and the a-verse with double alliteration is less favourable to increasing numbers of syllables than the a-verse with single alliteration (Table 2.47). Of further interest is the predominance of the a-verse with double alliteration when the first drop is occupied by a prefix alone: out of ten relevant examples (752a unsculdige scole, 1602a, 1890a, 3097a, 3351a, 3720a, 4082a, 4593a, 5661a, 206b), seven (70 per cent) constitute a-verses with double alliteration (752a, 1890a, 3097a, 3351a, 3720a, 4082a, 4593a). This figure would seem out of proportion in the

MOSP-02.qxd

5/11/04

2:38 PM

Page 81

2.4. Type B1

81

Table 2.46. Distribution of the varying number of syllables in the first drop of type B1 for the a-verse with double alliteration, the a-verse with single alliteration, and the b-verse number of x 1 2 3 4 5 6 7 8 9 10 11 total

a-verse with double alliteration

a-verse with single alliteration

49 20% 83 34% 60 24% 31 13% 14 6% 7 3% 2 1% 1 less than 1% 0 0% 0 0% 0 0%

38 8% 136 27% 149 30% 108 21% 42 8% 12 2% 15 3% 4 1% 1 less than 1% 0 0% 0 0%

247 100%

505 100%

b-verse 52 2% 390 16% 560 23% 522 22% 429 18% 259 11% 94 4% 61 3% 18 1% 4 less than 1% 1 less than 1% 2390 100%

Table 2.47. Verse distribution according to the number of syllables in the first drop of type B1 number of x

a-verse with double alliteration

a-verse with single alliteration

b-verse

total

1 2 3 4 5 6 7 8 9 10 11

49 35% 83 14% 60 8% 31 5% 14 3% 7 3% 2 2% 1 2% 0 0% 0 0% 0 0%

38 136 149 108 42 12 15 4 1 0 0

52 37% 390 64% 560 73% 522 79% 429 88% 259 94% 94 85% 61 92% 18 95% 4 100% 1 100%

139 609 769 661 485 277 111 66 19 4 1

27% 22% 19% 16% 9% 4% 14% 6% 5% 0% 0%

100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%

context of the overall distribution pattern for the realisation of the first drop as a monosyllable, as indicated in Table 2.47: only about 35 per cent of all occurrences with the monosyllabic first drop constitute a-verses with double alliteration. On closer consideration, however, we may provide a reasonable account of the apparently anomalous preponderance of the a-verse with double alliteration here. The proportion of the a-verse with double alliteration to the other two verses (i.e., the a-verse with single alliteration and the b-verse) becomes maximal when the first drop is associated with a single syllable, as pointed out above. In other words,

MOSP-02.qxd

5/11/04

2:38 PM

82

Page 82

2. Metrical types and positions

the a-verse with double alliteration shows the strongest preference for minimal prominence falling on its first drop. Since (monosyllabic) prefixes are characterised as less prominent than any other instances of monosyllables, namely, independent words (see section 2.1.7 above; see also Hofmann 1991: 139), it seems natural that the a-verse with double alliteration occurs with by far the greatest relative frequency when the first drop is realised by a monosyllabic prefix, although use of prefixes as a sole occupant of the first drop is marginal for type B1 as a whole. Hofmann (1991: 146) divides type B1 into four subtypes according to the language material that realises the two verse-internal positions / ⫻. In view of the wide variety of type A1 (section 2.1), it would be of interest to see whether Hofmann’s subtypes and their further variants are treated differently with regard to the realisation of their first drop. To be considered are the following eight configurations, all subsumable under type B1: (i) x . . . P#xP (Hofmann’s Bx1.x1); e.g., 2282b endi im forð giuuêt; (ii) x . . . Px#P (Hofmann’s Bx2.1); e.g., 76b That uuas sô sâlig man; (iii) x . . . PX#P (Hofmann’s Bx2.1); e.g., 20b Habda im uualdand god; (iv) x . . . Pxx#P (Hofmann’s B5x3a.1); e.g., 552b endi he frâgoda sân; (v) x . . . PXx#P (Hofmann’s B5x3a.1); e.g., 467b habda im hêlagna gêst; (vi) x . . . Pxx#P (Hofmann’s Bx2⫹.1); e.g., 359b thar iro beiðero uuas; (vii) x . . . PxS (Hofmann’s dbx3b); e.g., 51a an thesan middilgard; (viii) x . . . Px#xP (Hofmann’s x2.x1); e.g., 484b nu mi the uuilleo gistôd. The distribution patterns of these various configurations are given in Tables 2.48 through 2.55. As it turns out, the overall pattern is no different from subtype to subtype, with the exception of the configuration x . . . PxS (Table 2.54), which reverses the general pattern by favouring the a-verse with single alliteration at the expense of the b-verse. No less important, the accumulation of syllables in the initial drop does not seem to be accompanied by a markedly increasing incidence of the b-verse; the a-verse with single alliteration and the b-verse are hardly distinguishable with respect to the range of the unstressed syllables associated with the first drop. Table 2.48. Distribution of x . . . P#xP according to the number of syllables in the first drop number of x 1 2 3 4 5 6 7 8 9 total

a-verse with double alliteration 2 33% 2 10% 4 12% 2 4% 0 0% 0 0% 0 0% 0 0% 0 0% 10

7%

a-verse with single alliteration 1 17% 1 5% 3 9% 5 11% 2 9% 0 0% 1 14% 0 0% 0 0% 13

9%

b-verse

total

3 50% 17 85% 27 79% 39 85% 21 91% 8 100% 6 86% 0 0% 1 100% 122

84%

6 20 34 46 23 8 7 0 1

100% 100% 100% 100% 100% 100% 100% 0% 100%

145 100%

MOSP-02.qxd

5/11/04

2:38 PM

Page 83

2.4. Type B1

83

Table 2.49. Distribution of x . . . Px#P according to the number of syllables in the first drop number of x 1 2 3 4 5 6 7 8 9 10 total

a-verse with double alliteration 31 46% 43 13% 37 9% 15 4% 7 3% 3 2% 2 3% 0 0% 0 0% 0 0% 138

8%

a-verse with single alliteration 18 81 80 57 21 8 8 2 0 0

26% 24% 19% 15% 8% 5% 13% 5% 0% 0%

275 16%

b-verse

total

19 28% 213 63% 294 72% 307 81% 245 90% 145 93% 54 84% 37 95% 10 100% 3 100% 1327

76%

68 337 411 379 273 156 64 39 10 3

100% 100% 100% 100% 100% 100% 100% 100% 100% 100%

1740 100%

Table 2.50. Distribution of x . . . PX#P according to the number of syllables in the first drop number of x

a-verse with double alliteration

a-verse with single alliteration

b-verse

1 2 3 4 5 6 7 8 9 10

0 0% 0 0% 0 0% 1 7% 0 0% 1 17% 0 0% 1 100% 0 0% 0 0%

0 0% 1 5% 2 6% 2 14% 1 7% 0 0% 0 14% 0 0% 0 0% 0 0%

6 100% 19 95% 34 94% 11 79% 13 93% 5 83% 3 100% 0 0% 1 100% 0 0%

total

3

6

92

3%

6%

91%

total 6 20 36 14 14 6 3 1 1 0

100% 100% 100% 100% 100% 100% 100% 100% 100% 0%

101 100%

On the strength of the unique properties of the configuration x . . . PxS mentioned in the preceding paragraph, one might claim, as does Hofmann (1991), that this configuration should be distinguished from type B1 as a separate type, specifically type dbx3b according to Hofmann’s system of scansion. That is, the configuration in question, one might argue, should be characterised as a variant of type D (subtype D2b; see section 2.7 below), in which the first lift is replaced by a drop, as with types A3 and B3. Accordingly, the following derivation might be postulated for this hypothetical metrical type: / / ⫻ \ —⬎ ⫻ / ⫻ \; compare the derivations for types A3 and B3: / ⫻ / ⫻ —⬎ ⫻ ⫻ / ⫻ (type A3; section 2.3); ⫻ / ⫻ / —⬎ ⫻ ⫻ ⫻ / (type B3; section 2.5).

MOSP-02.qxd

5/11/04

2:38 PM

84

Page 84

2. Metrical types and positions

Table 2.51. Distribution of x . . . Pxx#P according to the number of syllables in the first drop number of x

a-verse with double alliteration

a-verse with single alliteration

1 2 3 4 5 6 7 8 9 10

2 40% 4 13% 2 5% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%

2 40% 15 47% 10 24% 5 15% 2 8% 0 0% 0 0% 0 0% 1 100% 0 0%

total

8

35

5%

22%

b-verse

total

1 20% 13 41% 29 71% 28 85% 24 92% 11 100% 5 100% 2 100% 0 0% 0 0% 113

72%

5 32 41 33 26 11 5 2 1 0

100% 100% 100% 100% 100% 100% 100% 100% 100% 0%

156 100%

Table 2.52. Distribution of x . . . PXx#P according to the number of syllables in the first drop number of x

a-verse with double alliteration

a-verse with single alliteration

b-verse

total

1 2 3 4 5 6 7 8 9

1 8% 3 27% 1 14% 0 0% 1 12.5% 0 0% 0 0% 0 0% 0 0%

2 3 0 1 1 0 0 0 0

10 77% 5 45% 6 86% 5 83% 6 75% 4 83% 2 100% 1 100% 0 0%

13 11 7 6 8 4 2 1 0

total

6 12%

7 13%

39

52 100%

15% 27% 0% 17% 12.5% 0% 0% 0% 0%

75%

100% 100% 100% 100% 100% 100% 100% 100% 0%

To be sure, the complete absence of double alliteration in the configuration at issue would seem consistent with the suggested scansion. A fundamental difficulty, however, is attached to this scansion: the configuration x . . . PxS is attested in the b-verse as well as in the a-verse with single alliteration; by contrast, types A3 and B3, the other two configurations with a single lift, are categorically excluded from the b-verse, because alliteration has to occur on the penultimate lift of the line, namely, the first of the two lifts present in the b-verse. In other words, the b-verse is structurally incompatible with the absence of a second lift. That the configuration x . . . PxS may be used in the b-verse therefore demonstrates that it consists of

MOSP-02.qxd

5/11/04

2:38 PM

Page 85

2.4. Type B1

85

Table 2.53. Distribution of x . . . Pxx#P according to the number of syllables in the first drop number of x

a-verse with double alliteration

a-verse with single alliteration

b-verse

total

1 2 3 4 5 6 7 8 9

1 0 1 2 1 0 0 0 0

1 50% 0 0% 0 0% 0 0% 1 17% 0 0% 1 100% 0 0% 0 0%

0 10 8 7 4 4 0 0 1

0% 100% 89% 78% 67% 100% 100% 0% 100%

2 10 9 9 6 4 1 0 1

total

5 12%

3

34

81%

50% 0% 11% 22% 17% 0% 0% 0% 0%

7%

100% 100% 100% 100% 100% 100% 100% 0% 100%

42 100%

Table 2.54. Distribution of x . . . PxS according to the number of syllables in the first drop number of x

a-verse with double alliteration

a-verse with single alliteration

b-verse

total

1 2 3 4 5 6 7 8 9

0 0 0 0 0 0 0 0 0

4 80% 13 76% 19 76% 21 68% 2 67% 0 0% 3 100% 1 50% 0 0%

1 20% 4 24% 6 24% 10 32% 1 33% 2 100% 0 0% 1 50% 0 0%

5 17 25 31 3 2 3 2 0

total

0 0%

63

25

88 100%

0% 0% 0% 0% 0% 0% 0% 0% 0%

72%

28%

100% 100% 100% 100% 100% 100% 100% 100% 0%

two lifts, and accordingly that the final position occupied by S constitutes the second lift.10 The unique distribution pattern of the configuration x . . . PxS in distinction from other variants of type B1 is also observed in Beowulf (see Suzuki 1996a: 345). When the second lift is occupied by a primary-stressed syllable, more than seven out of every ten instances are b-verses; by contrast, when the position in question is filled by

10

The present argument against scanning the configuration x . . . PxS as a variant of type D also applies to Hofmann’s identification of the configurations x . . . PSx, x . . . Psx, x . . . PXx, and x . . . Pxx as variants of the same metrical type (type dax3a in Hofmann’s term), rather than of type C (section 2.6).

MOSP-02.qxd

5/11/04

2:38 PM

86

Page 86

2. Metrical types and positions

Table 2.55. Distribution of x . . . Px#xP according to the number of syllables in the first drop number of x

a-verse with double alliteration

a-verse with single alliteration

1 2 3 4 5 6 7 8 9 10 11

6 38% 12 16% 9 7% 9 9% 4 4% 3 4% 0 0% 0 0% 0 0% 0 0% 0 0%

7 11 20 15 10 3 2 1 0 0 0

total

43

69 13%

8%

44% 14% 16% 15% 11% 4% 8% 5% 0% 0% 0%

b-verse

total

3 19% 54 70% 94 76% 77 76% 81 85% 65 92% 23 92% 19 95% 5 100% 1 100% 1 100% 423

79%

16 77 123 101 95 71 25 20 5 1 1

100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%

535 100%

a non-primary-stressed syllable, we obtain a reverse pattern, whereby the a-verse accounts for more than 60 per cent of all instances of the configuration in question. As discussed in Suzuki (1996a: 344–6), the demonstrably fewer occurrences of the configuration x . . . PxS in the b-verse is attributed to the structural rigidity and unmarked status of the b-verse that is less compatible with the marked association of the lift with a non-primary-stressed syllable. Thus, the Heliand poet seems to build on the traditional practice in distinguishing between the sequences x . . . PxP and x . . . PxS by polarising their distribution pattern to a larger extent.11 Also standing out from others is the configuration x . . . PXx#P (Hofmann’s B5x3a.1), though the distinction concerns a few minor points. First, even when the first drop is realised by a monosyllable, this subtype exhibits preference for the b-verse contrary to the general pattern noted above. A second peculiarity of the configuration is the low incidence of the b-verse and the inversely high incidence of the a-verse when the first drop is occupied by two syllables. In section 2.4.2.3 below, these and other characteristics will be shown to be consequences of the structural ambivalence of the configuration x . . . PXx#P, which stands at the intersection of types B1 and E. The inverse proportion of the size of the first drop to the preference for the a-verse with double alliteration that we find characteristic of type B1 in general would strike us as peculiar when we recall that the converse relation obtains elsewhere: for type A1 the larger the number of syllables associated with the first drop the higher the frequency of the a-verse with double alliteration (section 2.1); and for types D and E the more prominent the first drop is, the more likely it occurs in the a-verse with

11

In section 2.6 below, we will return to this issue by relating it to a wider context of reorganisation involving comparable variants of type C.

MOSP-02.qxd

5/11/04

2:38 PM

Page 87

2.4. Type B1

87

Table 2.56. Distribution of the varying number of syllables in the second drop of type B1 for the a-verse with double alliteration, the a-verse with single alliteration, and the b-verse number of x

a-verse with double alliteration

a-verse with single alliteration

b-verse

1 2 3 4 5

169 68% 67 27% 6 2% 4 2% 1 less than 1%

371 73% 127 25% 6 1% 1 less than 1% 0 0%

1689 71% 658 28% 41 2% 2 less than 1% 0 0%

total

247 100%

505 100%

2390 100%

Table 2.57. Verse distribution according to the number of syllables in the second drop of type B1 number of x

a-verse with double alliteration

a-verse with single alliteration

b-verse

1 2 3 4 5

169 8% 67 8% 6 11% 4 57% 1 100%

371 127 6 1 0

1689 658 41 2 0

17% 15% 11% 14% 0%

total 76% 77% 77% 29% 0%

2229 852 53 7 1

100% 100% 100% 100% 100%

double alliteration (sections 2.7 and 2.9). However, type C is treated much in the same way as type B1 (section 2.6.1); so is anacrusis, as demonstrated in section 2.11 below. Such divergence in distribution in terms of the size of the first drop will become subject to a principled account after we have placed in perspective a whole array of relevant distribution patterns in section 2.12 below. 2.4.2. The second drop of type B1 2.4.2.1. The range of variation on the number of syllables The linguistic realisation of the second drop of type B1 differs significantly from that of the first drop with respect to the number of syllables involved as well as the distribution of the varying number of syllables in the three verse categories. The second drop of type B1 has access to a far more limited range of syllable sequences for association than the first drop (Tables 2.56 and 2.57). The maximum number of syllables appearing in it is five, attested in a single verse (3097a githolos undar thesaru thiod), as against eleven for the first drop, also attested singly (1494b than ne sî he imu eo sô suuîðo an sibbiun bilang). The commonest number of syllables realised is one for the second drop, three for the first counterpart. And for all the three verse categories alike, the vast majority of instances, going well beyond 90 per cent, contain one or two syllables for the second drop. By contrast, the frequency in occurrence of one or two syllables in the first drop ranges between slightly over 50 per cent (the a-verse with double alliteration) and under 20 per cent (the b-verse), as indicated in Table 2.46 above.

MOSP-02.qxd

5/11/04

88

2:38 PM

Page 88

2. Metrical types and positions

In regard to verse distribution according to the varying number of syllables used as an occupant of the second drop (Table 2.57), the b-verse invariably accounts for about three quarters of the total inasmuch as the syllables involved in the sequence are three or less. Thus, as far as the number of syllables associated is relatively small, the second drop of type B1 is insensitive to the size of the syllable sequences appearing in it, and accordingly hardly affected by the kind of polarisation between the a-verse and the b-verse we find characteristic of the first drop of type A1 (section 2.1). For the longer sequences consisting of four or five syllables, however, the pattern reverses in favour of the a-verse with double alliteration. Yet the rarity of such longer sequences makes polarisation between the a-verse and the b-verse less pronounced than for the first drop of type A1. The whole distribution pattern thus obtained for the second drop stands diametrically opposed to that for the first drop, for which the longer the sequence is, the higher the proportion for the b-verse is, as observed in section 2.4.1 above. The limited size of syllable sequences available to the second drop for realisation is a consequence of the difference in relative metrical strength between the first and second drops, which is expressed in more general terms in the Linearity-based Metrical Strength Scale (section 1.2). According to the scale, the metrical positions of the same category (the lift, the normal drop, or the heavy drop) are finely graded in terms of their strength according to their location in the verse: the closer a position is to the beginning of the verse, the greater strength it has at its disposal, and consequently the more extensive and varied its realisation forms are. Martin (1907: 53–4) and Russom (1998: 139–44) among other metrists call into question the very legitimacy of type B1 verses with more than two syllables appearing in the second drop. Martin proposes rescanning these long type B1 verses as type E. As will be discussed at length later in this section, however, his proposal proves to be metrically untenable. On the other hand, Russom invokes three manipulative operations to reduce these verses to the commoner configuration that has the same position realised by a disyllabic sequence. Russom (1998: 141–3) lists thirty-four such type B1 verses all together. First, by appealing to elision of two adjacent unstressed vowels as suggested long ago by Kauffmann (1887: 324; cf. Martin 1907: 53–4), Russom regards verses like 2345b thea he cûðde obar al as containing two syllables in the second drop. Russom (1998: 141 n. 26) finds thirteen such examples: 591a, 512b, 874b, 890b, 2345b, 2545b, 3220b, 3230b, 3386b, 3581b, 3646b, 4009b, 5376b. Second, Russom observes that, when attested in both manuscripts, the trisyllabic second drop is often found only in the M manuscript. Drawing then on Sievers’s (1876) view that the C scribe seemed more reliable in metrical terms, Russom (1998: 141–2) adopts readings based on the C manuscript and therefore excludes the long type B1 scansion from the following eight verses: 1890a (M getholon undar theru thiod; C githolon under thero thioda), 3097a, 3535a, 4898a, 1699b, 1889b, 2317b, 4369b. Third, by referring to a scansion based on syncopation in Old English metre, Russom (1998: 142–3) disregards as purely scribal what would appear to be a third unstressed syllable in the second drop of type B1, as in verse 5045b thô imu is thiodanes gisuêk. Thirteen examples (4064a, 279b, 1571b, 1594b, 2171b, 3028b, 3107b, 3838b, 4270b, 4423b, 4505b, 5045b, 5846b) are thus claimed to be restored to their originals with the second drop realised by two syllables.

MOSP-02.qxd

5/11/04

2:38 PM

Page 89

2.4. Type B1

89

Several objections may be raised to Russom’s reductive approach. First, Russom’s particular scansions are largely based on Hofmann’s. Given an array of configurations x . . . Px . . . P with more than two unstressed syllables standing between the two lifts, however, Hofmann generally favours scansions as type E with anacrusis over type B1, interpretations that will be shown to be inappropriate in our discussion on the configurations x . . . Pxx#xP, x . . . Psx#xP, and x . . . PSx#xP in section 2.4.2.3 below. Taking this into consideration, and re-examining the corpus, we obtain double the number of type B1 verses with the second drop realised by three or more syllables, amounting to sixty in all.12 To be added to Russom’s list are thus the following verses: 1495a ne iro mâgskepi sô mikil, 1602a, 1607a, 2231a, 2263a, 2429a, 2522a, 2704a, 3415a, 4483a, 4593a, 4713a; 633b, 662b, 1061b, 1552b, 2028b, 2183b, 2350b, 2381b, 2592b, 2615b, 3114b, 4928b, 5394b, 5680b, 5944b. Second, there does not seem to be any strong independent reason for postulating elision as a significant metrical device for Old Germanic metre in general and the Heliand metre in particular. Third, inasmuch as we use Behagel’s text as the standard edition and take into full account the uniqueness of the C manuscript in terms of date and provenance (as noted in section 1.1 above), we see no point in replacing the adopted readings taken from the M manuscript with those found in C, even when they would strike as simply idiosyncratic or anomalous in metrical form or other. Finally, we should better adhere to the manuscript reading with minimal interference with the transmitted text; we ought to be particularly careful about making emendations for reasons of metre when engaged in metrical studies. After all, Russom’s textual manipulations are motivated by theory-internal concerns, necessitated as they are by his own metrical theory, which ‘predicts’ that the absence of the word form PxxxS in Old Saxon should be incompatible with the occurrence of three or more unstressed syllables in the second drop of type B1 (cf. Russom 1998: 139). Unless one is subscribed to this theory, however, one is free to consider the data as they are recorded in the text. The verses of the composition x . . . PxxxP are therefore as well-qualified for being scanned as type B1 as are those with one or two syllables in the second drop (x . . . Px(x)P). 2.4.2.2. The configurations x . . . PS#P and x . . . PS#px The variation on the realisation of the second drop of type B1 concerns not only the number of syllables used as examined in the preceding section, but their prosodic and morpholexical status in ways somewhat similar to the treatment of the first drop of type A1 discussed in section 2.1 above. Of particular interest is a parameter-based gradation of the variety of non-primary-stressed syllables belonging to the same word as the preceding lift. In what follows, then, we examine at length the behaviour of these varied word constituents that realise the second drop of type B1. The second drop of type B1 may be filled by the second element of a disyllabic compound (PS): (44) 221a Thô sprac ên gêlhert man 3794b the thar anduuard stôd

12

Verse 1889b thar sculun gi arbidies sô filu should be subtracted from the list, since it is subsumed under the ambivalent configuration x . . . PXx#(x)P (see section 2.4.2.3 below).

MOSP-02.qxd

5/11/04

90

2:38 PM

Page 90

2. Metrical types and positions

Table 2.58. Distribution of x . . . PS#P (to be modified as Table 2.62 below) a-verse with double alliteration

a-verse with single alliteration

b-verse

12 (32%)

4 (11%)

21 (57%)

a-verse with double alliteration: 23a, 1205a, 1217a, 1558a, 1624a, 1740a, 1763a, 1789a, 3515a, 3779a, 4208a, 5702a; a-verse with single alliteration: 221a, 2416a, 2555a, 3100a; b-verse: 106b, 255b, 1323b, 1458b, 2137b, 2534b, 2658b, 3128b, 3299b, 3333b, 3414b, 3437b, 3467b, 3794b, 4182b, 4283b, 4323b, 4541b, 4828b, 5139b, 5615b

We find similar verses in Beowulf, too (Suzuki 1996a: 92), as given below: (45) Beo 202a Ðone siðfæt him 202a, 615a, 821a, 232b, 537b, 1940b, 2109b Alongside these verses, there are a comparable number of verses that contain a resolvable disyllable in final position instead of a long stressed syllable as above. Such verses, as follows, may be scanned as type B1 or subtype A1s with anacrusis (Sievers 1893: §109): (46) 195a suîðo godcund gumo 450a untthat that friðubarn godes 2737b uuarð im thar gladmôd hugi a-verse with double alliteration: 195a, 667a, 4870a; a-verse with single alliteration: 450a, 1128a, 1140a, 3899a, 4024a; b-verse: 527b, 760b, 1011b, 1156b, 1163b, 1872b, 2099b, 2737b, 2813b, 3022b, 3469b, 3529b, 3559b, 3767b, 3789b, 3836b, 3943b, 4292b, 4494b, 4607b, 4938b, 5349b, 5526b, 5829b, 5932b, 5982b Kauffmann (1887: 299–308), Martin (1907: 51), and Hofmann (1991: 34; 1991, II: 60–1) scan the above verses as subtype A1s, whereas Sievers (1893: §107.2) identifies them with type B1. Three arguments might be adduced for scanning the above verses as subtype A1s with anacrusis. First, since anacrusis may generally appear at the beginning of any verses that start with a lift (section 2.11), there seems no principled reason why it should not occur before subtype A1s as well. On the view that the verses listed in (46) above constitute type B1 rather than subtype A1s, one could scarcely account for why anacrusis may not affect subtype A1s. Second, to scan these verses as type B1 would have to be faced with the otherwise unknown property of type B1: when the second drop is occupied by the second element of a compound, the immediately following, second lift would be just about as likely to be associated with a resolvable disyllable as with a long syllable. Resolution, however, is generally thought to be a marked phenomenon and a long stressed syllable is accordingly expected to occur more frequently than a resolved disyllable. On the other hand, to scan verses like 195a in (46) above as subtype A1s with anacrusis would readily deal with the absence of comparable type B1 verses (x . . . PS#px): since resolution is a marked phenomenon, it would seem natural that it does not happen to be attested as a special case of the marginal variant x . . . PS#P.

MOSP-02.qxd

5/11/04

2:38 PM

Page 91

2.4. Type B1

91

Third, the kind of compounds (PS#) used for the two groups of verse under consideration, x . . . PS#P and x . . . PS#px, is largely different, as illustrated below: for the configuration x . . . PS#P, the majority, namely, thirty out of a total of thirtyseven, are quasi compounds (-lîc, -sam, un-) or adverbs ending in -uuard; by contrast, for the configuration x . . . PS#px true compound nouns or adjectives are used in all but one case (5526b). (47) x . . . PS#P: gêlhert ‘playful’; -lîc: diurlîc ‘precious’; egislîc ‘frightful’; gêstlîc ‘spiritual’; gimêdlîc ‘foolish’; gôdlîc ‘good’; holdlîc ‘gracious’; lêðlîc ‘destructive’; leoblîc/lioflîc ‘lovely’; mislîc ‘different’; seldlîc ‘seldom’; sôðlîc ‘true’; uuîslîc ‘wise’; -sam: langsam ‘lasting’; uuonodsam ‘delightful’; uunsam ‘delightful’; treulôs ‘deceitful’; un-: unhold ‘evil’; unmet ‘immeasurably’; unreht ‘wrong’; -uuard: anduuard ‘present’; geginuuard ‘opposite’; tôuuard ‘ensuing’; uuiðeruuard ‘hostile’; uuîrôc ‘incense’ x . . . PS#px: bihêtuuord ‘threatening word’; ênfald ‘simple’; firinuuerc ‘evil deed’; firiuuit ‘curiosity’; frâhmôd/frômôd ‘happy’; friðubarn ‘child of peace’; gladmôd ‘glad’; godcund ‘divine’; harmuuerc ‘bad deed’; inuuid ‘evil’; lôsuuord ‘evil word’; thrîstmôd ‘confident’; uuilspel ‘welcome news’; uuiðermôd ‘hostile’ This divergence would make it less plausible to assume that the two groups realise the same metrical type, type B1. Since there would appear to be no reason why occurrence of resolution has to determine the choice of words for the immediately preceding positions / ⫻, one might go on to argue, to scan the configuration x . . . PS#px, along with the configuration x . . . PS#P, as type B1 would seem to create more problems than it solves. On closer consideration, however, we would be able to refute the above arguments one by one, and claim on the contrary that the configuration x . . . PS#px should better be scanned as type B1. As will be demonstrated below, our proposed scansion is not only capable of avoiding the alleged difficulties pointed out above but empirically supported by the distributional facts. To address the second objection first, the high incidence of the resolved type B1 variant x . . . PS#px comparable to that of the configuration x . . . PS#P may be attributed to the particular syntagmatic relation involved: the second lift is immediately preceded by a lexical-stressed syllable, which occupies a position which counts as weaker, namely a drop. As will be discussed in section 3.1.3 below, resolution is most likely to be implemented on a position which needs to be provided with extra prominence in order to be realised in its most prominent form in distinction from an adjacent position of lesser yet relatively high metrical strength. Immediately preceded by a weaker position which is filled by a stressed syllable, the second lift may require additional prominence so that it may be correctly perceived as a lift in its maximally salient form. Thus, much as with the configuration Px . . . PS (type A2; section 3.1.3), the immediate adjacency of a lexical-stressed syllable may be held responsible for the frequent occurrence of the resolvable disyllable as the second lift in verse-final position after x . . . PS#. The foregoing consideration furthermore offers a plausible answer to the third question, namely the divergence in lexical items used for PS# in the two supposedly

MOSP-02.qxd

5/11/04

92

2:38 PM

Page 92

2. Metrical types and positions

related configurations, x . . . PS#P and x . . . PS#px. As substantiated in (47) above, the sequence PS# is realised in large measure by quasi compounds (un-, -lîc, -sam) or compound adverbs ending in -uuard in x . . . PS#P on the one hand and by nominal compounds in x . . . PS#px on the other. The differentiation may be ascribed to the specific syntagmatic relation brought to bear in the preceding paragraph. The final position needs extra prominence for its full realisation on the surface due to the presence of -S# immediately preceding it. Other things being equal, then, the final position should best be realised by a resolved disyllable due to the maximal metrical strength achieved with it. When the device for maximising prominence is not in use, however, there is still another way to bring about a similar effect: it is to lower prominence of the immediately preceding position by associating it with a syllable of lesser salience. Because quasi compounds are less salient than true compounds by virtue of the boundedness of one of their constituents (section 2.12.5), and because adverbs belong to class 2 of lexical items (section 3.2.2), use of these less salient words rather than of maximally stressable class 1 words would have presented itself as a viable alternative for effecting a maximal distinguishability between -S# and the following P in the configuration x . . . PS#P. Finally on the first alleged difficulty: scanning the configuration x . . . PS#px, embodied by the verses in (46), as type B1 would entail that anacrusis does not apply to the configuration PS#px, subtype A1s. We would therefore be required to provide a credible account for this restriction against the background of the extended use of anacrusis in the Heliand (section 2.11). The configuration PS#px is characterised as a marked variant of type A1 with a demonstrably diminished frequency in occurrence, as treated in section 2.1.8 above. Since anacrusis in itself constitutes a marked metrical device which is optional in implementation, it should be hardly surprising that it does not occur on the marked configuration at issue in the corpus, even when the device becomes extended in scope in general.13 Truly surprising would be the contrary situation, whereby anacrusis exclusively affects marked variants of given types. At this point, we may offer two pieces of evidence for our claim that the configuration x . . . PS#px should be identified as type B1. Both concern distribution, one on the distinct patterning of the configurations x . . . PS#px and PS#px, the other on the distinct patterning of anacrusis in type A1 in general and the initial syllables in x . . . PS#px. As Table 2.59 indicates, the configurations x . . . PS#px and PS#px are markedly distinguished in verse distribution: while subtype A1s (PS#px) exhibits a slight preference for the b-verse over the a-verse with double alliteration, the configuration x . . . PS#px shows strong preference for the b-verse, with a concomitant avoidance of the a-verse with double alliteration. The divergence in distribution may thus cast doubt on the identification of the two configurations. More important still is that the patterning of x . . . PS#px is highly similar to that of type B1 in general, as observed above. The shared distribution accordingly may lend support to our scansion. Furthermore, if one regarded the sequence x . . . PS#px as subtype A1s with anacrusis, one would run into another disadvantage of having to explain the nearly identical overall frequency of the basic (thirty-nine instances, c. 53 per cent) and anacrustic verses

13

In actuality, however, we seem to have an instance of subtype A1s with anacrusis, as pointed out in note 14 below.

MOSP-02.qxd

5/11/04

2:38 PM

Page 93

2.4. Type B1

93

Table 2.59. Distribution of x . . . PS#px and PS#px configuration

a-verse with double alliteration

a-verse with single alliteration

b-verse

x . . . PS#px PS#px

3 (9%) 13 (33%)

5 (15%) 9 (23%)

26 (76%) 17 (44%)

Table 2.60. Distribution of x . . . PS#px according to the number of syllables in initial position number of syllables

a-verse with double alliteration

a-verse with single alliteration

b-verse

1 2 3 4 5 6

0 (0%) 2 (67%) 1 (33%) 0 (0%) 0 (0%) 0 (0%)

2 (40%) 1 (20%) 1 (20%) 0 (0%) 0 (0%) 1 (20%)

0 (0%) 11 (42%) 6 (23%) 6 (23%) 2 (8%) 1 (4%)

total

3 (100%)

5 (100%)

26 (100%)

(thirty-four instances, c. 47 per cent): such an even distribution is extremely rare in general, and totally unknown in particular to the configurations that fail to show maximal preference for the a-verse with double alliteration (cf. Table 2.94, section 2.11.1). Moving onto the second evidence, the distribution of varying numbers of verseinitial syllables in the configuration x . . . PS#px is characterised by the rarity of monosyllables, as shown in Table 2.60. This stands in sharp contrast to the distribution pattern of anacrustic syllables that are used at the beginning of type A1 verses, as summarised in Table 2.61 and discussed in detail in section 2.11 below (see also Suzuki 1999: 193): the monosyllable is the most frequent realisation of anacrusis for type A1. Such a difference in distribution may suggest that verse-initial syllables in the configuration x . . . PS#px should not be equated with anacrusis, thereby enhancing the credibility of our proposed scansion of it as type B1. All things considered, therefore, the configuration x . . . PS#P includes the sequence x . . . PS#px as a resolved variant, and the whole distribution is obtained as in Table 2.62.14 As mentioned earlier, Beowulf contains only seven instances of the configuration x . . . PS#P, a figure notably lower than the one for the Heliand. Moreover, the

14

The scansion of the configuration x . . . PS#px as type B1 as substantiated above might lend some credibility to scanning verse 2648b that godes barn sehan also as type B1, the only example of the configuration x . . . P#P#px. This verse might then have arisen by substituting -S# with #P# by analogical extension, a mode of generalisation often observed to be in force elsewhere in the metre. However, given the most plausible scansion of the configuration P#P#px as subtype A1s (section 2.1.8) and the monosyllabicity of the verse-initial weak element (that), it would seem more reasonable to identify it as an anacrustic variant of subtype A1s.

MOSP-02.qxd

5/11/04

94

2:38 PM

Page 94

2. Metrical types and positions

Table 2.61. Distribution of type A1 with anacrusis according to the number of anacrustic syllables number of syllables

a-verse with double alliteration

a-verse with single alliteration

b-verse

1 2 3 4 5 6 7 8 9 10

471 (77%) 97 (16%) 27 (4%) 16 (3%) 3 (less than 1%) 1 (less than 1%) 0 (0%) 0 (0%) 0 (0%) 0 (0%)

32 (34%) 16 (17%) 31 (33%) 6 (6%) 3 (3%) 4 (4%) 2 (2%) 0 (0%) 0 (0%) 0 (0%)

148 (37%) 64 (16%) 46 (12%) 38 (10%) 40 (10%) 27 (7%) 24 (6%) 5 (1%) 2 (1%) 2 (1%)

total

615 (100%)

94 (100%)

396 (100%)

Table 2.62. Distribution of x . . . PS#P (revised) a-verse with double alliteration

a-verse with single alliteration

b-verse

15 (21%)

9 (13%)

47 (66%)

distribution pattern in Beowulf seems deviant with the b-verse markedly underrepresented: 14 per cent (one example) for the a-verse with double alliteration, 29 per cent (two examples) for the a-verse with single alliteration, and 57 per cent (four examples) for the b-verse. In view of the extreme rarity of examples as well as of the atypical distribution, we may be led to conclude that the configuration x . . . PS#P constitutes a highly exceptional, ill-integrated variant of type B1 in Beowulf. The extreme marginality of the configuration x . . . PS#P in Beowulf seems fully in keeping with the complete absence of its marked counterpart x . . . PS#px. Thus, the configuration x . . . PS#px, though still marginal, became more common in the Heliand, which may be understood as another manifestation of the expansion in the range of language materials for the normal drop in it. 2.4.2.3. The configurations x . . . Px#xP, x . . . Pxx#xP, x . . . PS#xP, x . . . Psx#(x)P, x . . . PSx#(x)P, and x . . . PXx#(x)P Having firmly established the scansion of the configurations x . . . PS#P and x . . . PS#px as type B1, we move on to consider verses with more than one syllable standing between the two lifts in type B1. Such configurations may be subject to two competing scansions: type B1 and type E with anacrusis (cf. Sievers 1893: §109; Martin 1907: 53; Hofmann 1991: 112–16). Specific configurations that may be open to these two conflicting analyses are the following six: (i) x . . . Px#xP; e.g., 2783b endi it thar theru thiornun fargaf; (ii) x . . . Pxx#xP; e.g., 1571b that gi iuuues drohtines gibed (cf. Kauffmann 1887: 324; Martin 1907: 53); (iii) x . . . PS#xP; e.g., 1242b endi mundburd

MOSP-02.qxd

5/11/04

2:38 PM

Page 95

2.4. Type B1

95

gihêt; (iv) x . . . Psx#(x)P; e.g., 5727b Im ni uuelda thie folctogo thuo; (v) x . . . PSx#(x)P; e.g., 376b he thit erðrîki herod; (vi) x . . . PXx#(x)P; e.g., 2458a undar fîundo folc. The configuration x . . . Pxx#P (e.g., 2850b endi thô te is iungarun sprak), a form minimally distinct from (i) x . . . Px#xP and (ii) x . . . Pxx#xP, is unambiguous in scansion: it scans as nothing other than type B1. There is no example of the configuration Pxx#P that would otherwise be identifiable as a variant of type E without anacrusis (section 2.9). In the complete absence of the sequence Pxx#P, we may be justified in concluding that the configuration x . . . Pxx#P constitutes type B1. Unlike the configuration x . . . Pxx#P noted above, there are indeed instances of Px#xP (e.g., 279a scadouuan mid skimon) on the one hand and of Pxx#xP (e.g., 929a ênhuuilic ni bist) on the other, both being variants of type E without anacrusis, which might be regarded as a basis for x . . . Px#xP (2783b endi it thar theru thiornun fargaf) and x . . . Pxx#xP (1571b that gi iuuues drohtines gibed), that is, their anacrustic counterparts, precisely the scansion that Hirt (1891: 152) claimed. As far as the configuration Px#xP (including PX#xP) is concerned, however, we have only four examples: 39a uualdand gisprak, 279a, 1526b, 3570b. By contrast, there are well over five hundred instances of x . . . Px#xP (Table 2.55, section 2.4.1). Such an extremely unbalanced distribution pattern in favour of the configuration x . . . Px#xP would make it highly unlikely, Hofmann (1991: 112–13) persuasively argues, that the configuration x . . . Px#xP may reduce to Px#xP as a marked (i.e., anacrustic) variant. Still rarer is the configuration Pxx#xP (section 2.9): verse 929a ênhuuilic ni bist would appear the only instance. Since ênhuuilic is a compound, however, it would be safer to conclude that we do not have a single definite example of the configuration Pxx#xP. Accordingly, scanning the configuration x . . . Pxx#xP as an anacrustic variant of type E as do Martin (1907: 53) and Hofmann (1991, II: 208–9) would have to be rejected as empirically unfounded, much as with the subsumption of the configuration x . . . Pxx#P under type E. We should therefore follow Kauffmann (1887: 324) and Sievers (1893: §112.3) in identifying the configuration x . . . Pxx#xP as type B1. As unlikely is the configuration x . . . PS#xP (1242b endi mundburd gihêt) to be identified as an anacrustic counterpart of the type E variant PS#xP (e.g., 1623a grimuuerc fargeban), despite Kauffmann’s (1887: 343–4) and Hofmann’s (1991: 113–14) scansion to the contrary. The distribution of these two configurations is diametrically opposed (Table 2.63): while the a-verse with double alliteration is predominant for PS#xP, the b-verse is prevalent for x . . . PS#xP. As discussed at length in section 3.2.1.3 below, anacrusis tends to reinforce the distribution pattern of a non-anacrustic verse form, rather than reversing the underling pattern. In this light, it would be reasonable to determine that the configuration x . . . PS#xP should be scanned as type B1, rather than type E with anacrusis. The proposed scansion, furthermore, seems to conform to the traditional composition of the same configuration: Beowulf offers a single such verse, Beo 2441a Þæt wæs feohleas gefeoht, which cannot be scanned as anything but type B1, because type E is incompatible with anacrusis in the traditional metre (Suzuki 1996a: 317). Much the same applies to the opposition between the configurations Psx#(x)P (e.g., 1742a uuînberi uuesan) and x . . . Psx#(x)P (e.g., 5727b Im ni uuelda thie folctogo thuo). While the sequence Psx#(x)P is extremely rare in occurrence, these

MOSP-02.qxd

5/11/04

2:38 PM

Page 96

96

2. Metrical types and positions

Table 2.63. Distribution of PS#xP and x . . . PS#xP

PS#xP x . . . PS#xP

a-verse with double alliteration

a-verse with single alliteration

b-verse

8 (67%) 0 (0%)

3 (25%) 1 (25%)

1 (8%) 3 (75%)

PS#xP a-verse with double alliteration: 1397a, 1623a, 2220a, 2323a, 2630a, 4295a, 4320a, 4990a; a-verse with single alliteration: 167a, 1304a, 3244a; b-verse: 2293b x . . . PS#xP a-verse with single alliteration: 1697a; b-verse: 1242b, 3272b, 3680b

Table 2.64. Distribution of Psx#(x)P and x . . . Psx#(x)P

Psx#(x)P x . . . Psx#(x)P

a-verse with double alliteration

a-verse with single alliteration

b-verse

5 (83%) 6 (29%)

0 (0%) 1 (4%)

1 (17%) 14 (67%)

Psx#(x)P a-verse with double alliteration: 1098a, 1742a, 2993a, 3001a, 3033a; b-verse: 3697b x . . . Psx#(x)P a-verse with double alliteration: 1495a, 2646a, 2704a, 3968a, 4450a, 5827a; a-verse with single alliteration: 5226a; b-verse: 322b, 363b, 895b, 2666b, 3940b, 4135b, 4474b, 4487b, 5263b, 5369b, 5727b, 5793b, 5875b, 5886b

rare instances show a marked preference for the a-verse with double alliteration (Table 2.64). By contrast, the configuration x . . . Psx#(x)P is attested more than three times as many and characterised by a conversely strong preference for the b-verse. Hence it may follow that the two sequences should not reduce to variants of the same metrical type, type E with and without anacrusis. Rather, the configuration x . . . Psx#(x)P should best be identified as type B1, the characterisation that turns out to be corroborated by the overall distribution pattern of type B1 noted above. Kauffmann’s (1887: 318–24) scansion is thus supported on empirical grounds. By contrast, while properly scanning verses with the second drop associated with a heavy suffix such as 895b Ic bium an is bodskepi herod as type B1 (Hofmann 1991, II: 138–43), Hofmann (1991, II: 205–6, 208–9) incorrectly identifies as type E with anacrusis those verses with the second element of a true compound in the corresponding position, such as verse 5793b thuo sia thena lîchamon thar. The divergent distribution of the configurations Psx#(x)P and x . . . Psx#(x)P thus indicates that the disyllable -sx# may be equated with -Sx# as well as with -S# in metrical terms according to context: while the configuration Psx#P is grouped with PSx#P (type E; section 2.9) and distinguished from PS#P, which is non-existent, the sequence x . . . Psx#P is treated as equivalent to x . . . PS#P (type B1) rather than to x . . . PSx#P (type E with anacrusis; see immediately below). The scansion of the configuration x . . . Psx#P as type B1 as argued above seems to find support in traditional practice as well. In Beowulf, we find a single instance of the same configuration, namely, Beo 1941b þeah ðe hio æ¯nlicu sy. In view of the

MOSP-02.qxd

5/11/04

2:38 PM

Page 97

2.4. Type B1

97

Table 2.65. Distribution of PSx#(x)P and x . . . PSx#(x)P

PSx#(x)P x . . . PSx#(x)P

a-verse with double alliteration

a-verse with single alliteration

b-verse

80 (48%) 20 (47%)

36 (22%) 8 (19%)

51 (31%) 15 (35%)

PSx#(x)P a-verse with double alliteration: 8a, 224a, 237a, 252a, 336a, 537a, 563a, 567a, 590a, 622a, 744a, 1125a, 1133a, 1137a, 1150a, 1242a, 1305a, 1328a, 1411a, 1424a, 1563a, 1641a, 1668a, 1716a, 1775a, 1776a, 1790a, 1799a, 1840a, 1991a, 1997a, 2151a, 2168a, 2183a, 2191a, 2236a, 2239a, 2301a, 2307a, 2352a, 2359a, 2378a, 2414a, 2677a, 2735a, 2797a, 2867a, 2890a, 2891a, 2911a, 2982a, 3019a, 3141a, 3146a, 3399a, 3540a, 3545a, 3578a, 3735a, 3831a, 3908a, 3931a, 4137a, 4178a, 4234a, 4264a, 4535a, 4548a, 4737a, 4853a, 4879a, 4935a, 4977a, 5004a, 5040a, 5090a, 5457a, 5491a, 5559a, 5829a; a-verse with single alliteration: 317a, 352a, 619a, 624a, 750a, 947a, 1182a, 1202a, 1268a, 1272a, 1428a, 1433a, 2132a, 2161a, 2391a, 2432a, 2798a, 2893a, 3078a, 3421a, 3449a, 3594a, 3814a, 4384a, 4721a, 4761a, 4811a, 4893a, 5038a, 5081a, 5241a, 5247a, 5355a, 5363a, 5576a, 5938a; b-verse: 176b, 188b, 297b, 549b, 577b, 614b, 616b, 801b, 861b, 878b, 976b, 1087b, 1251b, 1448b, 1461b, 1504b, 1691b, 1850b, 1885b, 2087b, 2155b, 2487b, 2624b, 2700b, 2705b, 2763b, 2932b, 3305b, 3694b, 4018b, 4050b, 4085b, 4169b, 4186b, 4254b, 4294b, 4297b, 4358b, 4554b, 4597b, 4628b, 4674b, 4747b, 5060b, 5071b, 5132b, 5201b, 5427b, 5546b, 5692b, 5783b x . . . PSx#(x)P a-verse with double alliteration: 483a, 631a, 635a, 1658a, 1721a, 1876a, 2219a, 2548a, 2591a, 2650a, 3270a, 3466a, 4101a, 4122a, 4284a, 4427a, 4730a, 4841a, 5096a, 5452a; a-verse with single alliteration: 159a, 345a, 1058a, 3446a, 4185a, 4527a, 4801a, 5095a; b-verse: 71b, 294b, 376b, 997b, 1035b, 1083b, 1556b, 1644b, 2568b, 2856b, 2860b, 4260b, 5268b, 5399b, 5705b

categorical incompatibility of anacrusis with type E (Suzuki 1996a: 317), this verse has to be identified as type B1. Turning to the pair of configurations PSx#(x)P (e.g., 1133a mancunneas mên) and x . . . PSx#(x)P (e.g., 376b he thit erðrîki herod), we observe that their distribution is nearly identical: the a-verse with double alliteration is most favoured, accounting for about the half of all occurrences. In the light of the close similarity in distribution, as well as in the absence of predominant occurrence of the b-verse characteristic of type B1, we may accordingly assume that the sequence in question constitutes an anacrustic variant of the configuration PSx#(x)P, type E. We are thus in accord with Kauffmann’s (1887: 343–5) and Hofmann’s (1991: 114–16; 1991, II: 205–6, 208–9) analysis in this respect. By contrast, not a single instance of the configuration x . . . PSx#(x)P is attested in Beowulf. This should be hardly surprising, however, given that anacrusis categorically does not apply to type E in it, as noted above (Suzuki 1996a: 317). Finally, in regard to the configuration x . . . PXx#(x)P (e.g., 2458a undar fîundo folc), things become less than clear-cut. There are a total of fifty-eight examples of this configuration in the Heliand, as exemplified below, which are distributed in the three verse categories (the a-verse with double alliteration, the a-verse with single alliteration, and the b-verse) as shown in Table 2.66. On the other hand, the configuration PXx#(x)P (a variant of type E; for details, see section 2.9 below), a sequence

MOSP-02.qxd

5/11/04

2:38 PM

98

Page 98

2. Metrical types and positions

Table 2.66. Distribution of PXx#(x)P and x . . . PXx#(x)P

PXx#(x)P x . . . PXx#(x)P

a-verse with double alliteration

a-verse with single alliteration

b-verse

21 (20%) 7 (12%)

32 (30%) 8 (14%)

52 (50%) 43 (74%)

PXx#(x)P a-verse with double alliteration: 575a, 689a, 1047a, 1357a, 1450a, 2196a, 2259a, 2331a, 2398a, 3004a, 3789a, 4009a, 4588a, 4592a, 4814a, 4931a, 5224a, 5488a, 5638a, 5744a, 5914a; a-verse with single alliteration: 327a, 332a, 460a, 680a, 686a, 1049a, 1466a, 1965a, 2179a, 2278a, 2306a, 2456a, 3058a, 3079a, 3102a, 3136a, 3180a, 3383a, 3416a, 3729a, 3918a, 4102a, 4238a, 4293a, 4609a, 4715a, 4887a, 5541a, 5815a, 5819a, 5848a, 5978a; b-verse: 11b, 21b, 90b, 179b, 277b, 335b, 381b, 469b, 701b, 962b, 1026b, 1050b, 1294b, 1403b, 1984b, 2004b, 2030b, 2237b, 2248b, 2494b, 2695b, 2803b, 2912b, 3031b, 3057b, 3607b, 3618b, 3638b, 3671b, 3705b, 3758b, 3780b, 3818b, 3863b, 3917b, 3921b, 3922b, 4027b, 4168b, 4528b, 4708b, 4759b, 4803b, 4886b, 4965b, 5143b, 5265b, 5422b, 5441b, 5524b, 5684b, 5983b x . . . PXx#(x)P a-verse with double alliteration: 520a, 1190a, 2458a, 3587a, 4607a, 5032a, 5134a; a-verse with single alliteration: 840a, 1267a, 1279a, 3422a, 3822a, 3966a, 4949a, 5086a; b-verse: 28b, 52b, 121b, 202b, 300b, 467b, 475b, 782b, 1041b, 1186b, 1189b, 1222b, 1264b, 1365b, 1883b, 1389b, 2602b, 2688b, 2965b, 3070b, 3256b, 3334b, 3444b, 3620b, 3681b, 3793b, 3991b, 4079b, 4116b, 4137b, 4210b, 4372b, 4848b, 4857b, 5064b, 5163b, 5380b, 5381b, 5602b, 5638b, 5660b, 5703b, 5934b

minimally differentiated from x . . . PXx#(x)P by the absence of initial weak elements, is distributed in the manner not radically different from that of the configuration x . . . PXx#(x)P. In sharp contrast to the frequency of the configuration Px#xP relative to that of x . . . Px#xP, the proportion of the verse form with initial weak elements (x . . . PXx#(x)P) to the one without (PXx#(x)P) would appear in keeping with the general pattern obtained for basic and anacrustic variants (section 2.11.1). Furthermore, the occurrence of weak elements in initial position is accompanied with a decrease in incidence of the a-verse with double alliteration in favour of the b-verse. As substantiated in section 3.2.1.3 below, such a trading-off relationship is characteristic of the anacrustic verses that are based on the variants with markedly low incidence of the a-verse with double alliteration. Since the configuration PXx#(x)P, a variant of type E, is characterised by the underrepresentation of the a-verse with double alliteration, the overall distribution pattern of the configuration x . . . PXx#(x)P proves to be conforming to the general picture. Brought together, then, these observations might lead us to conclude that the configuration x . . . PXx#(x)P constitutes an anacrustic type E variant, rather than a type B1 variant. There are, however, contrary indications that cast doubt on the above characterisation. First, in contrast to the minimally distinct sequence x . . . PSx#(x)P treated above, the overall distribution of the configuration x . . . PXx#(x)P is found analogous to that of the group of type B1 variants examined earlier. For the sake of comparison, we may reproduce relevant data in Table 2.67. Not only analogous in broad shape, the distribution of the configuration x . . . PXx#(x)P is subject to the same specific generalisation that the others follow:

MOSP-02.qxd

5/11/04

2:38 PM

Page 99

2.4. Type B1

99

Table 2.67. Distribution of type B1 variants x . . . Px#P, x . . . PX#P, x . . . Pxx#P, and x . . . PXx#(x)P variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

x . . . Px#P x . . . PX#P x . . . Pxx#P x . . . PXx#(x)P

138 (8%) 3 (3%) 8 (5%) 7 (12%)

275 (16%) 6 (6%) 35 (22%) 8 (14%)

1327 (76%) 92 (91%) 113 (72%) 43 (74%)

the use of a long medial syllable instead of a short one favours the b-verse more strongly. Thus, much as the sequence x . . . PX#P is distinguished from its minimally distinct counterpart x . . . Px#P by the greater frequency of the b-verse, the configuration x . . . PXx#(x)P exhibits a higher incidence of the b-verse than its minimal opposite, x . . . Pxx#P. Such a structural parallelism may plausibly suggest that the configuration x . . . PXx#(x)P constitutes a firmly integrated type B1 variant. Second, as shown in Table 2.68 through Table 2.70, the configuration x . . . PXx#(x)P shows closer resemblance to the configuration x . . . Pxx#(x)P than to x . . . PSx#(x)P as regards the distribution of the varying number of unstressed syllables in verse-initial position. With respect to the composition of the a-verse, the distribution pattern for x . . . PSx#(x)P exhibits a marked concentration on the two lowest numbers (one and two; Table 2.70), a distinguishing feature of anacrusis (section 2.11.1). On the other hand, the other two configurations (x . . . PXx#(x)P, x . . . Pxx#(x)P) are distinguished by a more diffused distribution (Tables 2.68 and 2.69). Moreover, the single syllable does not rank the highest for these two configurations, in distinction from the demonstrably anacrustic configuration x . . . PSx#(x)P. Likewise, the b-verse is characterised by the same opposition between compact and diffuse distribution, although the range of distribution involved is more extended, which constitutes a recurrently observed feature of the b-verse. We are accordingly confronted with the contradictory indications as to the categorisation of the configuration x . . . PXx#(x)P, type E with anacrusis (comparable to x . . . PSx#(x)P) on the one hand or type B1 (comparable to x . . . Pxx#(x)P) on the other. In the absence of conclusive evidence, and in view of the prototype-based non-categorical systematicity of the metre (section 1.2), it would seem most reasonable to conclude that the configuration under consideration is structurally ambiguous and indeterminate in scansion: standing on the boundary between the two metrical categories (types B1 and E), the sequence at issue appears compatible with either scansion on different grounds, yet it is not fully subsumable under either of them. We are thus in disagreement with Kauffmann (1887: 318–24) and Hofmann (1991: 116; 1991, II: 138–42), who scan the majority of these verses as type B1. Yet in fairness Hofmann (1991: 115) seems well aware of the fuzzy distinction between type B1 and type E with anacrusis. On the other hand, Hirt (1891: 150) argues for the scansion as type E, which cannot be accepted, either. In contrast, Sievers (1893: §107.3) seems cautious enough to avoid categorical decision. From a comparative perspective, the wide attestation of the configuration x . . . PXx#(x)P sharply distinguishes the Heliand from Beowulf. As discussed in

MOSP-02.qxd

5/11/04

2:38 PM

Page 100

100

2. Metrical types and positions

Table 2.68. Distribution of the varying number of verse-initial unstressed syllables in x . . . PXx#(x)P number of x 1 2 3 4 5 6 7 8

a-verse 3 6 2 1 3 0 0 0

total

20% 40% 13% 7% 20% 0% 0% 0%

15 100%

b-verse 10 6 7 6 7 4 2 1

23% 14% 16% 14% 16% 9% 5% 2%

43 100%

Table 2.69. Distribution of the varying number of verse-initial unstressed syllables in x . . . Pxx#(x)P and x . . . Pxx#(x)P number of x

a-verse

1 2 3 4 5 6 7 8 9 10

6 21 13 7 4 0 1 0 1 0

total

54 100%

11% 39% 24% 13% 7% 0% 2% 0% 2% 0%

b-verse 1 23 46 37 31 16 5 2 1 0

1% 14% 28% 22% 19% 10% 3% 1% 1% 0%

165 100%

Suzuki (1996a: 93), Beowulf offers only four comparable examples (Beo 501b wæs him Beowulfes sið, 932b, 949b, 1830b).15 In the face of the structural indeterminacy distinctively characteristic of the configuration x . . . PXx#(x)P, as well as of its increased occurrences in the Heliand, we should address the following general questions: (i) why is this configuration, but not others, involved in categorial ambiguity? (ii) what was responsible for the extensive

15

The existence of the same configuration in Beowulf might possibly be adduced as further support for scanning it as type B1, rather than type E with anacrusis. In the presence of convincing evidence to the contrary as discussed above, however, the practice in Beowulf would hardly be of decisive significance.

MOSP-02.qxd

5/11/04

2:38 PM

Page 101

2.4. Type B1

101

Table 2.70. Distribution of the varying number of verse-initial unstressed syllables in x . . . PSx#(x)P number of x

a-verse

1 2 3 4 5 6 7 8

14 7 5 0 0 1 1 0

total

28 100%

50% 25% 18% 0% 0% 4% 4% 0%

b-verse 3 7 2 2 0 0 0 1

20% 47% 13% 13% 0% 0% 0% 7%

15 100%

use of this configuration in the Heliand? The first question may be rephrased more specifically in regard to the organisation of the second drop of type B1, the issue we are directly concerned with in this section: why is the configuration x . . . PXx#(x)P scannable as type B1 as opposed to the minimally distinct configuration x . . . PSx#(x)P, which is not? To provide a principled explanation, we may focus on the word internal structures involved in the minimally distinct configurations x . . . PXx#(x)P and x . . . PSx#(x)P, namely -Xx# (constitutive of a non-compound) and -Sx# (constitutive of a compound). As substantiated above, the opposition between -X- and -S- determines metrical categorisation: x . . . PXx#(x)P is indeterminate in scansion, while x . . . PSx#(x)P scans definitely as type E with anacrusis. Such a sharp distinction between the sequences -Xx# and -Sx#, however, is recurrently in evidence in the metre: with type A1, while both structures may realise the first drop, the one constitutive of a non-compound (-Xx#) is distinguished by its weaker preference for the a-verse with double alliteration than that of a compound (-Sx#; section 2.1.7); and in types D and E, as fully discussed in sections 2.7 and 2.9 below, the constituent of a non-compound (-Xx#) exhibits a strong preference for the b-verse, and the one of a compound (-Sx#) for the a-verse with double alliteration. Since type B1 as a whole is characterised by its demonstrably strong preference for the b-verse, one might wish to ascribe the avoidance of the disyllabic sequence constitutive of a compound (-Sx#) to this inherent property of type B1. By examining fully the nature of the second drop of type B1 in structural terms, however, we may arrive at a deeper understanding of the issue, as shown below. The increased occurrence of the sequence -Xx# in the second drop of type B1 in the Heliand may be understood as another manifestation of category extension: the second drop of type B1 would have been generalised in such a way that it became capable of association with a greater range of language material than traditionally permitted. As remarked above, the second drop of type B1 in Beowulf is highly restricted regarding its linguistic realisation: the lexical-stressed configurations -S#, -sx#, and even the non-lexical-stressed one -Xx# are vanishingly rare in

MOSP-02.qxd

5/11/04

102

2:38 PM

Page 102

2. Metrical types and positions

occurrence. Thus, in Beowulf the second drop of type B1 is hardly compatible with the presence of a foot, precisely as the normal drop in general in the traditional metre. In the Heliand, by contrast, the presence of a foot that is constructed over a long medial -X- does not disqualify for mapping onto the second drop of type B1. This relaxation on qualifications for the association with the position at issue makes sense given that the normal drop in general allows for association with footed as well as unfooted materials in the Heliand. More specifically, the widespread occurrence of the long medial syllable -X- in the second drop of type B1 would have been facilitated by the indistinguishability between -Xx# on the one hand and -X# and -xx# on the other in terms of foot organisation. The loss of Defooting in Old Saxon phonology, which contributed to the restoration of syncopated vowels (section 1.3.1), would have deprived the medial long syllable -X- of its distinguishability from the latter two strings because they were all footed in the new phonology. Since the strings -X# and -xx# were qualified for association with the second drop of type B1 in the traditional metre, the long medial syllable lacking lexical stress -X- came to be treated analogously so that -Xx# gained in acceptability as a realisation of the same position and became correspondingly more common in occurrence. By contrast, the string -Sx# remained unqualified to realise the position at issue, because -S- was still sharply differentiated from -X-, -X#, and -xx# through a lexical stress falling on it. Furthermore, in the light of the prototypical realisation of the normal drop by the word-final unstressed syllable, the sequence -Xx# was counted as less deviant from the prototype -x# than the string -Sx#, and therefore would have met with lesser resistance in occupying the second drop of type B1. On the other hand, the sequence -Sx#, far more removed from the prototype, remained prohibited from association. Thus, while largely limited to unfooted material for realisation in the traditional metre, the second drop of type B1 became compatible through reconfiguration with any language material lacking lexical stress in the Heliand metre. Although the preference of -Xx# to -Sx# as an occupant of the second drop of type B1 is understandable by reference to its greater similarity to the prototype due to the absence of lexical stress, we are still left with a problem to be solved. As discussed in section 2.1 above, type A1 allows the first drop to be filled by the string -Sx#. By contrast, the same string is unqualified for mapping to the second drop of type B1, as the configuration x . . . PSx#(x)P is excluded as a legitimate variant of type B1. The complete absence of the sequence -Sx# as an occupant of the second drop of type B1 then calls for a fuller explanation. The question to be explored is: what is the structural ground for such divergent treatment in types A1 and B1? We may provide a principled account of the divergence by bringing to bear on the differing paradigmatic relations involving the two drops under discussion within the traditional metre: while the first drop of type A1 stood in opposition to the heavy drop of type A2a, the second drop of type B1 did not have an analogous structural opposite (it will be recalled that the configuration ⫻ / \ / was non-existent; for discussion, see Suzuki 1996a: 93–4, 134–5). The unavailability of the configuration ⫻ / \ / means that the third position of type B1 in general was not realised by a lexical-stressed syllable, apart from isolated exceptions. Thus, the traditional versecraft simply failed to offer a stock of exemplary instances on the basis of which the innovative poet would have composed novel yet comparable verses

MOSP-02.qxd

5/11/04

2:38 PM

Page 103

2.4. Type B1

103

by slightly reorganising the existing body of metrical knowledge; to allow the sequence -Sx# to fill the drop in question would have struck the poet as too radical a departure from the norm and hence an unacceptable practice. By contrast, in the presence of the traditionally acknowledged type A2a alongside type A1, the poet would have readily encountered numerous examples of the form (PS#Px) along with those of Px . . . Px, which he would then have felt free to reconfigure by systematising the underlying metrical form as deemed appropriate. More specifically, the poet arrived at a fuller integration of the wide variation on the realisation of the first drop of type A1 by minimally altering the conventional range of metricality. At issue in the case of the first drop of type A1 was therefore the reconfiguring of the traditional system without creating radically new, totally unprecedented verses. Stated in formal terms, the first drop of type A1 became subject to the obliteration of the categorical distinction between normal and heavy drops, and thereupon to a consequent restructuring based on a fine gradation patterning. Thus, the reorganising process did not directly disrupt the traditional body of verse expressions.16 Standing in contrast to the categorical exclusion of -Sx# is the still marginal but increasingly acceptable use of -S# and -sx# (section 2.4.2.2) in the second drop of type B1. The diversified treatment of these three strings may be ascribed to the absence versus presence of an unstressed syllable in word-final position. The association with the foot expanded with the word-final syllable (-Sx#) would result in a greater deviation from the prototypical realisation of the normal drop (-x#) than that with a single foot standing alone (-S# or -sx#): a foot in combination with a following unstressed syllable would be perceived as relatively more prominent than a minimal bimoraic foot pure and simple; as a consequence, the string exceeding the minimal quantity of a bimoraic foot would be counted as more aberrant as a realisation of a normal drop than the minimal feet -S# and -sx#. In addition to the excessive prominence on the string -Sx# would be also at work the prototypical metrical values of its constituent syllables -S- and -x#. Notably, the presence of the word-final unstressed syllable -x# in the sequence -Sx# would prompt association with a normal drop on its own by virtue of its prototypical metrical value, thereby potentially dissolving the unity of the sequence as a single metrical position. Put another way, the string -Sx# would be inherently more likely to fill two separate positions, their respective prototypical metrical positions, namely, a heavy drop and a normal drop, than a single normal drop. Unlike types A1 and D, however, such peripheral members of type B1 are no different from other more central realisations in terms of their verse distribution patterning: the b-verse is at least as twice in number as the a-verse. The insensitivity of the second drop in type B1 to varying language materials in terms of verse distribution may also be attributed to the absence of a heavy drop as an integral position for class B: in the absence of the heavy type of class B that would have been distinguished as type B2 (⫻ / \ /) by its distinct verse distribution from the normal counterpart, type B1 (⫻ / ⫻ /), the extended use of relatively heavy language materials 16

As pointed out in section 2.2.1 above, the poet’s reluctance to use -S# in the first drop of type A1 may be viewed as his strategy for preventing the reorganisation from becoming too disturbing to tradition.

MOSP-02.qxd

5/11/04

2:38 PM

104

Page 104

2. Metrical types and positions

for the second position only led to the creation of heavier tokens of the normal type without acquiring an independent status of a metrical type and therewith manifesting a correspondingly distinct verse distribution. In summary, the variety of medial and final syllables that may realise the second drop of type B1 is characterised in terms of the following two parameters: (i) presence versus absence of lexical stress, and (ii) presence versus absence of a word-final unstressed syllable standing after the lexical foot identified by the first parameter; put another way, the opposition between the lexical foot standing alone and the one followed by a word-final syllable. The second parameter is thus subordinate to the first in scope: it is limited in effect to the subset that is marked by the first parameter. Accordingly, a sequence of a lexical-stressed long syllable and a following syllable (-Sx#) is categorically ruled out as a possible realisation of the second drop of type B1; by contrast, the minimal stress bearing units standing alone (-S# ot -sx#) may realise the position in question if only marginally. The computation of parametric values for the group of variants of type B1 is represented in Table 2.71. The entire group falls into three subgroups according to the distinct values assigned. This three-way distinction largely corresponds to the overall distribution pattern of these type B1 variants, recapitulated in Table 2.72. The strings -S# and -sx#, which count as marked by virtue of the presence of lexical stress, are distinguished from their minimally distinct sequences -X# and -xx#, respectively, by the significantly low incidence in occurrence and also by the relatively high frequency of the a-verse with double alliteration. Assigned the least optimal value as a realisation of type B1, the sequence -Sx# is characterised by the marked decrease in the incidence of the b-verse, and correspondingly deviates maximally from the prototypical distribution pattern of type B1. Correspondingly, the string -Sx# is rated as an optimal concatenation of heavy and normal drops, and the whole configuration x . . . PSx#(x)P consequently falls under a separate metrical type, type E with anacrusis. By virtue of the above parameter-based variation, the word-internal constituents are thus involved in gradation in terms of their metrical acceptability as realisations of the second drop of type B1. Table 2.71. Parameter-based evaluation of various posttonic forms as candidates for realising the second drop of type B1 configuration

lexical stress

-x# after a lexical foot

-x# -X# -xx# -Xx#

total value 0 0 0 0

-S# -sx#

⫹ ⫹

-Sx#

⫹

Key: ‘⫹’ ⫽ identified as marked

1 1 ⫹

2

MOSP-02.qxd

5/11/04

2:38 PM

Page 105

2.4. Type B1

105

Table 2.72. Distribution of type B1 variants according to the association of the second drop with the variety of word-internal constituents variant

a-verse with double alliteration

x . . . Px#P 138 (8%) x . . . PX#P 3 (3%) x . . . Pxx#P 8 (5%) x . . . Pxx#P 5 (12%) x . . . PXx#P 6 (12%) x . . . PS#P 15 (21%) x . . . Psx#P 4 (21%) x . . . PSx#P

a-verse with single alliteration

b-verse

275 (16%) 6 (6%) 35 (22%) 3 (7%) 7 (13%) 9 (13%) 1 (5%)

1327 (76%) 92 (91%) 113 (72%) 34 (81%) 39 (75%) 47 (66%) 14 (74%)

8 (26%)

13 (42%)

10 (32%)

The foregoing discussion explains why the configuration x . . . PXx#(x)P in contrast to x . . . PSx#(x)P may be used as a type B1 variant. We are then left with the question why it cannot be exclusively identified as type B1 as it may also be scanned as type E with anacrusis. This question will be postponed until we come to address the metrical status of anacrusis in the Heliand in section 2.11 below. 2.4.2.4. The configurations x . . . P#x#P, x . . . P#xx#P, x . . . P#x#x#P, x . . . P#x#xP, and x . . . Px#xx#P Having determined the varying ways the second drop of type B1 is associated with word internal constituents, we move on to consider how the position is filled by a syllable or a syllable sequence that is external to the preceding lift in terms of word constituency. Let us begin with the simplest case in which the second drop of type B1 is realised by an independent (for the most part, non-lexical) word as exemplified below (cf. Hofmann 1991, II: 94–8): (48) 232a legda im êna bôc an barm 3269a that thu man ni slah 117b endi thîn uuord sô self Table 2.73. Distribution of x . . . P#x#P a-verse with double alliteration

a-verse with single alliteration

b-verse

9 (10%)

7 (8%)

74 (82%)

a-verse with double alliteration: 232a, 1244a, 1332a, 1881a, 2034a, 2116a, 2709a, 4054a, 4781a; a-verse with single alliteration: 1395a, 1497a, 3269a, 3813a, 4561a, 4639a, 4735a; b-verse: 78b, 117b, 142b, 145b, 315b, 459b, 734b, 774b, 925b, 959b, 1007b, 1180b, 1225b, 1397b, 1428b, 1459b, 1512b, 1523b, 1570b, 1620b, 1638b, 1740b, 1809b, 1890b, 2016b, 2026b, 2127b, 2150b, 2254b, 2397b, 2631b, 2739b, 2781b, 2876b, 3053b, 3197b, 3227b, 3271b, 3436b, 3473b, 3548b, 3725b, 3753b, 3754b, 3757b, 3772b, 3953b, 3988b, 3999b, 4058b, 4087b, 4120b, 4298b, 4346b, 4374b, 4381b, 4444b, 4514b, 4605b, 4711b, 4764b, 4777b, 4793b, 4863b, 4881b, 5034b, 5078b, 5484b, 5525b, 5609b, 5647b, 5655b, 5695b, 5769b

MOSP-02.qxd

5/11/04

106

2:38 PM

Page 106

2. Metrical types and positions

Some of the above verses, however, might be read as type C: verses 3197b he uuas is an is hugi iu than and 5695b uuârun im quica noh than, for example, are scanned as type C by Hofmann (1991, II: 150, 153). Further candidates of type C would be verses 4458a that nu obar tuâ naht sind and 1373b Sô uuirðid them, the that godes uuord scal, although these are scanned as type B by Hofmann. Yet verses 4458a and 1373b would seem more amenable to the scansion as type C than would verses 3197b and 5695b and the like, because the words appearing in penultimate position (naht, uuord) are members of class 1 (section 3.2.2), which are most likely to serve as a lift. However, we have no definite instance of the type C a-verse with double alliteration that has its final position mapped to an independent word, though there is a single instance that has the second element of a compound as an occupant of that position (2062a is thit folc frômôd; see section 2.6.3 below). It might appear feasible to assume then that type C does not allow its final position to be occupied by an independent word, and accordingly the two verses at issue might possibly be identified as type B1. As will be discussed further in section 2.6.3 below, however, verses 4458a and 1373b are scanned as type C with greater plausibility. To be examined next is a variety of disyllabic sequences that begin with an independent function word. More specifically, we are concerned with the verses that are of the configurations x . . . P#xx#P (or x . . . P#Px#P), x . . . P#x#x#P (or x . . . P#P#x#P), and x . . . P#x#xP (or x . . . P#P#xP). These verses would appear ambiguous, as they would be scannable as type B1 (x . . . P#xx#P; x . . . P#x#x#P; x . . . P#x#xP) as well as subtype D2b with anacrusis (x . . . P#Px#P; x . . . P#P#x#P; x . . . P#P#xP), depending on how the function words involved (as underlined in the following exemplification) are scanned, as the second drop (type B1) or the second lift (subtype D2b). Hofmann (1991, II: 194–5) scans most of these verses as equivalent to subtype D2b, rather than type B1. (49) 4342a that than is sân after thiu (x . . . P#xx#P or x . . . P#Px#P) 2202b Thie rinc up asat (x . . . P#x#xP or x . . . P#P#xP) 5633b Thuo hreop up te gode (x . . . P#x#x#P or x . . . P#P#x#P) Particularly noteworthy about the above verses are the following two features. First, as indicated in Tables 2.73 and 2.74, with an outstanding concentration in the b-verse, the three configurations as a whole show a distribution broadly similar to that of the configuration x . . . P#x#P. Such a parallelism in distribution may lend plausibility to identifying these configurations as type B1, that is, x . . . P#xx#P, x . . . P#x#xP, Table 2.74. Distribution of x . . . P#xx#P (or x . . . P#Px#P), x . . . P#x#xP (or x . . . P#P#xP), and x . . . P#x#x#P (or x . . . P#P#x#P) a-verse with double alliteration

a-verse with single alliteration

b-verse

0 (0%)

8 (17%)

39 (83%)

a-verse with single alliteration: 222a, 243a, 630a, 2760a, 3894a, 4342a, 4758a, 5898a; b-verse: 43b, 298b, 435b, 449b, 715b, 800b, 995b, 1108b, 1634b, 1758b, 1763b, 1798b, 2010b, 2054b, 2202b, 2219b, 2260b, 2265b, 2667b, 2673b, 2852b, 2872b, 2947b, 3164b, 4545b, 4891b, 4970b, 5041b, 5106b, 5137b, 5146b, 5155b, 5172b, 5354b, 5633b, 5659b, 5867b, 5907b, 5954b

MOSP-02.qxd

5/11/04

2:38 PM

Page 107

2.4. Type B1

107

and x . . . P#x#x#P, respectively, the scansion that will be substantiated as we proceed. Second, regarding the words that appear immediately after the first lift, none of them are nouns or adjectives: all the words involved belong to the classes comprising adverbs, prepositions, finite verbs, and the like, the word classes (classes 2 and 3; section 3.2.2) that are hardly capable of constituting a lift on their own, in distinction from the class of nouns and adjectives (class 1). To be contrasted with the above three configurations are subtype D2b verses of the configurations P#Px#(x)P and P#P#x(x)P, as follows: (50) 626a liof landes uuard 4778a uuacon êne tîd 1135b godes êgan barn For a full list of examples, see under Table 2.80, section 2.7 below. Of particular interest here are the following two properties of these subtype D2b verses: first, a vast majority contain a noun or adjective as a realisation of the versemedial P# or Px# (for a small group of notable exceptions, see below); second, more than half of the verses at issue occur in the a-verse with double alliteration, which is twice as frequent as the b-verse (for the exact distribution pattern, see Table 2.80, section 2.7 below). Given the large discrepancies in distribution as well as in the kinds of lexical items used, it seems scarcely justifiable to identify the verses given in (49) and under Table 2.74 with the subtype D2b configurations P#Px#(x)P and P#P#x(x)P; rather, they should be scanned as type B1 in all probability. We have yet to pre-empt a possible argument for scanning the verses under consideration as subtype D2b, however. Nearly two-thirds of these verses given under Table 2.74, twenty-nine out of a total of forty-seven to be more exact, end in the phrase aftar/after thiu/thiu, as with 4342a that than is sân after thiu and 1763b endi sculun is uuerc aftar thiu. Precisely the same expression often appears at the end of type B1 (with alliteration on af-) on the one hand, and of a minority of the subtype D2b verses referred to above (without alliteration on af-) on the other, as exemplified below: (51) 2755a that ik thi than aftar thiu (type B1; see also 3073a, 3208a, 304b, 1709b, 1796b, 2395b, 2425b, 2567b, 2994b, 3186b, 3195b, 3287b, 3325b, 4613b) 1596a sân aftar thiu (subtype D2b; see also 2100a, 3108a, 196b, 2632b) The vast majority of the type B1 verses in question occur in the b-verse; and no instances of nouns and adjectives are found preceding the phrase aftar thiu. Even a few verbs like bigan ‘begin’, geng ‘go’ that appear before the phrase are so general in content that they may be viewed as closest to function words. Given the treatment of aftar as the first lift of type B1 on the one hand, and that of the same word as the second lift of subtype D2b on the other, one might be led to claim that the ambiguous verses at issue given in (49) and under Table 2.74 should be identified as subtype D2b with anacrusis, with the second lift placed on non-alliterating af- or comparable word-initial syllables: verse 5907b uuarð im sân after thiu would

MOSP-02.qxd

5/11/04

108

2:38 PM

Page 108

2. Metrical types and positions

accordingly be scanned as xxP#Px#P (subtype D2b with anacrusis; cf. Hofmann 1991: 84–5), rather than xxP#xxP (type B1). The suggested scansion as subtype D2b with anacrusis, however, proves to be suspicious on distributional grounds. The anacrustic verses of the group of subtype D2b verses given in (50) on the one hand and the alleged subtype D2b with anacrusis listed under Table 2.74 on the other diverge in their distribution so sharply that a nearly complementary pattern may obtain: while the demonstrable subtype D2b with anacrusis is rare in the b-verse with a single example out of a total of thirteen (c. 8 per cent), the ones given under Table 2.74 are predominant in the b-verse with the relative frequency of c. 83 per cent, as indicated in the table. To subsume under a single type these two configurations with diametrically opposed distribution patterns would therefore seem extremely doubtful. Furthermore, if the verses at issue were scanned as subtype D2b with anacrusis, the resultant whole distribution of the variant subtype D2b P#Px . . . P would be polarised depending on presence verse absence of anacrusis: without anacrusis, the majority is in the a-verse with double alliteration, but with anacrusis, in the b-verse. As substantiated in section 2.11.1 below, however, anacrusis is hardly capable of affecting distribution pattern, and such a polarity in distribution would be unlikely in the extreme. Thus, the scansion of the verses given in (49) above as subtype D2b would create the question of why this subtype has to be used with such overwhelming frequency in the b-verse when a word other than a noun or an adjective comes after the first lift and at the same time anacrusis occurs, in contrast to the majority of verses given in (50) above. Of further significance, the remaining variant of subtype D2b, namely, the one that is structurally unambiguous with its second drop occupied by the second element of a compound (that is, P#Px . . . S; e.g., 5899a erl ellanruof ), occurs nearly exclusively in the a-verse with double alliteration (see section 2.7 below). Moreover, anacrusis is infrequent in this variant: only two verses, 197a that uuîf uurdigiscapu and 3882a theru idis aldarlago, are provided with anacrusis; and if anacrusis occurs, anacrustic syllables tend to be minimal: monosyllabic (197a), disyllabic (3882a). Thus, one would have to explain why the alleged anacrustic variants of subtype D2b x . . . P#Px#(x)P and x . . . P#P#x(x)P lack all typical properties of subtype D2b in terms of composition and distribution. By contrast, scanning these verses as type B1 would not seem to run into the difficulties that confront their scansion as subtype D2b. We may provide a principled account for the absence of x . . . P#Px#(x)P and x . . . P#P#x(x)P, where #Px# and #P# in medial position correspond to a noun or an adjective: because type B1 does not allow the second drop to be filled by a string more prominent than a single foot with a secondary stress (-S# or -sx#; section 2.4.2.3), it may naturally follow that these sequences are disallowed categorically, much as is the sequence x . . . PSx#P treated in the preceding section. At the same time, the predominance of the b-verse, as well as the invariable presence of a long sequence of unstressed syllables in verseinitial position in the configurations x . . . P#xx#P, x . . . P#x#xP, and x . . . P#x#x#P, receive explanation: type B1 has strong preference for the b-verse, and requires a sequence of unstressed syllables at the beginning as one of its constituent positions, a notably long sequence for the b-verse in particular (section 2.4.1). Of related interest is the configuration x . . . Px#xx(x)#P (or x . . . Px#Px(x)#P), which may be scannable as types B1, D* (subtype D*2b, with anacrusis; section 2.8),

MOSP-02.qxd

5/11/04

2:38 PM

Page 109

2.4. Type B1

109

and E* (with anacrusis; section 2.10) alike, as exemplified below (cf. Kauffmann 1887: 324; Martin 1907: 53–4): (52) 1602a 502b 512b 662b 874b 890b 1061b 1552b 2183b 2345b 3220b 3230b 3386b 3646b 4009b 4928b 5376b

geuuîhid sî thîn namo Thiu thiorna al forstôd Thô uuas siu uuidouua aftar thiu The sterro liohto skên endi them uueroda allan dag That is hêrro obar al the fîund nâhor geng endi antfâhis eft than thu uuili Thiu môder aftar geng thea he cûðde obar al ac uuese imu mildi an is hugi Ef imu than is sundea aftar thiu it is hêr sô thikki undar ûs that im mildi aftar thiu than uuirðit iuuua gilôbo after thiu Thie fîund eft geuuitun thie habit hier rîki ober ûs

On the interpretation that the configurations x . . . P#xx#P and x . . . P#x#x(#)P (where #xx# and #x# are non-nominal) constitute type B1 as submitted above, we would be led to conclude by following a similar line of reasoning that the sequence under consideration constitutes type B1 as well, largely in accord with Kauffmann (1887: 324, 340, 346), and in disagreement with Martin (1907: 53) and partially so with Hofmann (1991, II: 196–7).17 All the above words used as the second drop of the configuration x . . . Px#xx(x)#P belong to the word classes that are distinguished by their weaker propensity to form a lift than the class of nouns and adjectives (section 3.2.2). To scan the verses listed in (52) as types D* or E* would have to leave this significant fact unexplained.18 At this point, it will be instructive to examine the situation in Beowulf: does the poem attest type B1 variants with the second drop occupied by an independent word, analogous to those found in the Heliand? Beowulf offers several examples of seemingly comparable compositions as follows: (53) Beo 80a Beo 525a Beo 1696a Beo 1763a Beo 1864a Beo 2870a 17

18

He beot ne aleh (x . . . P#x#xP) Ðonne wene ic to þe (x . . . Px#x#x#P) geseted ond gesæ¯d (xpx#x#xP) þæt þec adl oððe ecg (x . . . P#xx#P) ge wið feond ge wið freond (x . . . P#x#x#P) ower feor oððe neah (x . . . P#xx#P)

Hofmann would appear less than consistent in this respect. While scanning the majority of cases as subtype D*2b (in our term), he subsumes several similar examples (874b, 890b, 2345b, 3220b, 5376b) under type B1 (Hofmann 1991, II: 129–38). On analogous grounds, verse 1607a Gef ûs dago gehuuilikes râd, which could be scanned as type B1, type D* (subtype D*2b), or type E*, should better be scanned as type B1.

MOSP-02.qxd

5/11/04

2:38 PM

110

Page 110

2. Metrical types and positions Beo 272b Beo 455b Beo 469b Beo 1182b Beo 1616b Beo 1876b Beo 2574b

Þu wast, gif hit is (x . . . P#x#x#P) Gæ¯ð a wyrd swa hio scel (x . . . P#x#x#P) se wæs betera ðonne ic (x . . . Px#xx#P) gyf þu æ¯r þonne he (x . . . P#xx#P) wæs þæt blod to þæs hat (x . . . P#x#x#P) Wæs him se man to þon leof (x . . . P#x#x#P) swa him wyrd ne gescraf (x . . . P#x#xP)

As it turns out, however, the similarity is more apparent than real. The words involved are all functional or non-lexical, i.e., pronouns, conjunctions, and prepositions, belonging as they do to the class of words that are generally incapable of forming a lift in non-final position (class 3; section 3.2.2), as opposed to the class comprising adverbs and finite verbs (class 2). Specifically, there are no examples attested in which the words at issue serve as the first lift of type B1. The above verses therefore should better be compared to the following examples from the Heliand, which are unambiguously type B1 verses: (54) 591a 1522a 2547a 4898a 1746b 5542b

undartuisc erða endi himil (x . . . Px#xx#P) queðe iâ, gef it sî (x . . . P#x#x#P) ge that corn ge that crûd (x . . . P#x#x#P) he suiltit imu eft (x . . . Px#xx#P) nec it ôc god ni gescôp (x . . . P#x#xP) huand sia ni uuitun, huat sia duot, quathie (x . . . P#x#x#P(xx))19

Accordingly, it would seem most plausible to characterise the words and word groups under consideration as constituting drops and thus to analyse the verses concerned as type B1. Furthermore, in Beowulf occurrence of type D2b/D*2b is disallowed in the b-verse, as is anacrusis (Suzuki 1996a). Consequently, we have no choice but to scan the verses under discussion as type B1. In this way, Beowulf permits only words of the lowest ranking class (class 3) to appear in the second drop of type B1, while the Heliand allows the class next higher up in ranking (class 2) to be used as part of the second drop as well. The highest ranking class (class 1), however, is generally disallowed from this position. The foregoing consideration thus shows that the apparently ambiguous configurations under consideration in this section reduce to the configurations x . . . P#xxP, x . . . Px#xxP, and the like by constituting type B1 variants in the ways highly characteristic of the Heliand. To conclude, the Heliand expanded the variety of type B1 by making available a wider choice of language materials as possible realisations of its second drop. As regards word internal constituents, the Heliand increasingly allowed the sequence -Xx#, and more moderately -S# and -sx# as well, to occupy the second drop of type B1, whereas in the traditional metre these sequences occurred merely as isolated exceptions. With respect to use of independent words, the Heliand permitted the intermediate word class (class 2) to serve as an exponent of the position in question by relaxing the traditional restriction to the lowest ranking class (class 3). In a sense, 19

The phrase quathie is extrametrical, as noted in section 1.4 above.

MOSP-02.qxd

5/11/04

2:38 PM

Page 111

2.5. Type B3

111

such relaxations brought about an obscuration of the phonological- and lexicalbased, conventional categorical distinction between lift-forming and drop-forming classes: the Heliand accepted a subset of the lift-forming words and their constituents as qualified to form a drop as well. Despite the drastic transformation on the surface, this metrical change constituted only a slight shifting of the organising principle in the system. In the traditional metre, the second drop of type B1 was limited to unfooted language materials, that is, those lacking stress regardless of its origin (lexical or derived). What ascended to figure as central to the new organisation was the presence versus absence of lexical stress that appears prototypically in the most salient word categories, that is, nouns and adjectives. This parameter, however, was not constructed anew by the Heliand poet. Rather it had presumably been in existence in the traditional versecraft, as had the second parameter, the presence of an additional unstressed syllable in word-final position (-x#): these two parameters would have been jointly responsible for differentiating the string -Sx# from -S# and -Xx# as a totally unacceptable realisation of the second drop of type B1. With the reintroduction of feet in word-final position in Old Saxon (section 1.3.1), the foot was no more capable of functioning as a prime organiser for the position under consideration than for the first drop of type A1 (section 2.1.7.2). Thereupon the other two parameters (lexical stress and -x#), which remained undisturbed by this phonological change, came to figure more centrally as metrical organisers in the Heliand metre. Consequently, the word internal sequence -Xx# lacking lexical stress was increasingly allowed to appear in the second drop of type B1, while the strings -S# and -sx#, still marginal, became more acceptable and common as variants of the second drop of type B1. By contrast, the sequence -Sx# continued to be excluded categorically from the position at issue in the Heliand metre as well. Thus, the Heliand poet reorganised the language materials that had traditionally been ruled out as highly exceptional or totally unmetrical by making them graded in terms of their metricality. Intrigued by linguistic change, the poet achieved such a reconfiguration of traditional distinctions by minimally manipulating the inherited underlying generalisations.

2.5. Type B3 (⫻ ⫻ ⫻ /) There are seven verses in which the sole alliterative element comes at the end of a verse in the form of P or px, as follows (cf. Kauffmann 1887: 324): (55) 573a 1533a 2367a 2756a 3693a 4038a 5745a

huuand im habde forliuuan that gi sô ni uurecan huô sie scoldin gehalon sô hues sô thu mi bidis huô thu noh uuirðis behabd that he it thi sân fargibid huô sia eft te them grabe

Notable about these verses is that all but one example end in a resolvable disyllable px.

MOSP-02.qxd

5/11/04

112

2:38 PM

Page 112

2. Metrical types and positions

In the absence of a further alliterative element located nearer the beginning of a verse, and in the absence of a further lexical word qualified for alliteration by nature (see section 3.2.2 below), we have to assume that these verses contain only a single lift. Therefore, the whole configuration should be characterised as ⫻ ⫻ ⫻ /. It would then seem only natural to regard this configuration as a reduced type of class B that has only a single lift, namely type B3 (Sievers 1893: 157). Put another way, this configuration may be derived by substituting the otherwise expected first lift with a drop in a way comparable to what is responsible for the derivation of type A3 from type A1. Alternative scansions might suggest themselves, however: the verses in question might be scanned as subtype A3s, as type C3, or as type E3, as is proposed by Hofmann (1991: 65), and also by Suzuki (1996a: 133) for comparable verses in Beowulf. We should now consider these possible candidates in turn. Least plausible of all would seem scanning as subtype A3s, which would constitute a short variant of type A3, much as does subtype A1s (typically PS#px; section 2.1.8) as against type A1. While in keeping with the vast majority of examples ending in the resolvable disyllable px, this alternative would fail to account for verse 3693a ending in a long syllable P rather than a resolvable disyllable px. Moreover, lacking a demonstrably stressed word in the immediately preceding position, even the remaining verse-final words of the form px are unqualified for suspension of resolution, and the verses in question therefore cannot be identified as the short variant of type A3. Slightly better than subtype A3s would appear type C3 as a candidate, which would be derived by replacing the first lift with a drop, the whole configuration being accordingly ⫻ ⫻ / ⫻. This variant would be compatible with the presence of px at the end of a verse. We may note here in advance that type C is often subject to suspension of resolution on its second lift (section 3.1.1.2). It might thus be tempting to derive the candidate from this variant with the unresolved second lift (x . . . P#px). However, the existence of verse 3693a ending in P poses difficulty, as with the first alternative scansion. No less problematic would be found the scansion as type E3. Derived by replacing the first lift with a drop, this variant would continue to have a heavy drop in the second position (⫻ \ ⫻ /). Yet there seems to be no word that requires a lexical stress (a prototypical occupant of a heavy drop) in the portion of the verse before the alliterative word. Of course, one might be free to give prominence to one of these constituent words presumably as a rhythmical consequence, but such a practice, largely contextdependent and even arbitrary at times, would be a matter of performance that goes beyond the strictly metrical concerns. To be added to the above list are verses 235a and 3962a as follows: (56) 235a Thô nam he thia bôk an hand 3962a an is hêlagun uuord Unlike the seven examples considered earlier, these two verses have two content words each (bôk and hand, hêlagun and uuord, respectively) that are qualified for forming a lift. There is no question accordingly that these verses should be scanned as instances of type B1. Yet the verses manifest alliteration on the second lift alone, namely the alliterative pattern xa, rather than aa or ax as would otherwise be expected. As will be discussed at length in section 3.2.2 below, given two substantives the

MOSP-02.qxd

5/11/04

2:38 PM

Page 113

2.6. Type C

113

first one takes precedence over the second in realising alliteration (the principle of left dominance; cf. Sievers 1893: §23; Suzuki 1996a: 282). The verses under consideration therefore have to be regarded as violating this generalisation: by treating the first noun bôk and adjective hêlagun as drops, these verses constitute examples of type B3; in the presence of a single lift available to type B3, then, there would be no other way than to select the verse-final nouns hand and uuord as the sole alliterative elements. The existence of these exceptional verses may in turn be adduced as corroborating evidence for the scansion of the seven verses given earlier as type B3. Furthermore, the nearly exclusive appearance of resolution on the verse-final lift may be understood as a consequence of avoiding exclusive alliteration at the end of a verse (section 3.2.1.1), rather than a defining property inherent in the underlying verse form. All in all, then, the seven exceptional verses, alongside verses 235a and 3962a, should best be scanned as a reduced type of class B, namely type B3, which arises through substitution of the first lift with a drop.

2.6. Type C (⫻ / / ⫻) 2.6.1. The variation on the first drop of type C As with type B1, the first drop of type C is realised by a varying number of unstressed syllables ranging between one and eleven, as shown in Tables 2.75 and 2.76 (cf. Kauffmann 1887: 325–33; Hofmann 1991, II: 143–73). Also in parallel to type B1, the b-verse tends to be associated with a larger number of syllables for the first drop than the a-verse: while the range from one to four syllables accounts for about 90 per cent of all occurrences for the a-verse, with double and single alliteration alike, the same range amounts to less than 70 per cent for the b-verse (Table 2.75); and for each of the higher numbers of syllables involved, the b-verse constitutes a vast majority, accounting for more than 80 per cent of all instances (Table 2.76). Yet this contrast between the a-verse and the b-verse seems somewhat less pronounced in comparison with type B1 (see Tables 2.46 and 2.47 above), particularly because the proportion of the verses with a single syllable in initial position to those with two or more syllables is scarcely distinguishable among the three verse categories (Table 2.75), and because even for such minimal verses the b-verse accounts for well over the half of all instances (Table 2.76). Equally less outstanding is the contrast between the a-verse with single alliteration and the one with double alliteration: the use of a single syllable does not show a marked preference for the a-verse with double alliteration in type C. Of further interest among the type C verses having a single syllable in the first drop is the minimally prominent variant that begins with a prefix, as exemplified below (Hofmann 1991: 140): (57) 189a forsehen selbo 1394a biholan uuerðan 3119b gicoren habde

MOSP-02.qxd

5/11/04

2:38 PM

114

Page 114

2. Metrical types and positions

Table 2.75. Distribution of the varying number of syllables in the first drop of type C for the a-verse with double alliteration, the a-verse with single alliteration, and the b-verse number of x 1 2 3 4 5 6 7 8 9 10 11

a-verse with double alliteration 15 34 26 19 5 2 1 1 0 0 0

total

15% 33% 25% 18% 5% 2% 1% 1% 0% 0% 0%

103 100%

a-verse with single alliteration 48 158 141 65 26 16 6 5 0 0 0

10% 34% 30% 14% 6% 3% 1% 1% 0% 0% 0%

465 100%

b-verse 124 10% 258 21% 248 20% 219 18% 169 14% 97 8% 60 5% 24 2% 11 1% 10 1% 2 less than 1% 1222 100%

Table 2.76. Verse distribution according to the number of syllables in the first drop of type C number of x

a-verse with double alliteration

a-verse with single alliteration

b-verse

total

1 2 3 4 5 6 7 8 9 10 11

15 34 26 19 5 2 1 1 0 0 0

48 158 141 65 26 16 6 5 0 0 0

124 66% 258 57% 248 60% 219 72% 169 85% 97 84% 60 90% 24 80% 11 100% 10 100% 2 100%

187 450 415 303 200 115 67 30 11 10 2

8% 8% 6% 6% 3% 2% 1% 3% 0% 0% 0%

26% 35% 34% 21% 13% 14% 9% 17% 0% 0% 0%

100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%

a-verse with double alliteration: 189a, 1913a, 5796a, 5857a; a-verse with single alliteration: 1394a, 3199a, 3342a, 3350a, 3979a; b-verse: 105b, 132b, 245b, 375b, 423b, 426b, 452b, 457b, 472b, 1101b, 1297b, 1396b, 1839b, 1983b, 2337b, 2642b, 3119b, 3158b, 3259b, 3432b, 3470b, 3577b, 3652b, 3662b, 3736b, 3826b, 4157b, 5157b, 5272b These verses are noteworthy in several respects. First, none of the examples have their second lift suspended from resolution (cf. Hirt 1891: 153): thus, we find no

MOSP-02.qxd

5/11/04

2:38 PM

Page 115

2.6. Type C

115

instances of xP#px in contrast to xP#Px, although suspension of resolution is otherwise common for type C (section 3.1.1.2). Second, as pointed out by Hofmann (1991: 140–1), the vast majority (thirty-seven out of thirty-eight, c. 97 per cent) have their first lift resolved, with an apparent single exception being verse 5857a gisald selbo.20 In the light of the correlation substantiated in section 3.1.1.6 below, the predominant occurrence of resolution on the first lift would seem compatible with the avoidance of suspension of resolution from the second lift, yet we should explore a fuller account that explains the complete absence of the unresolved second lift. A credible account seems forthcoming by assuming that there is a minimum amount of prominence required of the verse, and also that there should be a balance (or a trading relation) between different positions in terms of their prominence realised. Given the minimal amount of prominence accorded to the first drop that is realised by a monosyllabic prefix, the unresolved second lift would make the resultant verse too weak in prominence by virtue of being short in syllable quantity: the configuration xP#px would thus be ruled out as failing to satisfy the minimum amount of prominence required for a well-formed verse. Furthermore, in favour of the well-balanced arrangement of prominence in metrical positions the minimum prominence falling on the first drop occupied by a prefix should be compensated for by a maximal prominence on the first lift, which is obtained by resolution, as shown below; thus, the configuration xpx#Px is chosen at the expense of xP#Px. Third, unless in combination with other unstressed material, a prefix never appears in the variants of type C the second lift of which is occupied by language material other than an independent word. Put another way, the configurations xPSx, xPsx, xPXx, and xPxx never begin with a prefix alone as an occupant of the first drop. Yet there seems no reason why these configurations should not be accompanied with a prefix alone; more specifically, we would predict the following words to constitute verses on their own: gisunfader ‘son and father’; gisîðskepi ‘companionship’; gilîknessi ‘image’. Particularly puzzling seems the absence of the configuration xPXx, given that verb forms, notably present participles, may be realised as the sequence #PXx#. We should accordingly expect that there are verses which are constituted by such a verb form preceded by a prefix alone, like *gifulleanne (cf. 976a te gifulleanne). Such verses, however, are non-existent. Of related interest in this connection would be the notable constraint on the occurrence of the construction te ⫹ gerund when used as a type C (Hofmann’s type dax3a) verse on its own, such as 1023a te giuuinnanne (Hofmann 1991: 143–4): all seventeen examples as Hofmann (1991: 143) lists, namely, verses 976a, 1023a, 1188a, 1467a, 2329a, 2377a, 2433a, 2531a, 3803a, 3903a, 4027a, 4055a, 4291a, 4512a, 4687a, 5531a, 5830a, have a gerund expanded with a verbal prefix; in other words, te ⫹ gerund without a prefix, such as *te uuinnanne, seems to be disallowed as an independent type C verse. By contrast, the same preposition is found capable 20

There is reason to suspect, however, that this verse would not have constituted an exception to the generalisation. According to Kuhn (1929: 50) and Hofmann (1991: 141), the original form of this seemingly exceptional verse would have been *giselid selbo, with the disyllabic variant -selid of the past participle of -sellian ‘give’ in the place of the monosyllabic counterpart -sald. The postulation of such a disyllabic alternate form for Old Saxon is made plausible by the existence of the analogous pair in Old High German, giselit/gisalt.

MOSP-02.qxd

5/11/04

116

2:38 PM

Page 116

2. Metrical types and positions

of constituting the first drop of type C on its own, as in 1867b te hebenrîkea. Following Hofmann’s (1991: 143) insight, the non-occurrence of te ⫹ gerund without a prefix as a verse in its own right may as well be ascribed to the minimal prominence falling on te, in parallel to prefixes as observed in the last paragraph: the preposition te, functioning here as a gerund marker, is treated as scarcely different from a verbal prefix in its morpholexical capacity, and therefore counted as minimally prominent for metrical purposes. We are then confronted with the following question: why are the configurations xPSx, xPsx, xPXx, and xPxx (where the initial x is a prefix) excluded as realisations of type C in contrast to x . . . PSx, x . . . Psx, x . . . PXx, and x . . . Pxx? Building on the account adduced earlier, we may assume that the hypothetical forms in question weigh too light to constitute a verse. Since even the primary-stressed short syllable (p-) falls short of the amount of prominence due in the context at issue, the syllables with lesser degrees of stress (-S-, -s-, -X-, -x-) are naturally excluded as no more qualified. By contrast, the presence of a primary stress on the long syllable in the second lift, as in the configurations xP#Px and xpx#Px, may well compensate for the minimal weight in the first drop by virtue of the relatively great prominence on it. Yet on another score, the configuration xP#Px is less than optimal in comparison with the configuration xpx#Px (with resolution on the first lift). As argued at length in section 3.1.3 below, type C suffers a relative weakening of prominence due to the contiguity of lifts: by virtue of the immediately following second lift of more or less equivalent prominence, the first counterpart tends to be perceived as less prominent than when appearing elsewhere. In order to make its inherent prominence fully perceivable, the first lift needs an extra prominence. This needed means of enhancing prominence is provided by resolution, which adds strength to the position concerned by making an additional syllable available for association: by resolution, the disyllable px instead of the monosyllable P comes to realise the position involved. Overall, then, the configuration xpx#Px is rendered capable of realising a sufficient amount of prominence that enables it to serve as a verse on its own despite the minimal prominence on the first drop. A further feature of this variety of verse needs to be addressed. As we have seen in section 2.4 above, when the first drop of type B1 is occupied by a prefix, the a-verse with double alliteration accounts for the majority of relevant instances (c. 71 per cent). Type C, however, fails to show a comparable feature: only four out of thirty eight examples (c. 11 per cent; 189a, 1913a, 5796a, 5857a) are a-verses with double alliteration. The lack of preference for the a-verse with double alliteration on the part of the variant of type C at issue may be attributed to the lesser degree of dependence of type C in general for its verse distribution patterning on the varying number of syllables in the first drop. More specifically, for type C, in contrast to type B1, the appearance of a single syllable (of whatever morpholexical status) in the first drop does not constitute a particularly favourable condition for occurrence of the a-verse with double alliteration, as observed at the beginning of this section. Given the weaker sensitivity to the varying size of the initial drop, then, it is hardly surprising that type C does not show preference for the a-verse with double alliteration even when the position in question is realised by a prefix, the language material of minimal prominence available.

MOSP-02.qxd

5/11/04

2:38 PM

Page 117

2.6. Type C

117

There is a further indication that type C is less sensitive to the varying size of its first drop than type B1. At issue here is the relatively higher frequency with which the minimally prominent material, namely a prefix, may constitute the first drop of type C on its own: such minimal verses account for c. 3 per cent of all occurrences of type C verse, whereas the corresponding figure drops sharply to c. 0.3 per cent for type B1. The greater tolerance of type C for the minimal prominence may again be understood as part of a more general property. Type C exhibits higher incidence of the variant having the first drop realised by a single syllable than does type B1: such verses account for approximately 15 per cent of all instances of type C (Tables 2.75 and 2.76), and about 6 per cent of type B1 (Tables 2.46 and 2.47). 2.6.2. The variation on the second lift of type C While type C is commonly characterised by the overall distribution pattern shown in Tables 2.75 and 2.76 above, it is subject to gradation in finer points of distribution primarily according to the varying ways the second lift is realised. On the basis of such subtle distinctions, the following six configurations may be distinguished as major variants of type C:21 (i) x . . . P#px (e.g., 313a suîðo gôd gumo); (ii) x . . . P#Px (e.g., 1871a thene uueg uuîsit); (iii) x . . . Psx (e.g., 730a thurh iro handmagen); (iv) x . . . PSx (e.g., 3492a an thene uuîngardon); (v) x . . . Pxx (e.g., 4250a iro drohtines); (vi) x . . . PXx (e.g., 5830a te gihôrianne).22 The differing distribution patterns are shown in Table 2.77. As shown below, the distribution of type C differs considerably from variant to variant, largely conditioned by language materials that occupy the second lift. On the basis of syllable properties, the relevant materials may fall into three classes, each subdivided into two groups. The three major classes are identified by differing degrees of stress, primary, secondary, and weak. Each class thus identifiable is further divided into two subclasses according to the length of the syllables involved, long versus short. We accordingly obtain the following three classes, each divided further into two subclasses: (58) Class 1: (i) x . . . P#Px; (ii) x . . . P#px Class 2: (i) x . . . PSx; (ii) x . . . Psx Class 3: (i) x . . . PXx; (ii) x . . . Pxx 21

22

Hofmann (1991: 146–7) distinguishes between x’1.2 (⫽ x . . . px#Px) and x1.2 (⫽ x . . . P#Px). This proves to be a distinction without a difference, however, because these two classes show similar verse distribution patterns, as follows. For the configuration x . . . px#Px, the a-verse with double alliteration: sixty-two examples (13 per cent); the a-verse with single alliteration: forty-eight examples (10 per cent); the b-verse: 374 examples (77 per cent). For the configuration x . . . P#Px, the a-verse with double alliteration: nineteen examples (10 per cent); the a-verse with single alliteration: twenty-three examples (12 per cent); the b-verse: 150 examples (78 per cent). After all, Hofmann’s tripartite distinction of type C into x . . . P#Px, x . . . px#Px, and x . . . P#px is not motivated on structural grounds, as he himself concedes (Hofmann 1991: 146–7). We may recall here that the configuration x . . . Pxx (type C; e.g., 1229a that sie ûses drohtines) should be categorically distinguished from the configuration x . . . Pxx (type A3; e.g., 1221a thoh sie thar alle be gelîcumu), which is characterised by the association of the verse-final position with a disyllabic ending (-xx#) (see section 2.3 above).

MOSP-02.qxd

5/11/04

2:38 PM

118

Page 118

2. Metrical types and positions

Table 2.77. Distribution of major variants of type C variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

total

x x x x x x

14 3% 81 12% 0 0% 4 2% 4 3% 0 0%

46 9% 71 11% 87 48% 141 56% 76 63% 44 96%

452 523 96 109 40 2

512 675 183 254 120 46

. . . . . .

. . . . . .

. . . . . .

P#px P#Px Psx PSx PXx Pxx

88% 77% 52% 43% 33% 4%

100% 100% 100% 100% 100% 100%

x . . . P#px a-verse with double alliteration: 200a, 313a, 949a, 970a, 3401a, 3821a, 3893a, 4145a, 4454a, 4467a, 5071a, 5132a, 5570a, 5716a; a-verse with single alliteration: 17a, 73a, 94a, 115a, 177a, 227a, 784a, 896a, 1015a, 1079a, 1123a, 1131a, 1278a, 1363a, 1387a, 1391a, 1456a, 1484a, 2079a, 2143a, 2216a, 2303a, 2326a, 2469a, 2537a, 2539a, 2671a, 2731a, 2816a, 2932a, 3266a, 3582a, 3634a, 3678a, 4050a, 4198a, 4324a, 4362a, 4600a, 4628a, 4786a, 5234a, 5599a, 5623a, 5631a, 5770a; b-verse: 2b, 10b, 14b, 23b, 81b, 87b, 95b, 118b, 122b, 128b, 129b, 164b, 168b, 169b, 172b, 180b, 182b, 184b, 192b, 205b, 209b, 216b, 229b, 260b, 273b, 280b, 324b, 331b, 352b, 368b, 382b, 391b, 425b, 431b, 437b, 460b, 466b, 474b, 479b, 487b, 501b, 522b, 531b, 544b, 545b, 565b, 572b, 575b, 584b, 626b, 628b, 637b, 650b, 651b, 689b, 693b, 702b, 707b, 712b, 714b, 721b, 740b, 767b, 770b, 787b, 825b, 827b, 831b, 840b, 852b, 856b, 866b, 886b, 911b, 913b, 915b, 919b, 929b, 932b, 946b, 955b, 999b, 1003b, 1004b, 1064b, 1065b, 1084b, 1086b, 1115b, 1154b, 1164b, 1196b, 1203b, 1260b, 1289b, 1337b, 1354b, 1355b, 1357b, 1359b, 1374b, 1376b, 1381b, 1392b, 1393b, 1407b, 1432b, 1480b, 1495b, 1550b, 1557b, 1563b, 1574b, 1583b, 1623b, 1627b, 1654b, 1675b, 1679b, 1724b, 1753b, 1775b, 1788b, 1797b, 1806b, 1825b, 1845b, 1853b, 1864b, 1869b, 1888b, 1901b, 1919b, 1927b, 1951b, 1968b, 1978b, 1980b, 1982b, 1985b, 1986b, 1996b, 2003b, 2018b, 2023b, 2046b, 2062b, 2066b, 2070b, 2082b, 2085b, 2105b, 2107b, 2158b, 2180b, 2192b, 2195b, 2197b, 2203b, 2225b, 2227b, 2253b, 2264b, 2267b, 2288b, 2289b, 2295b, 2307b, 2309b, 2312b, 2324b, 2334b, 2342b, 2346b, 2369b, 2376b, 2385b, 2407b, 2415b, 2419b, 2435b, 2438b, 2446b, 2469b, 2472b, 2474b, 2476b, 2478b, 2481b, 2485b, 2488b, 2497b, 2500b, 2538b, 2559b, 2564b, 2565b, 2571b, 2574b, 2591b, 2606b, 2633b, 2638b, 2649b, 2650b, 2656b, 2669b, 2675b, 2677b, 2710b, 2750b, 2777b, 2796b, 2807b, 2831b, 2851b, 2875b, 2881b, 2895b, 2913b, 2916b, 2918b, 2928b, 2929b, 2960b, 2962b, 2964b, 2998b, 3025b, 3050b, 3102b, 3125b, 3132b, 3152b, 3161b, 3173b, 3178b, 3214b, 3226b, 3233b, 3237b, 3262b, 3280b, 3284b, 3291b, 3302b, 3304b, 3322b, 3336b, 3340b, 3351b, 3374b, 3396b, 3402b, 3440b, 3481b, 3488b, 3547b, 3552b, 3560b, 3583b, 3642b, 3653b, 3686b, 3707b, 3727b, 3732b, 3740b, 3755b, 3760b, 3778b, 3785b, 3799b, 3804b, 3808b, 3810b, 3819b, 3830b, 3862b, 3878b, 3906b, 3910b, 3928b, 3939b, 3950b, 3951b, 3987b, 3994b, 4007b, 4008b, 4010b, 4032b, 4043b, 4047b, 4063b, 4077b, 4080b, 4081b, 4098b, 4122b, 4123b, 4133b, 4160b, 4161b, 4164b, 4171b, 4181b, 4191b, 4193b, 4194b, 4203b, 4230b, 4252b, 4257b, 4265b, 4274b, 4280b, 4282b, 4287b, 4289b, 4309b, 4317b, 4336b, 4348b, 4359b, 4366b, 4376b, 4379b, 4404b, 4419b, 4432b, 4441b, 4448b, 4463b, 4470b, 4482b, 4485b, 4491b, 4492b, 4503b, 4527b, 4534b, 4535b, 4549b, 4550b, 4578b, 4585b, 4588b, 4608b, 4609b, 4637b, 4665b, 4669b, 4670b, 4671b, 4687b, 4698b, 4723b, 4726b, 4727b, 4730b, 4738b, 4751b, 4770b, 4771b, 4779b, 4801b, 4808b, 4822b, 4838b, 4844b, 4845b, 4853b, 4867b, 4882b, 4883b, 4890b, 4903b, 4907b, 4920b, 4929b, 4933b, 4943b, 4946b, 4947b, 4957b, 4964b, 5015b, 5031b, 5032b, 5042b, 5052b, 5054b, 5055b, 5061b, 5074b, 5095b, 5101b, 5117b, 5122b, 5124b, 5130b, 5138b, 5178b, 5190b, 5203b, 5205b, 5206b, 5238b, 5261b, 5270b, 5283b, 5286b, 5296b, 5313b, 5336b, 5338b, 5342b, 5361b, 5365b, 5378b, 5390b, 5434b, 5464b, 5482b, 5520b, 5523b, 5534b, 5539b, 5540b, 5562b, 5590b, 5608b, 5646b, 5653b, 5669b, 5686b, 5691b, 5729b, 5825b, 5831b, 5840b, 5863b, 5865b, 5868b, 5870b, 5896b, 5908b, 5910b, 5940b, 5971b, 5980b x . . . P#Px a-verse with double alliteration: 104a, 189a, 228a, 293a, 331a, 549a, 577a, 669a, 807a, 945a, 982a, 1017a, 1042a, 1407a, 1477a, 1488a, 1521a, 1797a, 1871a, 1898a, 1913a, 1977a, 1998a, 2019a,

MOSP-02.qxd

5/11/04

2:38 PM

Page 119

2.6. Type C

119

2041a, 2062a, 2154a, 2184a, 2256a, 2402a, 2423a, 2500a, 2595a, 2701a, 2779a, 2850a, 2858a, 2894a, 2897a, 2906a, 2922a, 2925a, 2956a, 3094a, 3135a, 3234a, 3245a, 3295a, 4017a, 4030a, 4041a, 4069a, 4098a, 4184a, 4210a, 4555a, 4740a, 4964a, 5092a, 5162a, 5178a, 5285a, 5293a, 5315a, 5319a, 5371a, 5486a, 5555a, 5628a, 5757a, 5763a, 5783a, 5790a, 5796a, 5814a, 5832a, 5857a, 5871a, 5900a, 5954a, 5960a; a-verse with single alliteration: 14a, 103a, 132a, 272a, 283a, 429a, 444a, 457a, 475a, 528a, 653a, 706a, 806a, 977a, 1168a, 1241a, 1249a, 1394a, 1564a, 1793a, 1800a, 1858a, 1969a, 2237a, 2310a, 2314a, 2383a, 2479a, 2498a, 2509a, 2629a, 2787a, 2905a, 3199a, 3342a, 3350a, 3358a, 3427a, 3475a, 3514a, 3590a, 3632a, 3742a, 3748a, 3778a, 3901a, 3907a, 3941a, 3979a, 4022a, 4146a, 4149a, 4235a, 4372a, 4458a, 4534a, 4678a, 4753a, 4787a, 4809a, 4839a, 4862a, 5181a, 5188a, 5222a, 5442a, 5736a, 5780a, 5846a, 5895a, 5980a; b-verse: 19b, 25b, 41b, 70b, 94b, 105b, 120b, 132b, 163b, 165b, 193b, 208b, 210b, 214b, 215b, 218b, 225b, 232b, 235b, 245b, 251b, 253b, 257b, 258b, 265b, 276b, 283b, 287b, 296b, 299b, 302b, 323b, 327b, 328b, 357b, 375b, 385b, 412b, 423b, 426b, 428b, 442b, 451b, 452b, 457b, 461b, 472b, 473b, 489b, 494b, 499b, 503b, 541b, 576b, 578b, 592b, 596b, 607b, 623b, 636b, 642b, 644b, 654b, 668b, 673b, 674b, 679b, 695b, 713b, 731b, 737b, 752b, 758b, 773b, 777b, 790b, 793b, 794b, 809b, 810b, 817b, 819b, 835b, 841b, 850b, 896b, 916b, 923b, 939b, 948b, 949b, 957b, 968b, 994b, 1015b, 1018b, 1033b, 1039b, 1049b, 1057b, 1062b, 1070b, 1081b, 1101b, 1109b, 1158b, 1166b, 1174b, 1185b, 1197b, 1234b, 1258b, 1270b, 1297b, 1336b, 1342b, 1343b, 1348b, 1373b, 1394b, 1396b, 1420b, 1425b, 1447b, 1450b, 1460b, 1465b, 1487b, 1503b, 1530b, 1538b, 1543b, 1565b, 1577b, 1579b, 1587b, 1590b, 1597b, 1604b, 1613b, 1633b, 1637b, 1649b, 1652b, 1661b, 1662b, 1664b, 1669b, 1671b, 1676b, 1707b, 1711b, 1716b, 1722b, 1730b, 1732b, 1742b, 1754b, 1757b, 1771b, 1777b, 1784b, 1787b, 1789b, 1791b, 1793b, 1795b, 1813b, 1826b, 1832b, 1839b, 1847b, 1854b, 1856b, 1860b, 1862b, 1875b, 1896b, 1913b, 1930b, 1936b, 1937b, 1941b, 1943b, 1945b, 1948b, 1950b, 1954b, 1955b, 1956b, 1983b, 2015b, 2033b, 2053b, 2086b, 2101b, 2119b, 2122b, 2140b, 2159b, 2162b, 2177b, 2217b, 2233b, 2235b, 2239b, 2258b, 2305b, 2321b, 2323b, 2326b, 2327b, 2333b, 2337b, 2340b, 2370b, 2375b, 2406b, 2411b, 2434b, 2436b, 2452b, 2517b, 2521b, 2523b, 2557b, 2569b, 2588b, 2594b, 2607b, 2627b, 2640b, 2642b, 2654b, 2657b, 2663b, 2674b, 2682b, 2692b, 2698b, 2709b, 2711b, 2712b, 2713b, 2715b, 2725b, 2726b, 2727b, 2729b, 2752b, 2766b, 2769b, 2778b, 2786b, 2792b, 2832b, 2833b, 2840b, 2848b, 2866b, 2921b, 2935b, 2939b, 2945b, 2946b, 2963b, 2967b, 2970b, 2977b, 3019b, 3034b, 3044b, 3059b, 3060b, 3069b, 3081b, 3082b, 3103b, 3109b, 3117b, 3119b, 3148b, 3158b, 3189b, 3190b, 3194b, 3198b, 3201b, 3207b, 3210b, 3238b, 3243b, 3259b, 3289b, 3297b, 3300b, 3311b, 3317b, 3331b, 3343b, 3368b, 3369b, 3381b, 3385b, 3400b, 3432b, 3442b, 3460b, 3470b, 3483b, 3489b, 3515b, 3520b, 3536b, 3577b, 3584b, 3586b, 3605b, 3619b, 3621b, 3645b, 3652b, 3655b, 3658b, 3662b, 3664b, 3667b, 3670b, 3674b, 3679b, 3690b, 3695b, 3701b, 3719b, 3733b, 3736b, 3739b, 3752b, 3769b, 3776b, 3782b, 3802b, 3812b, 3817b, 3826b, 3839b, 3848b, 3857b, 3875b, 3893b, 3907b, 3924b, 3931b, 3932b, 3944b, 3952b, 3959b, 3983b, 4001b, 4020b, 4042b, 4046b, 4067b, 4071b, 4078b, 4083b, 4095b, 4099b, 4107b, 4110b, 4129b, 4138b, 4157b, 4173b, 4178b, 4192b, 4195b, 4205b, 4218b, 4231b, 4242b, 4248b, 4268b, 4295b, 4306b, 4307b, 4313b, 4318b, 4319b, 4320b, 4325b, 4328b, 4340b, 4343b, 4360b, 4377b, 4397b, 4401b, 4405b, 4410b, 4416b, 4420b, 4429b, 4434b, 4438b, 4450b, 4457b, 4459b, 4460b, 4465b, 4495b, 4496b, 4506b, 4539b, 4548b, 4570b, 4572b, 4599b, 4641b, 4642b, 4649b, 4658b, 4661b, 4678b, 4686b, 4703b, 4706b, 4709b, 4710b, 4733b, 4739b, 4748b, 4781b, 4787b, 4791b, 4792b, 4814b, 4817b, 4820b, 4829b, 4831b, 4841b, 4851b, 4871b, 4884b, 4937b, 4939b, 4940b, 4949b, 4975b, 4995b, 5057b, 5067b, 5069b, 5085b, 5112b, 5157b, 5160b, 5239b, 5254b, 5255b, 5267b, 5272b, 5278b, 5287b, 5300b, 5303b, 5311b, 5339b, 5346b, 5347b, 5359b, 5374b, 5383b, 5387b, 5388b, 5397b, 5400b, 5405b, 5410b, 5419b, 5421b, 5425b, 5432b, 5438b, 5445b, 5451b, 5502b, 5535b, 5561b, 5573b, 5585b, 5612b, 5616b, 5622b, 5698b, 5726b, 5744b, 5763b, 5765b, 5770b, 5790b, 5791b, 5792b, 5794b, 5795b, 5801b, 5804b, 5814b, 5821b, 5823b, 5826b, 5837b, 5856b, 5873b, 5874b, 5900b, 5903b, 5906b, 5950b, 5959b, 5960b, 5964b x . . . Psx a-verse with single alliteration: 44a, 108a, 138a, 230a, 285a, 309a, 344a, 362a, 709a, 730a, 789a, 870a, 875a, 931a, 1050a, 1413a, 1445a, 1553a, 1868a, 1899a, 1905a, 1909a, 2128a, 2441a, 2461a, 2610a, 2727a, 2728a, 2762a, 2783a, 2785a, 2979a, 3010a, 3260a, 3283a, 3285a, 3338a, 3366a, 3397a, 3555a, 3731a, 3736a, 3790a, 3801a, 3873a, 3885a, 3917a, 4045a, 4085a, 4099a, 4148a, 4156a, 4157a, 4197a, 4201a, 4219a, 4225a, 4228a, 4466a, 4477a, 4530a, 4557a, 4574a, 4623a, 4734a, 4773a, 4930a, 5005a, 5143a, 5305a, 5334a, 5340a, 5408a, 5409a, 5415a, 5422a, 5441a,

MOSP-02.qxd

5/11/04

120

2:38 PM

Page 120

2. Metrical types and positions

5461a, 5462a, 5465a, 5662a, 5672a, 5704a, 5772a, 5787a, 5852a, 5853a; b-verse: 58b, 100b, 241b, 249b, 343b, 533b, 539b, 582b, 622b, 786b, 879b, 920b, 921b, 1032b, 1046b, 1080b, 1095b, 1113b, 1119b, 1239b, 1401b, 1414b, 1438b, 1488b, 1540b, 1728b, 1750b, 1833b, 1871b, 1910b, 1932b, 2110b, 2271b, 2294b, 2301b, 2351b, 2503b, 2575b, 2604b, 2718b, 2733b, 2738b, 2874b, 3221b, 3327b, 3348b, 3375b, 3405b, 3453b, 3458b, 3468b, 3523b, 3528b, 3595b, 3792b, 3816b, 3834b, 3898b, 3905b, 4014b, 4221b, 4227b, 4352b, 4362b, 4437b, 4544b, 4553b, 4614b, 4620b, 4638b, 4742b, 4754b, 4799b, 5002b, 5058b, 5075b, 5128b, 5199b, 5200b, 5221b, 5246b, 5250b, 5256b, 5266b, 5277b, 5284b, 5285b, 5309b, 5389b, 5420b, 5489b, 5553b, 5583b, 5586b, 5724b, 5902b x . . . PSx a-verse with double alliteration: 1027a, 2144a, 3294a, 3447a; a-verse with single alliteration: 4a, 80a, 84a, 89a, 99a, 155a, 162a, 183a, 210a, 238a, 244a, 353a, 376a, 388a, 398a, 417a, 434a, 436a, 513a, 609a, 616a, 618a, 632a, 640a, 655a, 824a, 839a, 857a, 868a, 869a, 986a, 987a, 1034a, 1038a, 1111a, 1116a, 1121a, 1155a, 1193a, 1212a, 1226a, 1254a, 1290a, 1351a, 1410a, 1520a, 1534a, 1575a, 1636a, 1651a, 1817a, 1834a, 1839a, 1849a, 1874a, 1893a, 1953a, 1979a, 2029a, 2117a, 2131a, 2201a, 2287a, 2419a, 2437a, 2468a, 2487a, 2508a, 2605a, 2613a, 2624a, 2641a, 2651a, 2660a, 2679a, 2736a, 2757a, 2791a, 2811a, 2829a, 2880a, 3041a, 3049a, 3120a, 3150a, 3158a, 3167a, 3175a, 3259a, 3424a, 3492a, 3511a, 3592a, 3727a, 3744a, 3826a, 3843a, 3852a, 4169a, 4180a, 4269a, 4294a, 4363a, 4387a, 4412a, 4515a, 4551a, 4559a, 4594a, 4668a, 4694a, 4708a, 4803a, 4878a, 4927a, 4999a, 5026a, 5051a, 5131a, 5172a, 5174a, 5220a, 5221a, 5235a, 5236a, 5275a, 5290a, 5301a, 5328a, 5382a, 5386a, 5396a, 5403a, 5404a, 5521a, 5563a, 5645a, 5739a, 5807a, 5877a, 5896a; b-verse: 45b, 248b, 275b, 321b, 325b, 411b, 414b, 488b, 551b, 568b, 672b, 859b, 928b, 998b, 1001b, 1002b, 1022b, 1036b, 1043b, 1051b, 1096b, 1098b, 1153b, 1161b, 1200b, 1274b, 1303b, 1375b, 1422b, 1469b, 1490b, 1510b, 1615b, 1635b, 1867b, 1870b, 1878b, 1914b, 1929b, 2084b, 2133b, 2154b, 2275b, 2421b, 2473b, 2486b, 2492b, 2495b, 2511b, 2518b, 2600b, 2623b, 2626b, 2637b, 2645b, 2756b, 2842b, 2885b, 3080b, 3199b, 3217b, 3222b, 3330b, 3461b, 3597b, 3598b, 3617b, 3643b, 3659b, 3722b, 3762b, 4114b, 4141b, 4255b, 4277b, 4301b, 4356b, 4357b, 4422b, 4461b, 4473b, 4519b, 4573b, 4887b, 4918b, 4922b, 4971b, 4979b, 4983b, 5014b, 5043b, 5063b, 5094b, 5152b, 5193b, 5204b, 5224b, 5237b, 5280b, 5391b, 5444b, 5488b, 5498b, 5604b, 5714b, 5842b, 5920b, 5962b, 5969b x . . . PXx a-verse with double alliteration: 278a, 419a, 2723a, 4680a; a-verse with single alliteration: 143a, 188a, 213a, 402a, 553a, 613a, 638a, 719a, 864a, 876a, 918a, 976a, 1013a, 1023a, 1025a, 1026a, 1053a, 1060a, 1074a, 1188a, 1448a, 1467a, 1630a, 1838a, 1889a, 1928a, 2046a, 2061a, 2141a, 2157a, 2329a, 2377a, 2433a, 2531a, 2674a, 2695a, 2786a, 2803a, 2812a, 3217a, 3254a, 3280a, 3407a, 3433a, 3451a, 3470a, 3593a, 3803a, 3818a, 3903a, 3937a, 4027a, 4055a, 4291a, 4356a, 4365a, 4436a, 4512a, 4554a, 4582a, 4656a, 4679a, 4687a, 4717a, 4741a, 4749a, 4859a, 4969a, 5346a, 5433a, 5531a, 5543a, 5589a, 5606a, 5816a, 5830a; b-verse: 186b, 190b, 356b, 365b, 394b, 587b, 699b, 753b, 927b, 996b, 1044b, 1215b, 2058b, 2095b, 2103b, 2121b, 2276b, 2281b, 2809b, 2873b, 2883b, 3130b, 3171b, 3215b, 3345b, 3391b, 3517b, 3574b, 3889b, 3897b, 4028b, 4124b, 4199b, 4223b, 4331b, 4467b, 4556b, 4797b, 5167b, 5872b x . . . Pxx a-verse with single alliteration: 56a, 68a, 82a, 294a, 351a, 589a, 596a, 861a, 1087a, 1158a, 1229a, 1336a, 1714a, 1940a, 2088a, 2262a, 2603a, 2981a, 3032a, 3459a, 3671a, 3859a, 4239a, 4248a, 4250a, 4252a, 4390a, 4597a, 4646a, 4699a, 4772a, 4864a, 4962a, 4979a, 5018a, 5021a, 5063a, 5122a, 5135a, 5153a, 5243a, 5261a, 5610a, 5642a; b-verse: 2304b, 3519b

The three major classes are ranked in the decreasing order of frequency with which the b-verse is composed, as represented in (59) below. When the second lift is filled by a primary-stressed syllable (P or p), that is, class 1, type C occurs in the b-verse with overwhelming frequency. At the opposite extreme is found class 3, with the corresponding position realised by a syllable lacking lexical stress: the incidence of the b-verse is minimal here. Class 2, on the other hand, is more or less evenly distributed in the a-verse and the b-verse. Accordingly, we could hardly miss the generalisation that the greater the second lift is in terms of prominence, the greater the likelihood is to occur in the b-verse.

MOSP-02.qxd

5/11/04

2:38 PM

Page 121

2.6. Type C

121

(59) Ranking of the six major variants of type C in the decreasing order of frequency of the b-verse x . . . P#px (88%) 傻 x . . . P#Px (78%) 傻 x . . . Psx (52%) 傻 x . . . PSx (43%) 傻 x . . . PXx (33%) 傻 x . . . Pxx (4%) In addition to the difference in stress degrees, that of syllable length appears to play a secondary role in determining verse distribution in fine detail, as may be detected from the above representation. For class 1, the appearance of the short (unresolved) syllable in the second lift is distinguished by the diminished frequency of the a-verse with double alliteration and proportionately the higher incidence of the b-verse. This distribution pattern may be ascribed to the weaker prominence of the unresolved lift that makes alliteration less feasible to occur on it than on the fully realised lift, namely the long syllable. By contrast, the reverse pattern obtains for class 3, for which the occurrence of the short syllable enhances use of the a-verse. This distinction is a transparent reflection of the general pattern noted above, whereby the lesser the prominence involved, the likelier it is to appear in the a-verse. Between these two contrastive classes is located class 2, for which the opposition between long and short syllables would appear less conspicuous. Indeed, the proportion of the a-verse with single alliteration to the b-verse remains more or less even in overall terms regardless of syllable length, yet the short syllable does seem favourable to a higher incidence of the b-verse. We may provide a further, complementary account for the above group-internal difference related to syllable length as follows. The second lift of type C is characterised by near-obligatory suspension of resolution, as treated in section 3.1.1.3 below. In other words, the position is realised through association with a stressed syllable regardless of its exact length, rather than with a foot as in contexts in which resolution is normally implemented at the expense of suspension of resolution. Since the long stressed syllable is indistinguishable from a foot in surface manifestation, we may assume that the short counterpart constitutes the unmarked variant inasmuch as it is distinguished from a foot fully and unambiguously. The marked status of the long stressed syllable as an occupant of the second lift of type C is represented vicariously by the predominance of resolution on the first lift: as will be substantiated in section 3.1.1.6 below, the long primary-stressed syllable P associated with the second lift of type C induces resolution on the immediately preceding lift. Since the resolved foot otherwise occurs less frequently than the monosyllabic one, we may reasonably interpret the reversed pattern as a reflection of the marked status of the following entity. Given such difference in terms of markedness, it would be most natural that the unmarked variant, namely the short stressed syllable, is preferred to the long counterpart as a realisation of the second lift of type C. Accordingly, the b-verse, which is most in harmony with the type C variants having the second lift realised by a primary-stressed syllable, exhibits preference for the unmarked string x . . . P#px over the marked x . . . P#Px; and similarly but to a lesser extent, the same verse favours the sequence x . . . Psx as against x . . . PSx. Essentially the same reasoning applies to the pair of unstressed syllables, short and long. Since the prototypical realisation of the second lift of type C is implemented through association with the level of the syllable rather than the foot, being short may count as unmarked regardless of accentuation. The association of the second lift with an unstressed

MOSP-02.qxd

5/11/04

2:38 PM

122

Page 122

2. Metrical types and positions

syllable, however, is privileged by the a-verse rather than the b-verse. Therefore, the unmarked member, x . . . Pxx, occurs with greater frequency in the a-verse. Whatever the precise mechanism underlying the class-internal distinction based on syllable length, the frequency of the b-verse reaches the minimum when the second lift is occupied by the short medial syllable of a non-compound, while it is maximal when the position is filled by the primary-stressed short syllable. Given the neatly patterned distribution of type C variants in accordance with the varying degrees of prominence on their second lift in the Heliand, it will be appropriate to compare with the situation in Beowulf. Table 2.78 indicates the distribution of the corresponding type C variants in Beowulf. Of particular interest is the lack of differentiation among the variants with the second lift realised by non-primary-stressed syllables. While the configurations x . . . P#Px and x . . . P#px predominantly appear in the b-verse analogous to the Heliand, the remaining sequences are all favoured in the a-verse with single alliteration without significant distinction among them. More specifically, the distribution is consistently in favour of the a-verse with single alliteration regardless of whether the second lift is associated with the second element of a compound or a part of a non-compound word: thus, the configurations x . . . Psx and x . . . Pxx, which stand at the polar opposites among the four variants x . . . Psx, x . . . PSx, x . . . PXx, and x . . . Pxx in the Heliand, are treated as equivalent in Beowulf. We may accordingly interpret the situation in the Heliand as another manifestation of the reorganisation by the Heliand poet of metrical features towards greater degrees of patterning and gradation. The reconfigured patterning of type C observed in the Heliand may be contrasted with the conservative treatment of type B1 with respect to the distinction of x . . . PxP and x . . . PxS. As pointed out in section 2.4.1 above, the Heliand poet maintained the metrical opposition between x . . . PxP and x . . . PxS in its traditional form, whereas he reorganised the traditional patterning in regard to the distinction between x . . . P#Px and x . . . PSx. We may naturally wonder what would have motivated this divergent treatment. More specifically, why was only type C subject to a reorganisation based on gradation? The question may be meaningfully related to the differential treatment of the traditional types A2a and A2b. As discussed in section 2.2.1 above, only type A2a was recast into a scalar opposition, and type A2b was kept intact, standing as it did in binary opposition to type A1. Unlike type C, the second lift of type B1 was limited to P (including px) Table 2.78. Distribution of major variants of type C in Beowulf variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

total

x x x x x x

12 5% 26 10% 11 5% 5 3% 3 2% 0 0%

34 13% 9 3% 151 67% 145 74% 102 76% 25 71%

219 83% 229 87% 64 28% 46 23% 30 22% 10 29%

265 264 226 196 135 35

. . . . . .

. . . . . .

. . . . . .

P#px P#Px Psx PSx PXx Pxx

100% 100% 100% 100% 100% 100%

MOSP-02.qxd

5/11/04

2:38 PM

Page 123

2.6. Type C

123

or S (including sx) for association. The syllable lacking lexical stress (X or x) was excluded as a possible occupant on structural grounds: the sequences PxX and Pxx were categorically disallowed as realisations of the metrical configuration / ⫻ /. Thus, the poet had only two options for the second lift of type B1, P or S, much as he had a binary choice for the second drop of class A, the lexical-stressed or the unstressed syllables. Accordingly, given the minimal range of choice, there would have been no basis for introducing a gradient distinction as an improvement on the traditional binary opposition. 2.6.3. The variation on the second drop of type C The second drop of type C is more limited than that of type A1 in terms of the variety of linguistic realisations, despite both positions being commonly located in versefinal position: it is always associated with a word-final unstressed syllable, with the following rare exceptions. First, there are three examples in which the final drop is filled by the second constituent of a compound, a feature unknown to Beowulf (Suzuki 1996a: 95):23 (60) 2062a is thit folc frômôd 1057b uuânde that he god ênfald 1062b uuânda that he man ênfald In this connection, we may recall the following verses noted in section 2.4.2.4 above, which one may be tempted to regard as ending in an independent word serving as the second drop of type C: (61) 4458a that nu obar tuâ naht sind 1373b Sô uuirðid them, the that godes uuord scal However, the above two verses might be scanned as type B1 as well, as observed in section 2.4.2.4 above. The scansion as type B1 would seem to be supported by the absence of comparable a-verse with double alliteration in the corpus: that is, the configuration x . . . P#P#P (P ⫽ alliterating primary-stressed syllable) is totally absent, unlike the configuration x . . . P#PS mentioned above, for which verse 2062a given in (60) earlier provides empirical support. This argument for scanning the two verses as type B1, however, is undermined by the following distributional property of type C: the type C variant x . . . P#Px appears in the b-verse with overwhelming frequency, c. 80 per cent, as pointed out above (Table 2.77). In this light, and in view of the extreme rarity of relevant examples even of the a-verse with single alliteration and the b-verse (one instance for each; see (61) above), the absence of analogous a-verses with double alliteration would be hardly surprising on statistical grounds and therefore could scarcely be adduced as decisive evidence against scanning the two verses in question given in (61) above as type C. Furthermore, the verse-final words at issue (sind, scal) are not class 1 words (substantives) most 23

Yet Beowulf contains a single example of type C in which the final drop is occupied by the second element of a quasi compound, Beo 2309b Wæs se fruma egeslic, comparable to verse 3135a uuas thar gard gôdlîc in the Heliand.

MOSP-02.qxd

5/11/04

124

2:38 PM

Page 124

2. Metrical types and positions

likely to serve as a lift. In this respect, these two verses may be compared to the type A1 variant Px . . . P#x, such as 4618a dô that thu duan scalt and 5104a gihid that he god sî. The verse-final function words involved (scalt, sî) are demonstrably identical in their metrical capacity to the word-final unstressed syllable -x# serving as the final drop of type A1, as argued in section 2.2.1 above. Given this equivalence of the configuration Px . . . P#x with Px . . . Px, then, the two verses under consideration (61) may be scanned as x . . . P#P#x, and therefore identified with the configuration x . . . P#Px, type C. On balance, then, scanning these two verses as type C seems to have more to recommend it than their alternative scansion as type B1. Another variant of the second drop of type C may be realised by a disyllabic ending, as exemplified below by the configurations x . . . P#Pxx, x . . . Psxx, x . . . PXxx, and x . . . Pxxx: (62) 2850a an grase gruonimu 4741a that he im thero costondero 928b ef thu tharo forasagono These verses, particularly the configuration x . . . P#Pxx, should not be equated with the configuration x . . . P#Pxx, a variant of type D with anacrusis, such as 2854b thene meti uuîhide, 4635b endi gode thancode. As shown in section 2.7 below, this configuration is strongly favoured in the b-verse, occurring there nearly seven times out of every ten. Kauffmann (1887: 327, 329), Sievers (1893: §116.1), and Martin (1907: 32), however, failed to distinguish between these two configurations and scanned the configuration x . . . P#Pxx such as verse 18b sô uuârun thia man hêtana as type C, an error analogous to the one they committed in scanning the configuration Px#Pxx as type A1 (section 2.2.2). Thus, because of the lower incidence of the b-verse and the correspondingly higher frequency of the a-verse with single alliteration, the sequence x . . . P#Pxx has to be distinguished from x . . . P#Pxx and characterised as a variant of type C. We now have to account for the extremely rare occurrence in the second drop of type C of syllables other than the minimal one (i.e., the unstressed word-final syllable -x#) in contrast to the corresponding position of type A1 and type A2, which allows for a wider range of material for association (section 2.2). Most briefly here, the near-exclusive use of the unstressed word-final syllable may be ascribed to the syntagmatically motivated requirement that the final position of type C should best be kept to a minimum in prominence due to the contiguity of two strong positions immediately preceding it. For details on this restriction specific to type C, see section 3.1.1.6 below. Table 2.79. Distribution of the type C variant ending in -xx# a-verse with double alliteration

a-verse with single alliteration

b-verse

1 (11%)

4 (44%)

4 (44%)

a-verse with double alliteration: 2850a; a-verse with single alliteration: 2262a, 3049a, 3514a, 4741a; b-verse: 928b, 1422b, 3944b, 4107b

MOSP-02.qxd

5/11/04

2:38 PM

Page 125

2.7. Type D

125

2.7. Type D (/ / ⫻ ⫻) This type comprises three major variants, subtypes D1, D2a, and D2b, which are differentiated by the linguistic realisations of the two drops that constitute the last two positions. Subtype D1 is distinguished from subtypes D2a and D2b by the absence of a lexical-stressed syllable in either drop. Commonly characterised by the association of one of the drops with a lexical-stressed syllable, subtypes D2a and D2b in turn are differentiated by the location of the lexical stress involved, in the first drop (subtype D2a) on the one hand and in the second drop (subtype D2b) on the other. Subtype D1 has two realisation variants, depending on whether or not the second lift (second position) is occupied by a primary-stressed syllable. When this position is associated with a primary-stressed syllable, that is, the head of an independent word, we obtain the variant P#PX/xx, which is further subdivided into two configurations P#Pxx and P#PXx according to length of the syllable that occupies the first drop, as exemplified in (63) and listed under Table 2.80 below (cf. Kauffmann 1887: 333–6; Hofmann 1991, II: 174–8): (63) 1446a erl ôðarna 2366a sunu drohtines 3673b uuel huggendies Another realisation variant of subtype D1 has its second lift filled by the second constituent of a compound, and the whole verse (apart from anacrusis if any) is accordingly realised by a single compound word of the form Psxx, PsXx, PSxx, or PSXx, as exemplified below (cf. Hofmann 1991, II: 178–80). Analogous to the pair of configurations P#Pxx and P#PXx but in finer distinction, this variant may be divided into four configurations, Psxx, PsXx, PSxx, and PSXx, according to syllable length of the second lift as well as of the first drop. (64) 3186a 82b 222b 4316b

aðalkêsures hebancuninge aðalboranes erðbûandiun

The finer differentiation of compound word verses predicated on syllable length of the second lift stems from the limitation of suspension of resolution to the secondary-stressed syllable, as will be discussed at length in sections 3.1.1.3 and 3.1.1.6 below. As shown in Table 2.80 below, the overall distribution of subtype D1 is largely characterised by the predominance of the b-verse and conversely the low incidence of the a-verse with double alliteration. Seemingly deviant from this generalisation is the configuration P#PXx, which shows a marked preference for the a-verse with double alliteration. Such an apparent exceptionality, however, makes sense when looked at from a wider context: the six realisation variants of subtype D1 are graded along a scale by the varying extent to which the b-verse on the one hand and the a-verse with double alliteration on the other are used. The graded patterning is

MOSP-02.qxd

5/11/04

126

2:38 PM

Page 126

2. Metrical types and positions

conditioned by three parameters in the following order of decreasing scope, first by the syllable length of the second lift, second by the syllable length of the first drop, and finally by the stress properties (lexical versus derived) of the second lift. The association of the second lift with a short syllable through suspension of resolution on the one hand, and the realisation of the first drop by a short syllable on the other, maximise use of the b-verse and minimise incidence of the a-verse with double alliteration. Conversely, the b-verse occurs least frequently and the a-verse with double alliteration figures most prominently when the second lift is realised by a long primarystressed syllable, and the first drop by a long syllable. The scaling thus obtained is represented as follows: (65) Ranking of the six major variants of subtype D1 in the decreasing order of frequency of the b-verse and/or in the increasing order of frequency of the a-verse with double alliteration Psxx 傻 PsXx 傻 PSxx 傻 P#Pxx 傻 PSXx 傻 P#PXx Since the configuration P#PXx is accordingly removed maximally from the optimal values for the b-verse among the six variants in question, it is a matter of course that it should stand out by the ‘deviant’ feature of the relatively low frequency of the b-verse. On several occasions we have observed that the disyllabic ending -xx# is metrically distinguished from other disyllables -xx#, as in the realisation of the first drop of type A1 (section 2.1.4). The distinction also holds good for the composition of subtype D1. Specifically, there are no subtype D1 verses in which the disyllabic ending is associated with the two drops constitutive of it: the configuration P#Pxx is thus unmetrical in contrast to the metrical sequence P#Pxx treated above. When a disyllabic ending is preceded by another unstressed syllable (-x-), however, we obtain metrical though marginal variants as follows: (66) 2722a unsundigane (PSxxx) 2689b he ni uuas iro cûð ênigumu (x . . . P#Pxxx) In the above verses, the first constituent of the disyllabic endings involved is associated with the first drop along with the immediately preceding short unstressed syllable, as will be shown in section 2.12.4 below. Subtype D2a has its first drop (penultimate position) typically filled by the second element of a compound, and the whole configuration is accordingly P#PS/sx, as illustrated below (cf. Kauffmann 1887: 333–6; Hofmann 1991, II: 174–8): (67) 1041a hôh himilrîki 2337a god alomahtig 4421b godes andsacun The distribution pattern of this subtype as a whole is nearly the reverse of subtype D1 (Table 2.80): the a-verse with double alliteration prevails over the other two verses, accounting for two-thirds of all occurrences. On closer examination, this variant is found differentiated in its distribution pattern, conditioned by length of the

MOSP-02.qxd

5/11/04

2:38 PM

Page 127

2.7. Type D

127

Table 2.80. Distribution of major variants of type D variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

Psxx PsXx PSxx PSXx

0 0% 0 0% 0 0% 3 23%

0 0% 3 17% 3 38% 3 23%

2 100% 15 83% 5 63% 7 54%

PS/sX/xx

3

7%

9 22%

29

71%

P#Pxx P#PXx

13 13% 7 44%

22 22% 4 25%

65 5

65% 31%

P#PX/xx

20 17%

26 22%

70

60%

P#P#px P#P#Px

3 30% 3 75%

3 30% 0 0%

4 1

40% 25%

P#P#P/px

6 43%

3 21%

5

36%

13%

21

29%

P#Px . . . P

42

58%

P#Psx P#PSx

23 96% 36 58%

0 0% 11 18%

1 15

4% 24%

P#PS/sx

59 69%

11 13%

16

19%

P#Px . . . S

12

2

14%

86%

9

0

0%

Psxx b-verse: 222b, 464b PsXx a-verse with single alliteration: 130a, 1472a, 2114a; b-verse: 82b, 91b, 278b, 362b, 765b, 781b, 1120b, 1939b, 2344b, 2377b, 2443b, 2767b, 2981b, 4258b, 5575b PSxx a-verse with single alliteration: 3186a, 3195a, 3305a; b-verse: 678b, 3601b, 3610b, 3827b, 4275b PSXx a-verse with double alliteration: 1196a, 2918a, 2964a; a-verse with single alliteration: 343a, 2722a, 3086a; b-verse: 815b, 2771b, 2839b, 2909b, 4316b, 4919b, 5276b P#Pxx a-verse with double alliteration: 862a, 1502a, 2084a, 2261a, 2336a, 2370a, 2428a, 2708a, 3373a, 4586a, 5024a, 5410a, 5667a; a-verse with single alliteration: 446a, 516a, 702a, 770a, 2366a, 2808a, 2999a, 4026a, 4078a, 4136a, 4301a, 5110a, 5193a, 5202a, 5207a, 5313a, 5356a, 5431a, 5568a, 5788a, 5837a, 5850a; b-verse: 18b, 77b, 349b, 526b, 534b, 594b, 751b, 805b, 834b, 950b, 992b, 1045b, 1145b, 1418b, 1596b, 2005b, 2073b, 2090b, 2190b, 2199b, 2244b, 2259b, 2417b, 2621b, 2812b, 2854b, 2969b, 2974b, 2980b, 3029b, 3115b, 3350b, 3542b, 3546b, 3602b, 3603b, 3787b, 3980b, 3984b, 4012b, 4053b, 4088b, 4109b, 4259b, 4272b, 4590b, 4631b, 4635b, 4744b, 4790b, 4800b, 4855b, 4992b, 5056b, 5175b, 5188b, 5230b, 5241b, 5288b, 5475b, 5515b, 5702b, 5789b, 5848b, 5926b P#PXx a-verse with double alliteration: 384a, 721a, 1446a, 1897a, 4071a, 5685a, 5731a; a-verse with single alliteration: 289a, 485a, 1299a, 2872a; b-verse: 562b, 669b, 697b, 2689b, 3673b

MOSP-02.qxd

5/11/04

128

2:38 PM

Page 128

2. Metrical types and positions

P#P#px a-verse with double alliteration: 412a, 1163a, 2900a; a-verse with single alliteration: 4062a, 5481a, 5584a; b-verse: 1692b, 2908b, 3364b, 5605b P#P#Px a-verse with double alliteration: 2242a, 3059a, 5087a; b-verse: 2238b P#Px . . . P a-verse with double alliteration: 261a, 599a, 626a, 661a, 671a, 693a, 916a, 957a, 984a, 1295a, 1962a, 2145a, 2601a, 2813a, 2860a, 2862a, 2910a, 2917a, 3081a, 3117a, 3163a, 3198a, 3273a, 3325a, 3653a, 3669a, 3759a, 4256a, 4432a, 4495a, 4573a, 4611a, 5140a, 5169a, 5170a, 5380a, 5392a, 5743a, 5802a, 5907a, 5929a, 5966a; a-verse with single alliteration: 794a, 838a, 1010a, 1335a, 1596a, 2100a, 3085a, 3108a, 4778a; b-verse: 174b, 196b, 326b, 369b, 960b, 982b, 1135b, 1287b, 2000b, 2250b, 2291b, 2632b, 2762b, 4102b, 4241b, 4329b, 4622b, 4824b, 4879b, 5490b, 5803b P#Psx a-verse with double alliteration: 240a, 1097a, 1325a, 1396a, 1511a, 1946a, 2186a, 3044a, 3240a, 3718a, 3765a, 3922a, 4128a, 4279a, 4323a, 4449a, 4703a, 4866a, 5049a, 5314a, 5398a, 5476a, 5558a; b-verse: 4421b P#PSx a-verse with double alliteration: 256a, 297a, 620a, 775a, 823a, 905a, 913a, 915a, 998a, 1041a, 1500a, 1501a, 1622a, 1851a, 2007a, 2146a, 2332a, 2446a, 2574a, 2607a, 2620a, 3137a, 3411a, 3489a, 3512a, 3925a, 4040a, 4052a, 4106a, 4313a, 4350a, 4463a, 5443a, 5619a, 5741a, 5948a; a-verse with single alliteration: 245a, 813a, 1297a, 1334a, 1619a, 2337a, 2451a, 2459a, 3296a, 3508a, 4957a; b-verse: 153b, 930b, 1759b, 1766b, 1922b, 3139b, 3937b, 3954b, 3967b, 3974b, 4038b, 4159b, 4746b, 5635b, 5781b P#Px . . . S a-verse with double alliteration: 92a, 110a, 197a, 512a, 1356a, 1773a, 2266a, 2907a, 3882a, 4425a, 4448a, 5899a; b-verse: 2699b, 3628b

syllable that occupies the first drop, so that we may distinguish between the configurations P#Psx and P#PSx: the shorter configuration (P#Psx) is characterised by a higher frequency of the a-verse with double alliteration than the longer counterpart (Table 2.80). In addition, there is a further variant of subtype D2a, P#P#P/px, in which the first drop is realised by a primary-stressed syllable (long or short), as shown below: (68) 412a 4062a 2238b 5605b

fagar folc godes Krist godes sunu Segel up dâdun sehan lioht godes

As with the pair of configurations P#Psx/P#PSx, we may make a distinction between P#P#px and P#P#Px for this variant as well. Although the extreme rarity of relevant examples makes it difficult to generalise, we can hardly miss a significant difference between the two configurations in their metrical behaviour. The two configurations P#P#px and P#P#Px exhibit differing distribution patterns: the longer one, P#P#Px, shows a stronger preference for the a-verse with double alliteration than the shorter one, P#P#px. The configuration P#P#px, when attended by single alliteration, that is, P#P#px, would appear ambiguous in categorisation, scannable as it is either as subtype D2a or as subtype A1s, as observed in section 2.1.8 above and discussed more fully in section 3.2.2 below. By reference to syntactic constituency, however, this configuration will lend itself to disambiguation: the instances with a major syntactic break

MOSP-02.qxd

5/11/04

2:38 PM

Page 129

2.7. Type D

129

after the second word are scanned as subtype A1s (section 2.1.8), and those with a syntactic break before the second word (i.e., verses 4062a, 5584a, 2908b, 3364b, 5605b, as given in (68) above), as subtype D2a. Subtype D2b is distinguished from subtype D2a by the presence of a lexicalstressed syllable in the second drop (final position). This position may be occupied by the second element of a compound (P#Px . . . S; cf. Kauffmann 1887: 335–6; Hofmann 1991, II: 193–5, 197): (69) 1773a al irminthiod 2699b godes ambahtman As with the configuration P#PS/sx (subtype D2a), but to a fuller extent, the a-verse with double alliteration constitutes a typical instantiation of this subtype, as indicated in Table 2.80. Furthermore, the second drop of subtype D2b may be realised by an independent lexical word (P#Px . . . P), as follows, comparable to the first drop of subtype D2a treated above: (70) 626a liof landes uuard 1010a godes êgan barn 4102b Uueros gengun tô When attended by single alliteration (P#Px . . . P), the type D2b variant with the second drop associated with an independent word may appear structurally ambiguous, as it is also scannable as type E in terms of metrical configuration. By invoking difference in syntactic constituency, however, we are capable of making a proper scansion, as will be treated fully in section 3.2.2 below. For the sake of presentation, we are anticipating here the unambiguous scansion based on syntactic constituency. A further variation on the realisation of subtype D2b is manifested by the treatment of the first drop of subtype D2b in two different ways. First, the position may be occupied by a disyllable (cf. Kauffmann 1887: 335):24 (71) 599a giboran bald endi strang 957a garu gumono sô huuem 2145a hard helleo gethuing 5802a lîf langerun huîl a-verse with double alliteration: 197a, 512a, 599a, 693a, 957a, 2145a, 2860a, 2910a, 2917a, 3117a, 3273a, 5169a, 5802a, 5966a; b-verse: 4241b, 4824b, 5490b Compared with Beowulf, in which the corresponding position is occasionally realised by a disyllable as well, the Heliand allows a greater variety of language material to appear in this position. While the disyllable observed in Beowulf is either 24

There is further a single example which has its first drop filled by a trisyllabic sequence: 5929a bi namen neriendero best.

MOSP-02.qxd

5/11/04

130

2:38 PM

Page 130

2. Metrical types and positions

a sequence of a word-final syllable and a prefix -x#x- as in Beo 2774a eald enta geweorc or a sequence of medial syllables -xx- as in Beo 387a seon sibbegedriht (Suzuki 1996a: 161–2), we find in the Heliand a disyllabic function word #xx# (endi, 599a), a sequence of a word-final syllable and a function word -x#x# (-no sô, 957a), and a sequence of medial and word-final syllables -xx# (-gerun, 5802a), besides a sequence of a word-final syllable and a prefix -x#x- (-leo ge-, 2145a). No less significant is the higher proportion of such verses to those with a monosyllabic first drop: in the Heliand as many as seventeen out of a total of eighty-six verses, c. 20 per cent, contain a disyllabic first drop, while in Beowulf there are only six examples (8 per cent) out of a total of seventy-five type D2b examples. We witness here again a further manifestation of the expansion of the normal drop in its range of linguistic realisations. Second, the first drop of subtype D2b may be occupied by the second element of a compound, as exemplified by three instances. These verses are highly similar in composition, closing with the same expression -lîc thing, and two of them also share a preceding word, manag:25 (72) 1295a manag mârlîc thing 5380a sô manag mislîc thing 5907a sehan seldlîc thing Three points are noteworthy about this minor group. First, there are no examples in which the second and third positions are realised by a true compound rather than a quasi compound, that is, a complex word ending in a heavy derivational suffix like -lîc. Second, the first lift is resolved (manag, sehan) in all three examples. Finally, subtype D2a offers no comparable instance (P#P#PS) in which the second drop is filled by the second element of a compound, although there is a single example of P#PSS, scanned as subtype D2a (4957a magad unuuânlîc), in which the final drop as well as the penultimate one are realised by secondary-stressed syllables. These three properties seem reducible to a common structural factor, the arrangement of relative prominence on the horizontal axis. The succession of three primarystressed syllables such as would be embodied by the non-existent configuration P#P#PS would have been excluded because of the resulting prominence contour that is hardly differentiated and modulated among the constituents. On the other hand, the configuration P#PS#P would sound more tolerable as it consists of two primary stresses in succession followed by a less prominent syllable, which, realised by the second element of a quasi compound, was certainly even less pronounced than if occupied by the second element of a true compound. Nonetheless, with the occupation of a lexical stress in the third syllable, the whole contour involved would have been found not sufficiently modulated in prominence. In order to improve the situation, recourse would then have been taken to enhance prominence on the first lift by resolution and therewith to provide a greater modulation to the prominence contour involved. 25

Apparently ambiguous as it seems, verse 4329b ferid unmet grôt ought to be included in the group, given the syntax-based scansion referred to above.

MOSP-02.qxd

5/11/04

2:38 PM

Page 131

2.7. Type D

131

Bringing together all major realisation variants of type D (Table 2.80), we recognise an extended gradation pattern in their distribution which may be rearranged for the sake of exposition as follows in the decreasing order of the incidence of the a-verse with double alliteration: (73) Ranking of the major variants of type D in the decreasing order of frequency of the a-verse with double alliteration P#Px . . . S (86%) 傻 P#PS/sx (69%) 傻 P#Px . . . P (58%) 傻 P#P#P/px (43%) 傻 P#PX/xx (17%) 傻 PS/sX/xx (7%) The scale given above constitutes in essence an extension of the one provided earlier for subtype D1 (65). On average, subtypes D2a and D2b show a stronger preference for the a-verse with double alliteration than even the configuration P#PXx, which is characterised by the most extensive use of the a-verse with double alliteration among the six variants of subtype D1. Such a relative ranking may be explained in part by making a further distinction in stress properties for the drop in ways analogous to the second lift that provides a basis for distinguishing between PSX/xx and P#PX/xx. The occupation of one of the two drops by a lexical-stressed syllable correlates to a greater incidence of the a-verse with double alliteration. We are then confronted with the difference between primary and secondary stresses. As far as the variants of subtypes D2a and D2b are concerned, the configurations with a secondary-stressed syllable associated with a drop, that is, P#PS/sx and P#Px . . . S, are attended by a higher frequency of the a-verse with double alliteration than their counterparts with a primary-stressed syllable, P#P#P/px and P#Px . . . P, respectively. We may naturally wonder why the secondary-stressed syllable induces a higher frequency of the a-verse with double alliteration. By referring to a comparable distinction in the traditional metre (see below), we would rather expect the converse relation in which the primary-stressed syllable is attended by an increased incidence of the a-verse with double alliteration. The correlation that we have unexpectedly encountered, however, seems to constitute a regular feature, as we find it recurring in other metrical types as well, types D* (section 2.8) and E (section 2.9). Since a proper explanation of this phenomenon requires prior understanding on the mechanism of alliterative patterning for heavy verses (section 3.2.2), we may only briefly point out here that the multiple alliterative patterns made available exclusively to heavy verses led to attach priority to syntactic constituency in determining alliterative patterning, thereby overriding the otherwise applicable regularity based on purely metrical properties. Hardly less significant would seem another feature of the gradation: where syllable length differentiates between minimal pairs, further distinction is in evidence. In regard to the varying frequencies for the a-verse with double alliteration and conversely for the b-verse, we obtain the following rankings between the minimal pairs: PsXx 傻 Psxx; PSXx 傻 PSxx; P#PXx 傻 P#Pxx; P#P#Px 傻 P#P#px; P#Psx 傻 P#PSx. Such a group-internal articulation based on syllable length, however, would appear to lack consistency in patterning: while the short syllable correlates to a lower incidence of the a-verse with double alliteration for the four pairs, namely PsX/xx, PSX/xx,

MOSP-02.qxd

5/11/04

132

2:38 PM

Page 132

2. Metrical types and positions

P#PX/xx, and P#P#P/px, the pair with a secondary-stressed drop (P#Psx/P#PSx) shows a contrary pattern of the short syllable attended by a higher frequency of the a-verse with double alliteration. Such a reversal in correlation between syllable length and verse distribution is hardly unexpected, however; in fact, we have already encountered a similar reversal in the gradation pattern of type C in the preceding section, represented in (59) above. While the configuration x . . . PXx shows a stronger preference for the b-verse than the shorter counterpart x . . . Pxx, the pairs x . . . P#P/px and x . . . PS/sx exhibit a converse pattern of distribution by favouring the shorter members x . . . P#px and x . . . Psx, respectively, in the b-verse. What we are most concerned with here is not so much the specifics of the correlation between syllable length and verse distribution (which differ diametrically in the two cases, because distinct metrical types and contexts are involved) as with the recurring phenomenon of reversal. As argued in section 2.6.2 above, the underlying scheme of correlation is predicated on pairing the like with the like in terms of markedness: given two pairs of structural opposites to be matched with, the marked member of one pair goes with the marked one of the other. As far as the second lift of type C is concerned, it is the long syllable with lexical stress that should be characterised as marked, because this position is normally suspended from resolution, as will be substantiated in section 3.1.1.2 below. Suspension of resolution implements mapping with the level of the syllable, while resolution concerns association with the foot (Suzuki 1996a). While the long stressed syllable, bimoraic in terms of mora counting, is no different from the foot on the surface in its material extension, the short counterpart, monomoraic as it is, is differentiated from the bimoraic foot in unambiguous ways that the long one is not. The predominant occurrence of the configuration x . . . P#px as opposed to x . . . P#Px (section 3.1.4) may lend further credibility to the view that the second lift of type C is associated with the syllable node even when occupied by the long syllable. Given the three minimal pairs differentiated on the basis of the syllable length of the second lift, the shorter variants, x . . . P#px, x . . . Psx, and x . . . Pxx are identified as unmarked in opposition to the longer ones, x . . . P#Px, x . . . PSx, and x . . . PXx, respectively, according to the markedness relation thus established. With respect to verse distribution, because the type C b-verse favours the variants with the lexical-stressed second lift, the unmarked, prototypical members of these weighty variants, that is, x . . . P#px and x . . . Psx, are found most harmonious with it. Situated at the other extreme is the unmarked, minimal member of the light variants with the second lift associated with a lexically unstressed syllable, namely the sequence x . . . Pxx, which is less compatible with occurring in the b-verse than the configuration x . . . PXx, and conversely most congruent with the a-verse, which favours the light variants with the unstressed second lift. We should note that since heavy verses (i.e., verses with a heavy drop realised by a primarystressed syllable) are not at issue here for type C, and accordingly since syntactic constituency is incapable of overriding pure metrical consideration, the configurations x . . . P#px and x . . . Psx are treated exactly in parallel. A similar reasoning extends to the group-internal differentiation of type D based on its first drop. This position constitutes a normal drop regardless of its realisations, and is therefore categorically immune to resolution (section 3.1.1.4). Given that the normal drop is realised prototypically by the short unstressed syllable in word-final

MOSP-02.qxd

5/11/04

2:38 PM

Page 133

2.7. Type D

133

position, any syllables are rated as increasingly marked to the extent that they are farther removed from this prototype. In determining the relative ranking of the pair of the configurations P#Psx and P#PSx in terms of markedness, then, the shorter configuration P#Psx is characterised as less marked than the longer one P#PSx because the short stressed syllable -s- deviates less from the prototypical realisation of the normal drop -x#. As regards verse distribution, it is the a-verse with double alliteration that favours the weighty variants of type D, subtypes D2a and D2b as generalised in the scale given in (73) above. Given the binary opposed pair in question, the unmarked, typical member P#Psx is accordingly preferred and used with correspondingly greater frequency in the a-verse with double alliteration than its marked counterpart P#PSx. On the other hand, subtype D1 is privileged by the b-verse, and the variants (PSxx and P#Pxx) with the first drop closer to its prototype are favoured there accordingly, and conversely the marked counterparts (PSXx and P#PXx) occur more frequently in the a-verse with double alliteration. Apparently aberrant would seem the pair of heavy verses, P#P#px and P#P#Px; other things being equal, the short variant P#P#px, parallel to P#Psx, should be more frequent in the a-verse with double alliteration. This is precisely the point where the special character of heavy verses comes to the fore, which has to do with the complex of syntactic and phonological dimensions of alliterative patterning. As will be argued fully in section 3.2.2 below, the second lift is more compatible with lack of alliteration when the immediately following drop is realised by a short rather than a long stressed syllable on account of its lesser inherent prominence. Accordingly, it is exclusively the pair of configurations with a secondary-stressed rather than a primary-stressed drop that is affected by the reversal of ranking on the basis of contextual markedness. Subtype D2b has yet to be integrated in the scale more firmly. Specifically, we need to account for the maximal occurrence of the a-verse with double alliteration for the configuration P#Px . . . S in distinction from P#PS/sx, disregarding the latter’s group-internal finer articulation treated above. Obviously at stake here is the differing location of the drop that is realised by a lexical-stressed syllable. In this connection, we can hardly fail to recall a somewhat similar asymmetry that concerns the two configurations, PS#Px (type A1) on the one hand and Px#PS (type A2) on the other: the occupation of a lexical stress in the second drop in contrast to the first counterpart is encoded with the nearly exclusive occurrence of the a-verse with double alliteration (sections 2.1.1 and 2.2.1). The different location of the drop associated with a lexical stress thus makes a categorical distinction in metrical types. While analogous in broad terms, however, the contrast between subtypes D2a and D2b is less than categorical, and should accordingly be characterised in relative terms as variants of the same metrical type. To be sure, the configuration P#Px . . . S is used in the a-verse with double alliteration with predominant incidence, but it is also attested in the b-verse, the structural opposite of the a-verse with double alliteration. Moreover, the varying distribution of the configurations P#Px . . . S and P#Px . . . P is no different from that of the pair of P#PS/sx and P#P#P/px, which are fully integrated in the graded patterning and therefore identified as a variant of type D (subtype D2a), rather than an independent metrical type. Of further significance in determining the status of the configuration P#Px . . . S is its markedly low frequency of use in comparison with subtype D2a, which is hardly in doubt as a variant of type D by virtue of the gradation pattern that it embodies. By contrast, the configuration

MOSP-02.qxd

5/11/04

2:38 PM

Page 134

134

2. Metrical types and positions

Table 2.81. Distribution of major variants of types D1, D2a, and D2b in Beowulf variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

Psxx PsXx PSxx PSXx

0 0 0 6

0% 0% 0% 7%

0 0% 1 8% 1 50% 39 48%

2 100% 11 92% 1 50% 37 45%

PS/sX/xx

6

6%

41 42%

51

52%

P#Pxx P#PXx

7 9

6% 8%

12 10% 27 23%

105 82

85% 69%

P#Psx P#PSx

40 30

87% 73%

0 0

0% 0%

6 11

13% 27%

P#PS/sx

70

80%

0

0%

17

20%

P#P#px P#P#Px

2 100% 1 100%

0 0

0% 0%

0 0

0% 0%

P#P#P/px

3 100%

0

0%

0

0%

38 90% 33 100%

1 0

2% 0%

3 0

7% 0%

P#Px . . . S P#Px . . . P

Px#PS, which constitutes an independent metrical type (type A2), is distinguished sharply from its structural opposite PS#Px by its more extensive use, as well as its maximal polarity and lack of gradation in distribution patterning. In summary, the set of configurations Psxx, PsXx, PSxx, PSXx, P#Pxx, P#PXx, P#Psx, P#PSx, P#P#px, P#P#Px, P#Px . . . S, and P#Px . . . P all reduces to the common metrical type, type D, as its realisation variants, and they fall into three subtypes (D1, D2a, D2b) according to the prosodic properties (stress and syllable length) of language materials that occupy the drops. These variants are graded along a single scale of varying preference for the a-verse with double alliteration (or conversely the b-verse at the other extreme), as represented in (73), by the interplay of the prosodic properties of the second lift and the two drops, overridden in part by syntactic constituency as far as heavy verses are concerned. In order to appreciate fully the gradation pattern of type D in the Heliand, we need to put it in perspective by making a comparison with the situation in Beowulf. Table 2.81 indicates the distribution of the corresponding configurations in Beowulf.26 26

One might take issue with the scansion of the configurations P#P#px and P#Px . . . P as subtype A1s and type E rather than types D2a and D2b, respectively, that I proposed in Suzuki (1996a) and am assuming here as well. Yet the fact remains indisputable that there are nearly no such ambiguous instances in the a-verse. This would still demonstrate if with lesser force that Beowulf is distinguished from the Heliand by the stronger preference of the heavy verses of types D2a and D2b for the a-verse with double alliteration.

MOSP-02.qxd

5/11/04

2:38 PM

Page 135

2.7. Type D

135

To a large extent, the distribution pattern involved proves to be common to the Heliand and Beowulf: the realisation of the second lift by a short secondary-stressed syllable and that of the first drop by a short unstressed syllable (Psxx) maximises the incidence of the b-verse; the appearance of a long syllable in these two positions correlates to the decreased occurrence of the b-verse; the association of a drop with a lexical-stressed syllable favours use of the a-verse with double alliteration (P#PS/sx; P#Px . . . S), and conversely the absence of such a constituent increases the frequency of the b-verse (P#PX/xx); in the absence of lexical stress, the occurrence of the medial short syllable of a non-compound in the first drop (P#Pxx) correlates with higher incidence of the b-verse than the long counterpart (P#PXx); to the contrary, the configuration with a short stressed syllable (P#Psx) shows a stronger preference for the a-verse with double alliteration than the long counterpart (P#PSx); and as evidenced by the higher incidence of the a-verse with double alliteration for type D2b, the occurrence of a lexical stress on the final drop occasions more extensive use of the a-verse with double alliteration than that on the penultimate drop. Despite these shared features, however, the two poems differ significantly in important respects. In Beowulf the configuration PSxx is not as strongly represented in the b-verse as the configuration P#Pxx; by the same token, the configuration PSXx occurs in the b-verse with lower frequency than the configuration P#PXx. In other words, in Beowulf the varied realisation of the second lift in terms of stress properties (primary versus secondary) does not covary with verse distribution in the way it is patterned in the Heliand metre. On the other hand, the two configurations PSxx and PSXx are treated more similarly than distinctly, in contrast to their more diversified treatment in the Heliand. Correspondingly, the distinction between short and long secondary-stressed syllables has more significant consequences for verse distribution patterning in Beowulf, so that the four variants of compound-word verses, i.e., Psxx, PsXx, PSxx, PSXx, are dichotomised into the two largely polarised classes predicated on the distinction in syllable length of the second lift (PsX/xx versus PSX/xx). By contrast, these four variants are graded more evenly and continuously in the Heliand. As a consequence, the variants of type D1 in Beowulf are not organised by the same gradation pattern as in the Heliand. Obviously, the association of the second lift with a primary rather than a secondary stress does not figure as an organising principle in Beowulf. No less distinct are the specific manners of gradation that govern the variants of type D2a in Beowulf on the one hand and those of subtype D2a in the Heliand on the other. In overall terms whereby the configurations P#PSx and P#Psx are combined together, the frequency with which the a-verse with double alliteration appears with the second constituent of a compound is higher in Beowulf. Beowulf thus shows a greater degree of polarity on this score. When distinguishing between the two variants, however, a different way of polarisation emerges, whereby the short variant P#Psx is differentiated more sharply from the long counterpart in the Heliand. By contrast, much as with the pair of P#Pxx and P#PXx, Beowulf treats the configurations P#Psx and P#PSx less distinctly than the Heliand. By far the most significant difference, however, concerns the treatment of heavy verses, variants in which the first or second drop is realised by an independent word. While in the Heliand the heavy verses show a weaker preference for the a-verse with double alliteration than the configurations with a secondary-stressed syllable in the

MOSP-02.qxd

5/11/04

136

2:38 PM

Page 136

2. Metrical types and positions

corresponding position (Table 2.80), they occur in the a-verse with double alliteration with maximal incidence in Beowulf. We can thus hardly avoid concluding that in Beowulf the heavy verses do not follow a separate organising principle as in the Heliand; and therefore that in Beowulf the categorisation of types D1, D2a, and D2b is organised exclusively by a single principle based on the nature of stress. This exclusive organisation should obviously be ascribed to the uniformity of the alliterative pattern aax for heavy a-verses in Beowulf. As the foregoing comparison shows, the Heliand and Beowulf follow different principles for organising the distribution pattern of type D. Beowulf gives priority to maintaining group-internal homogeneity at the expense of the overall gradation patterning, so that the scale was divided into discrete categories, which constitute separate metrical types (i.e., types D1, D2a, and D2b). By contrast, the Heliand integrates the whole array of variants into a single hierarchy of gradation through their balanced and harmonious rearrangement along the entire scale. Consequently, the traditional group-internal integrity was disrupted, and the originally distinct metrical types, types D1, D2a, and D2b, came to be merged into a single type, type D, in the Heliand metre, with the concomitant radical remaking of their traditional distribution pattern. The resultant type D, then, had the metrical structure / / ⫻ ⫻, in which the last two positions were characterised as normal drops. Corresponding to the integration of the earlier heavy drops to the normal ones, the latter had access to an extended range of language materials for their realisation: no longer limited to syllables without lexical stress, they were allowed to be occupied by lexical-stressed syllables. In this way, as with type A1, but to a fuller extent, type D was subject to neutralisation of the opposition between heavy and normal drops, a metrical change that was motivated ultimately by the weakening of stress in Old Saxon (see sections 2.1.7.2 and 2.12.2 ).

2.8. Type D* (/ ⫻ / ⫻ ⫻) This metrical type has an extra drop added between the lifts; it is thus often referred to as an expanded variant of type D. Corresponding to the three subtypes of type D (subtypes D1, D2a, and D2b; section 2.7), we may distinguish three variants according to the differing relative prominence of the last two positions, subtypes D*1, D*2a, and D*2b, as each exemplified in (74) through (76) below (cf. Kauffmann 1887: 336–41; Hofmann 1991, II: 173–4, 180–90, 192–3, 195–9). And the distribution pattern of these subtypes including their realisation variants is shown in Table 2.82. The configurations Px . . . Psx and Px . . . P#px are apparently ambiguous, as they are scannable as type A2 and subtype D*2a alike. As will be demonstrated in section 3.1.3 below, however, there are reasons for scanning these configurations as subtype D*2a rather than type A2. Since disambiguation heavily draws on operation and suspension of resolution, it seems appropriate to treat the issue fully in the context of the metrical device involved. Verses like 2051a alloro erlo gehuilic are open to two alternative scansions (type B1 or type D*, subtype D*2b), depending on how we interpret the class 2 words involved, such as alloro. On the ambivalence of these verses, see section 3.2.2 below.

MOSP-02.qxd

5/11/04

2:38 PM

Page 137

2.8. Type D*

137

Also ambivalent would appear the configuration Px . . . Px . . . P with single alliteration, as it may be scanned as type D* (subtype D*2b) and type E*. The disambiguation has to await discussion on alliterative patterning of heavy verses, especially on the role of syntactic constituency in metrical type categorisation, as will be explored in section 3.2.2 below. Here we are simply anticipating the scansion based on syntactic constituency. (74) Subtype D*1 PxSxx 155b ôðarlîcaron Px . . . PX/xx 533a hêlag hîuuiski 2274a fêknea iungoron 3014b gôd te gifrummienne (75) Subtype D*2a Px . . . PS/sx 1007a manno mêndâdi 2400a hrosso hôfslaga 2585b barno mancunnies Px . . . PP/px 696a fulgengun godes lêrun 1180a sâlig barn godes 5974b Giuuêt imo up thanan (76) Subtype D*2b Px . . . Px . . . S 4165a alla irminthiod 2007b Gengun ambahtman Px . . . Px . . . P 2348a hôrien is hêlag uuord 5327a hreopun thuo alla samad 5775b Uuânom up astuod While type D* occurs predominantly in the a-verse with double alliteration, it is subject to gradation in fine detail of distribution. As with type D, the distribution of type D* is patterned in gradient terms depending on the two parameters, stress degrees and syllable length of the last three positions, particularly the two drops. By the first parameter of stress degrees, the configuration PxSxx, a minor variant of subtype D*1, stands out from the rest of type D* by virtue of the maximal representation of the b-verse. By the same token, the remaining variants of subtype D*1 (Px . . . Pxx and Px . . . PXx) are characterised by a weaker preference for the a-verse with double alliteration than the configurations Px . . . PS/sx (subtype D*2a) and Px . . . Px . . . S (subtype D*2b). By the second parameter of syllable length, this pair of subtype D*1 is further divided into two variants, Px . . . Pxx and Px . . . PXx, and the shorter variant Px . . . Pxx is distinguished from the longer one Px . . . PXx by its lower incidence of the a-verse with double alliteration.

MOSP-02.qxd

5/11/04

2:38 PM

Page 138

138

2. Metrical types and positions

Table 2.82. Distribution of major variants of type D* (subtypes D*1, D*2a, and D*2b) variant PxSxx

a-verse with double alliteration

a-verse with single alliteration

b-verse

0

0%

0

0%

2 100%

Px . . . Pxx Px . . . PXx Px . . . PX/xx

44 43 87

76% 90% 82%

5 0 5

9% 0% 5%

9 5 14

16% 10% 13%

Px . . . P#px Px . . . P#Px Px . . . P#P/px

54 64% 19 100% 73 71%

12 14% 0 0% 12 12%

19 0 17

23% 0% 17%

Px . . . Psx Px . . . PSx Px . . . PS/sx

147 146 293

98% 97% 97%

1 1 2

1% 1% 1%

2 4 6

1% 3% 2%

Px . . . Px . . . P Px . . . Px . . . S

113 30

88% 97%

3 0

2% 0%

13 1

10% 3%

PxSxx b-verse: 155b, 3123b Px . . . Pxx a-verse with double alliteration: 42a, 304a, 323a, 342a, 371a, 810a, 973a, 1134a, 1149a, 1261a, 1291a, 1592a, 1992a, 2000a, 2125a, 2285a, 2315a, 2413a, 2567a, 2696a, 2864a, 2984a, 3119a, 3176a, 3224a, 3238a, 3633a, 3769a, 4001a, 4338a, 4400a, 4414a, 4420a, 4499a, 4529a, 4572a, 4635a, 4804a, 4899a, 4938a, 5042a, 5127a, 5539a, 5634a; a-verse with single alliteration: 940a, 2274a, 2815a, 4371a, 5905a; b-verse: 308b, 513b, 1191b, 1309b, 2131b, 2541b, 2555b, 3526b, 5118b Px . . . PXx a-verse with double alliteration: 282a, 533a, 564a, 694a, 760a, 781a, 860a, 935a, 1011a, 1063a, 1266a, 1460a, 1599a, 2095a, 2546a, 2572a, 2575a, 2659a, 2806a, 2810a, 2996a, 3299a, 3310a, 3414a, 3441a, 4437a, 4525a, 4663a, 4771a, 4802a, 4831a, 4881a, 5030a, 5118a, 5271a, 5649a, 5887a, 5941a, 5947a, 5958a, 5963a, 5965a, 5982a; b-verse: 438b, 821b, 2017b, 3014b, 4593b Px . . . P#px a-verse with double alliteration: 49a, 193a, 276a, 399a, 598a, 648a, 895a, 1084a, 1282a, 1384a, 1400a, 1412a, 1485a, 1581a, 1642a, 1996a, 2122a, 2176a, 2251a, 2298a, 2355a, 2371a, 2403a, 2666a, 2847a, 2908a, 2937a, 2942a, 2943a, 3024a, 3070a, 3138a, 3248a, 3911a, 4011a, 4089a, 4115a, 4259a, 4270a, 4374a, 4722a, 4914a, 5084a, 5089a, 5133a, 5171a, 5332a, 5341a, 5395a, 5869a, 5882a, 5894a, 5946a, 5962a; a-verse with single alliteration: 7a, 518a, 798a, 1180a, 1440a, 2024a, 2038a, 2133a, 2325a, 2520a, 2975a, 5509a; b-verse: 400b, 812b, 832b, 847b, 1121b, 1165b, 1916b, 2172b, 2533b, 2975b, 3111b, 3293b, 3895b, 4796b, 4798b, 5093b, 5159b, 5974b, 5976b Px . . . P#Px a-verse with double alliteration: 696a, 1538a, 1726a, 1908a, 1921a, 1960a, 2269a, 2313a, 2644a, 2821a, 2946a, 2948a, 3277a, 3603a, 3919a, 4441a, 4451a, 4755a, 5655a Px . . . Psx a-verse with double alliteration: 229a, 301a, 341a, 346a, 360a, 378a, 386a, 411a, 424a, 454a, 473a, 480a, 535a, 627a, 651a, 668a, 672a, 679a, 691a, 727a, 742a, 765a, 871a, 980a, 1062a, 1105a, 1115a, 1129a, 1176a, 1191a, 1228a, 1240a, 1274a, 1322a, 1349a, 1359a, 1441a, 1530a, 1544a, 1646a, 1652a, 1661a, 1669a, 1722a, 1751a, 1783a, 1859a, 1860a, 1886a, 1896a, 1961a, 1976a, 1981a, 2004a, 2009a, 2063a, 2174a, 2181a, 2229a, 2300a, 2317a, 2400a, 2515a, 2543a, 2653a, 2730a, 2741a, 2766a, 2776a, 2780a, 2796a, 2855a, 2876a, 2886a, 2888a, 2889a, 2896a, 2899a, 2924a, 2926a, 2938a, 3053a, 3116a, 3126a, 3143a, 3335a, 3476a, 3536a, 3686a, 3687a, 3698a, 3800a, 3856a, 3943a, 3948a, 4033a, 4127a, 4190a, 4215a, 4242a, 4347a, 4443a, 4553a, 4634a,

MOSP-02.qxd

5/11/04

2:38 PM

Page 139

2.8. Type D*

139

4642a, 4662a, 4665a, 4743a, 4745a, 4759a, 4762a, 4766a, 4783a, 4812a, 4813a, 4897a, 4913a, 4944a, 4971a, 4981a, 5068a, 5125a, 5190a, 5195a, 5213a, 5242a, 5276a, 5303a, 5325a, 5338a, 5339a, 5367a, 5368a, 5497a, 5517a, 5549a, 5550a, 5637a, 5643a, 5646a, 5658a, 5708a, 5734a, 5747a, 5774a, 5798a, 5803a; a-verse with single alliteration: 3639a; b-verse: 3270b, 4753b Px . . . PSx a-verse with double alliteration: 55a, 60a, 109a, 121a, 172a, 233a, 251a, 258a, 274a, 300a, 381a, 488a, 494a, 507a, 521a, 526a, 565a, 581a, 637a, 659a, 796a, 883a, 902a, 924a, 967a, 974a, 1007a, 1057a, 1064a, 1082a, 1102a, 1119a, 1126a, 1141a, 1178a, 1214a, 1302a, 1340a, 1352a, 1361a, 1364a, 1381a, 1388a, 1444a, 1496a, 1504a, 1515a, 1526a, 1546a, 1574a, 1609a, 1620a, 1625a, 1631a, 1648a, 1756a, 1807a, 1842a, 1854a, 1870a, 1919a, 1949a, 1970a, 1989a, 2033a, 2081a, 2115a, 2121a, 2238a, 2264a, 2361a, 2395a, 2396a, 2447a, 2542a, 2583a, 2599a, 2640a, 2667a, 2748a, 2768a, 2774a, 2826a, 2913a, 2929a, 2994a, 3008a, 3017a, 3029a, 3040a, 3043a, 3045a, 3154a, 3221a, 3230a, 3253a, 3255a, 3297a, 3345a, 3363a, 3417a, 3479a, 3504a, 3537a, 3541a, 3602a, 3637a, 3704a, 3751a, 3764a, 3797a, 3879a, 3992a, 4003a, 4119a, 4123a, 4207a, 4221a, 4254a, 4352a, 4354a, 4402a, 4497a, 4500a, 4507a, 4698a, 4917a, 4968a, 5101a, 5173a, 5296a, 5364a, 5375a, 5413a, 5502a, 5721a, 5735a, 5742a, 5748a, 5797a, 5831a, 5864a, 5909a, 5935a, 5951a, 5967a; a-verse with single alliteration: 4973a; b-verse: 690b, 2168b, 2585b, 2957b Px . . . Px . . . P a-verse with double alliteration: 9a, 15a, 198a, 347a, 372a, 385a, 414a, 440a, 571a, 578a, 657a, 717a, 734a, 743a, 763a, 801a, 855a, 946a, 961a, 985a, 1005a, 1059a, 1107a, 1144a, 1199a, 1285a, 1348a, 1382a, 1451a, 1536a, 1540a, 1730a, 1748a, 1752a, 1795a, 1810a, 1826a, 1850a, 1914a, 2051a, 2235a, 2243a, 2279a, 2284a, 2348a, 2357a, 2444a, 2528a, 2553a, 2556a, 2576a, 2581a, 2594a, 2658a, 2853a, 2856a, 3005a, 3046a, 3099a, 3129a, 3144a, 3172a, 3344a, 3349a, 3359a, 3425a, 3432a, 3445a, 3482a, 3577a, 3617a, 3644a, 3646a, 3685a, 3711a, 3737a, 3802a, 3994a, 4090a, 4129a, 4194a, 4267a, 4285a, 4314a, 4315a, 4319a, 4325a, 4379a, 4381a, 4552a, 4633a, 4751a, 4819a, 4836a, 4852a, 4916a, 4972a, 5094a, 5158a, 5214a, 5231a, 5298a, 5420a, 5458a, 5473a, 5511a, 5561a, 5654a, 5678a, 5690a, 5782a, 5806a, 5973a; a-verse with single alliteration: 1085a, 4867a, 5327a; b-verse: 112b, 1134b, 2036b, 2161b, 2467b, 3228b, 3295b, 3822b, 3998b, 4103b, 5000b, 5755b, 5775b Px . . . Px . . . S a-verse with double alliteration: 329a, 340a, 365a, 531a, 629a, 725a, 764a, 1114a, 1186a, 1200a, 1324a, 1904a, 2190a, 3169a, 3235a, 3426a, 3825a, 4068a, 4165a, 4309a, 4419a, 4650a, 4661a, 4729a, 4810a, 4856a, 5064a, 5343a, 5466a, 5499a; b-verse: 2007b

As with type D (section 2.7), heavy verses of type D* (subtypes D*2a and D*2b) are differentiated from those equipped with a secondary-stressed drop by their weaker preference for the a-verse with double alliteration. Particularly noteworthy is the lowest incidence of the a-verse with double alliteration of the configuration Px . . . P#px, again comparable to the type D2a counterpart P#P#px (Table 2.80). Such weaker preference of heavy verses in general, and of the configuration with a short stressed syllable (Px . . . P#px) in particular, for the a-verse with double alliteration is a consequence of the enriched variety of alliterative patterns accessible to heavy verses in the Heliand, in conjunction with the minimal prominence of the short syllable, as pointed out in section 2.7 above and discussed fully in section 3.2.2 below. Overall, type D* is subject to graded patterning of distribution to a lesser extent than type D. Indeed, the configurations Px . . . PX/xx and Px . . . P#P/px show a gradation pattern, differentiated on the group-internal as well as group-external dimensions, in the manners largely parallel to subtype D1, yet the distinction observed is far less conspicuous; moreover, the configurations Px . . . PS/sx are scarcely involved in gradation. Type D* is thus distinguished from type D by the greater degree of homogeneity in distribution among its members.

MOSP-02.qxd

5/11/04

140

2:38 PM

Page 140

2. Metrical types and positions

This difference then brings us to a more outstanding feature about the distribution of type D* in comparison with that of type D (section 2.7): type D* appears in the a-verse and realises double alliteration with much greater frequency than type D. The three subtypes of type D* used as a-verses with double alliteration are far more frequent than their corresponding subtypes of type D; and even subtype D*1, the subtype of type D* with the lowest incidence of the a-verse with double alliteration, occurs in the a-verse with double alliteration with higher frequency than subtype D2a, which is in turn distinguished from subtype D1 by the marked preference for the a-verse with double alliteration (Table 2.80). No less notable is the wide difference between types D* and D in terms of the number of occurrences, and the relative frequency, of their corresponding subtypes, as indicated in Table 2.83. As should be clear from Table 2.83, the proportion of type D* to type D is so high that it would seem questionable to reduce type D* to type D as its marked variant. Of particular importance is the conspicuously higher proportion of subtype D*1 to subtype D1 in the Heliand than in Beowulf, in which subtype D*1 is extremely rare (Suzuki 1996a: 104): in the Heliand, subtype D*1 is only slightly lower than the half of the total of subtypes D1 and D*1. The remaining two subtypes D*2a and D*2b also are attested in the Heliand far more frequently than in Beowulf; of more significance still is that these two variants outnumber their corresponding subtypes D2a and D2b with a wide margin and thereby reverse the original proportion obtained in Beowulf. In view of the radically increased incidence with which all the subtypes D*1, D*2a, and D*2b appear in the Heliand, as well as of the overwhelming predominance of D*2a and D*2b over D2a and D2b, it would seem inappropriate to characterise them as derived marked variants. Rather than reduced to type D as in the traditional metre, the so-called expanded variant should accordingly be postulated as an independent metrical type, type D*, in the Heliand metre. The proposed analysis is supported also by the reverse preference patterns for the a-verse with double alliteration exhibited by the configuration PPxx on the one hand and the configuration Px . . . Pxx on the other, as shown in Tables 2.80 and 2.82 Table 2.83. Proportion of subtypes D*1, D*2a, D*2b to subtypes D1, D2a, D2b, respectively subtype

number

percentage

D1 D*1 total

157 108 265

59% 41% 100%

D2a D*2a total

101 405 506

20% 80% 100%

D2b D*2b total

86 160 246

35% 65% 100%

MOSP-02.qxd

5/11/04

2:38 PM

Page 141

2.8. Type D*

141

above. Should these two configurations be surface variations of a common metrical type, one would expect them to be treated similarly as regards their distribution. Of further significance is the frequent occurrence of verses of the configuration Px . . . Pxx in which the syllables associated with the first drop are closely bound with the following rather than the preceding stressed word: e.g., 3633a uueros auuahsane. This property characteristic of the Heliand in contrast to Beowulf would imply that the derivation of variant D* from type D through the association of the first lift with the head of a prosodic word as proposed for Beowulf (Suzuki 1996a: 24–35) does not apply to the Heliand, and thereby lend further credibility to the independent status of type D*. The extra drop of type D* then constitutes an autonomous (though optional) position of its own in the underlying metrical structure, while the corresponding language materials used in Beowulf are associated with the first lift of variant D* as its surface increments by encliticisation. Type D* in the Heliand is distinguished from its corresponding configuration in Beowulf also in regard to the variety of language material that may occupy the first drop. The variety of material used is summarised in Tables 2.84 and 2.85 (for the Heliand) on the one hand, and Tables 2.86 and 2.87 (for Beowulf) on the other, in terms of the quantity of syllable concatenations involved in association with the position in question. Type D* in the Heliand differs from its structural analogue, subtype D* in Beowulf in terms of the greater variety of its first drop, both quantitatively and qualitatively. The position may be filled by a much larger number of syllables in the Heliand than in Beowulf, as testified in the following tables. While the maximal number of syllables involved is three in Beowulf (Beo 473a Sorh is me to secganne; Suzuki 1996a: 29), it amounts to nine in the Heliand: 5195a antfâhad ina than eft under iuuue folcscepi. Indeed, the association with a single syllable is still the most common in the Heliand as in Beowulf, yet the range of numerical variability is markedly extended, and the whole distribution consequently becomes diffused with a less prominent concentration on the single syllable. Correspondingly, the use of a single syllable accounts for only slightly over the half of all occurrences of type D* in the Heliand (Table 2.85), whereas it constitutes an overwhelming majority in Beowulf (Table 2.87). Table 2.84. The varying number of syllables in the first drop of type D* number of syllables

a-verse with double alliteration

a-verse with single alliteration

b-verse

1 2 3 4 5 6 7 8 9

302 143 84 48 16 3 3 0 1

13 7 1 1 0 0 0 0 0

35 10% 13 8% 5 6% 2 4% 0 0% 0 0% 0 0% 0 0% 0 0%

86% 88% 93% 94% 100% 100% 100% 0% 100%

4% 4% 1% 2% 0% 0% 0% 0% 0%

MOSP-02.qxd

5/11/04

2:38 PM

Page 142

142

2. Metrical types and positions

Table 2.85. Distribution of type D* according to the number of syllables in the first drop number of syllables

number of all verses

percentage

1 2 3 4 5 6 7 8 9

350 163 90 51 16 3 3 0 1

52% 24% 13% 8% 2% less than 1% less than 1% 0% less than 1%

total

677

100%

Table 2.86. The varying number of syllables in the first drop of subtype D* in Beowulf number of syllables

a-verse with double alliteration

a-verse with single alliteration

b-verse

1 2 3

109 71% 3 60% 1 50%

33 22% 1 20% 0 0%

11 7% 1 20% 1 50%

Table 2.87. Distribution of subtype D* according to the number of syllables in the first drop in Beowulf number of syllables

number of all verses

percentage

1 2 3

153 5 2

96% 3% 1%

total

160

100%

A further notable point is that the position in question is occupied by more varied language material in the Heliand. Of special interest about individual occupants is that, as briefly noted above, it may be filled by a prefix, a phenomenon which is totally unknown in Beowulf (Suzuki 1996a: 34). The traditional avoidance of prefixes is a consequence of the position at issue being derivative due to encliticisation. The association with a prefix may thus show that contrary to the traditional metre encliticisation is not crucially involved in the composition of the first drop of type D* in the Heliand. Yet the proportion of prefixes in the metrical configuration in question (P#xPxx) is extremely low: there are only six such examples (2243a suang gisuerc an gimang, 2853a sat gesîði mikil, 3633a, 5101a, 308b, 5755b) out of the total of 350, that is,

MOSP-02.qxd

5/11/04

2:38 PM

Page 143

2.8. Type D*

143

c. 2 per cent of all occurrences of type D* with the first drop filled by a single syllable. This figure seems markedly lower than that for the configuration P#xPx (a variant of type A1), for example, for which 146 instances or c. 9 per cent out of a total of 1541 occurrences of the configuration PxPx realise a prefix. The predominance of word finals as an occupant of the second position of type D* would suggest that the Heliand did not totally subvert the traditional practice of limiting the position to the least prominent material, despite the notable expansion that the work allowed in size and variety of its surface manifestations. The traditional restriction on the position in question for its realisations is reflected further in the fact that it is rarely occupied by a lexical-stressed syllable in the Heliand as well, in particular the second element of a true compound or an independent word: we find only fourteen examples in which the first drop is occupied by such a relatively strong syllable including a heavy derivational suffix such as -lîc: 121a anduuard for them alouualdon; 424a blîðlîc bodskepi; 961a and 1005a diurlîc drohtines suno; 1592a diurlîc dôperi; 2181a ênan lîflôsan lîchamon; 2285a gôd uuerk mid is iungeron; 2515a sô mislîcan muodsebon; 3046a and 5806a diurlîc drohtines bodo; 3994a diurlîc drohtines thegan; 4215a hetelîc hardburi; 5042a breosthugi blôðora; 2975b Elithioda quam imu. Put another way, the sequences PS(x)#(x) and P#P(x)#(x) are only exceptionally allowed to realise the first two positions of type D* (/ ⫻), a generalisation that was in force Beowulf (Suzuki 1996a: 110). Of particular interest, the vast majority of these examples contain heavy derivational suffixes like -lîc and -lôs, rather than full lexical elements like uuerk, -hugi. Such predominance of -lîc and the like also constitutes an inherited property of the traditional metre (Suzuki 1996a: 110). Of further significance is the complete absence of the string PSx#x as a realisation of the first two positions, although other strings such as PS#, PSx#, PS#x . . . are attested. As may be recalled, the sequence -Sx# is least likely to be associated with a normal drop: in contrast to the sequences -S#x and -Xx#, the string -Sx# may not realise the second drop of type B1 (section 2.4.2.3). In this light, the string PSx#x, which is expanded by a word-initial syllable #x, would have been deemed as too prominent and hence too disrupting to the traditional practice due to its extra prominence involved. By contrast, the sequence -Xx#x . . . counts as sufficiently light so as not to contravene the traditional convention of avoiding relatively prominent material from the first drop of type D*. The string is thus permitted to occur, as in 3344a fan themu hêroston, the thes hûses giuueld, 3414a the hêrosto thes hîuuiskeas, 3441a thie hêrosto them hîuuiskes, 4374a sô fârungo uuarð that fiur kemen, 4663a ik fullêstiu iu uuiðer themu fîunde, 5946a gegnungo fan themo godes suno. The absence of lexical stress thus distinguishes -X- from -S-, precisely as in their differential treatment for the first drop of type A1 (section 2.1.7) and the second drop of type B1 (section 2.4.2.3), for example. Accordingly, we may generalise that the extra position between the lifts in type D* constitutes a normal drop that is hardly amenable to association with a lexicalstressed syllable, a restriction that owes to traditional practice. In this light, we can hardly avoid the conclusion that the first drop of type D* was allowed access to a minimally enlarged range of prominence for its surface realisations, although its underlying status was radically reorganised in the system.

MOSP-02.qxd

5/11/04

144

2:38 PM

Page 144

2. Metrical types and positions

Of no less significance, type D* in the Heliand stands out by its graded pattern of distribution along the parameter of the number of syllables that occupy the position under consideration: as the number of syllables involved increases, the frequency of verses gradually decreases, yet the relative proportion of the a-verse with double alliteration increases (Table 2.84). In Beowulf, by contrast, verses with more than a single syllable for the first drop of subtype D* are extremely rare, so much so that we find a sharp dichotomy in force, whereby a vast majority have a single syllable and the exceptional remainder contain two or three syllables (Table 2.87). Moreover, the distribution of these exceptional verses seems hardly patterned in Beowulf: they appear in the a-verse (with double or single alliteration), as well as in the b-verse (Table 2.86); and the increased number of syllables does not appear to induce an enhanced frequency of the a-verse with double alliteration contrary to what we would expect on the basis of the general pattern observed elsewhere (e.g., the first drop of type A1). Finally, it may be as well to recall that there are a number of type A1 verses which might at first glance appear to be instances of type D*. They are of the configuration Px . . . Pxx with a disyllabic ending (-xx#) in verse-final position as discussed in section 2.2.2 above. We may reproduce representative examples below: (77) 319a thiornun thînaro 936a drohtine mînumu 2223a haltaro endi hâbaro The rarity of the a-verse with double alliteration thus indicates that these sequences ending in a disyllabic ending should be scanned as type A1 rather than type D*: as demonstrated above, type D* is characterised by its strong preference for the a-verse with double alliteration. This would imply that the disyllabic ending is treated as different from the normal disyllabic sequence of a short unstressed syllable and another unstressed syllable -xx#, which constitutes the last two positions of type D*, as in 342a cuman fon them kêsure. More specifically, the disyllabic ending does not count as heavy enough to be provided with two separate positions. In section 2.1.4 above, however, we have seen that the disyllabic ending is treated as more prominent than the sequence -xx# in verse internal position: when mapped with the first drop of type A1, the ending in question is distinguished from the sequence -xx# by a stronger preference for the a-verse with double alliteration. The varied metrical treatment of the disyllabic ending -xx# will be treated fully in section 2.12.4 below. In conclusion, type D* (/ ⫻ / ⫻ ⫻), consisting of three subtypes, D*1 (PxSxx, Px . . . PX/xx), D*2a (Px . . . PS/sx; Px . . . P#P/px), and D*2b (Px . . . Px . . . S; Px . . . Px . . . P), should be identified as an autonomous metrical type irreducible to type D (/ / ⫻ ⫻) with three subtypes, D1 (PS/sX/xx, P#PX/xx), D2a (P#PS/sx; Table 2.88 (⫽ Table 2.42). Distribution of Px . . . Pxx a-verse with double alliteration

a-verse with single alliteration

b-verse

8 (22%)

15 (42%)

13 (36%)

MOSP-02.qxd

5/11/04

2:38 PM

Page 145

2.9. Type E

145

P#P#P/px), and D2b (P#Px . . . S; P#Px . . . P); in other words, this type is no longer a realisation variant of type D as in the traditional metre. Accordingly, the Heliand has an independent verse form at its disposal which consists of five metrical positions.

2.9. Type E (/ \ ⫻ /) This type comprises five variants which are distinguished according to the morphological status of the word that fills the second position (first drop; represented as \ for the time being, to be fully justified in due course). First, the position may be occupied by a medial or final syllable of a non-compound word. A vast majority of relevant instances have this position filled by a medial rather than a final syllable (PXx(x)#(x)P), as exemplified in (78) and listed under Table 2.89 below (cf. Kauffmann 1887: 341–5; Hofmann 1991, II: 201–5, 207–8): (78) 575a uualdandes uuord 3079a brinnandi fiur 4708b hêlagna gêst By contrast, we have only a small number of examples in which the position is associated with a word-final syllable (PX/x#x . . . P), as illustrated in (79) and listed under Table 2.89 below (cf. Hofmann 1991: 71; 1991, II: 200–1): (79) 39a uualdand gisprak 279a scadouuan mid skimon 1526b ôðres ni uili As far as the unstressed word-final syllable is concerned, there seems to be no restriction imposed on its length for filling the position in question: either a long or a short syllable is found associated with it. Thus, disyllabic simplex words Px# (e.g., 1526b) as well as PX# (e.g., 39a) may realise the first two positions of type E (/ \) in the Heliand as marginally as in Beowulf (Suzuki 1996a: 120). As regards the association of the same position with an unstressed medial syllable, however, syllable length seems to play a vital role: there are no instances in which the position is occupied by a single short syllable. Put another way, when a trisyllabic noncompound word is mapped onto the first three positions, the second position invariably corresponds with a long syllable. Thus, the first three positions of type E (/ \ ⫻) may be realised by the non-compound #PXx#, but not by the non-compound form #Pxx# such as drohtines, folgoda, hêtana, as pointed out in section 2.4.2.3 above. Apparently exceptional to this generalisation are the following verses: (80) 610a 929a 2804a 4915a

craftagoron cuning ênhuuilic ni bist craftigana Crist hatandiero hôp

MOSP-02.qxd

5/11/04

146

2:38 PM

Page 146

2. Metrical types and positions

Verse 4915a would appear doubly anomalous. First, this is the only example in which the two internal drops of type E (\ ⫻) are realised by a single short unstressed syllable each. Second, the two syllables involved constitute a disyllabic ending -ero#, which is normally treated as an inseparable unit. Following Hofmann (1991: 54), however, we may read hettendero, a variant form, for hatandiero in accordance with Manuscript C (the variant hetteand is also attested elsewhere in Manuscript M, as in 4658a hetteand herugrim). The verse in question would then be an instance of the well-attested verse form PXxx#P, as exemplified below: (81) 2031a hêleandoro bezt 5858a hettiandero hand 2635b menniscono barn a-verse with double alliteration: 5858a; a-verse with single alliteration: 2031a, 2180a, 3061a, 3558a, 4031a, 5218a, 5601a; b-verse: 2635b, 2678b, 2811b, 3156b, 3606b, 3630b, 4036b We are then left with the three exceptions (610a, 929a, 2804a). Common to these three is that there are three short unstressed syllables between the lifts. Still valid thus remains the observation that no genuine example is found of the sequence Pxx#P that would unambiguously demonstrate that the second position is associated with a single medial short syllable. We may accordingly assume that the above three verses receive resolution on the second position whereby it is mapped onto a foot governing a disyllable (e.g., -tago- in verse 610a), a metrical practice independently carried out as in 619a friðugumono bezt, 2993a uuamscaðon biuueri, 2700b langsamane râd (see section 3.1.1.5 below).27 Hofmann, who excludes resolution from secondary-stressed syllables – except for the word form Pxsx (Hofmann 1991: 69) –, would object to scanning these verses as resolved on the second position. Hofmann’s interpretation, however, would seem untenable, as will be criticised in section 3.1.1.5 below. One might object, further, that the proposed analysis based on resolution would be confronted with the difficulty of scanning verse 2804a: it would destroy the morphological integrity of the disyllabic ending involved by aligning its constituents with separate positions. Given craftigana, the two medial syllables -tigawould be associated with the second position, and the final syllable -na with the third. As may be recalled, however, a disyllabic ending behaves as a unit when associated with a metrical position. Furthermore, there is reason to believe that resolution normally affects lexical-stressed syllables (section 3.1.1.6). Thus, one might argue, to appeal to resolution in accounting for the anomalous examples at issue must be rejected as a misguided practice; rather, the proper scansion would have to align a single short unstressed syllable with the second position and the following two syllables with the third counterpart in accordance with the integrity of the disyllabic endings involved. There are several difficulties with the above alternative scansion. First, verses 610a and 929a do not have a disyllabic ending. The two verses therefore have to be 27

Verse 929a should therefore be characterised as equivalent to the configuration PX/x#xP, rather than Pxx#P.

MOSP-02.qxd

5/11/04

2:38 PM

Page 147

2.9. Type E

147

scanned by mapping the first two unstressed syllables and the third onto the second and the third positions, respectively. Otherwise, the first drop would be less prominent than the second, and the unmetrical configuration / ⫻ \ / would result accordingly. Even if one would scan these verses by aligning them to the metrical form / ⫻ ⫻ / rather than / \ ⫻ /, a problem would remain unsolved: in the light of the prominence profile specified in the Linearity-based Metrical Strength Scale (section 1.2), a drop closer to the beginning of a verse counts as more prominent. Accordingly, to associate only the first of the three unstressed syllables with the first drop would violate this generalisation. Second, the scansion under criticism here would have to leave the following question unanswered: what is marginally allowed (by whatever mechanism, to be specified below) as a realisation of the first three positions of type E is the sequence #Pxxx#, rather than #Pxx#. According to the alternative scansion without resolution, however, the sequence #Pxx# would constitute a better match with the metrical contour / \ ⫻ (or / ⫻ ⫻) than the sequence #Pxxx#, partly because the disyllabic ending involved counts as more prominent than the preceding short syllable, and partly because the disyllabic ending is generally regarded as marked in opposition to the monosyllabic one. Hence, one may be led to predict that of the two candidates at issue the word form #Pxx# would be selected as more fitting the metrical template contrary to the fact. By recourse to resolution, however, we may surmount the above difficulty by accounting for the preference for the sequence #Pxxx# including #Pxxx# over #Pxx#. Indeed, resolution normally affects lexical-stressed syllables. Yet this restriction derives from the fact that resolution is a metrical phenomenon occurring on strong positions, prototypically lifts, but under certain circumstances heavy drops as well (section 3.1.1.6). Since these positions are typically occupied by lexical-stressed syllables, the generalisation obtains that resolution applies to stressed syllables to the exclusion of unstressed counterparts, which are usually disallowed from such strong positions. In marked instances where special conditions obtain, however, strong positions may be filled by unstressed syllables, as may the second lift of type C (x. . . PX/xx; e.g., 82a uuârun sô gihôriga; section 2.6.2). By the same token, the second position of type E may occasionally be made susceptible to association with lexically unstressed syllables. Obviously, long syllables are chosen for the most part due to their inherent greater prominence, but short syllables may as well even if marginally in the extreme, in particular when immediately followed by a word boundary (see below). Contingent on the realisation of a heavy drop by short syllables without lexical stress, resolution comes into play as a viable metrical device for association: because the heavy drop is prototypically filled by a foot, it is as capable of being associated with a sequence of two short unstressed syllables that is organised into a foot medially or word-finally (section 1.3.1; for details on the mechanism of foot construction, see section 2.12.4 below). Now given the two sequences #Pxx# and #Pxxx# under consideration, and the stronger preference for implementation of resolution than for suspension of it in the metrical context in question (section 3.1.1.6), to invoke resolution would choose #Pxxx# over #Pxx# as a better match for the metrical target / \ ⫻: while #Pxx# is incompatible with the implementation of resolution on the heavy drop that would result in the association with two positions (/ \) rather than the required three (/ \ ⫻), the sequence #Pxxx#

MOSP-02.qxd

5/11/04

148

2:38 PM

Page 148

2. Metrical types and positions

is resolvable by mapping the medial disyllable -xx- onto the second position (\), thereby leaving the word-final syllable for association with the third position (⫻). Thus, by associating the footed medial disyllable with the heavy drop, the resulting overall contour would match the underlying template by virtue of the markedly greater prominence on the second position than on the third, which is filled by a single unstressed syllable. The exceptional mechanism for the association of the heavy drop with short unstressed syllables, however, sometimes incurs undesirable consequences: it does violence to the otherwise respected morphological unity of disyllabic endings as noted above. Such a potential disintegration of the morphological unity involved may be held responsible for the contrastive treatment of the sequence #Pxxx# in types A1 and E: this string is attested more frequently in type A1 (the first two positions / ⫻) than in type E (the first three positions / \ ⫻); and it also occurs more frequently than the string #PXxx# appearing in the identical positions of type A1 (section 2.1.4); by contrast, the same string #Pxxx# occurs only once in type E, the frequency significantly lower than the string #PXxx# in the same metrical type, as pointed out and listed in (81) above. Since the association of the string -xxx# with the second position of type A1 leaves the integrity of the disyllabic ending intact, it would not meet with the kind of resistance that it provokes in type E. In any event, the alignment of the second position of type E with short unstressed syllables is a highly marginal process because the process involved is far removed from the prototypical association of the heavy drop with a lexical-stressed syllable, and also because it may possibly come into conflict with otherwise valid metrical generalisations, particularly the metrical respect of the morphological integrity.28 Thus, while the configuration Px#xP as well as PX#xP scans as an acceptable (if marginal) type E verse, the sequence Pxx#P is totally unattested, in contrast to PXx#P, which is fully metrical. The differential treatment between Px#xP on the one hand and Pxx#P on the other may be related to what we have seen differentiating between Px#xPx and Pxx#Px for type A1 (section 2.1.3). That is, the string -x#x counts as more prominent than -xx# with the reverse order of x# by showing a stronger preference for the a-verse with double alliteration, when used as the first drop of type A1. Analogously, we may assume that the greater prominence of the string -x#x may make it more feasible for it to be provided with two positions than the less prominent counterpart -xx#, which would be more likely to be subsumed under a single position by virtue of its minimal prominence. Thus, the minimally prominent string Pxx#, in contrast to the more prominent one Px#x, is excluded from association with the first three positions of type E (/ \ ⫻). A second subtype of type E has the first drop occupied by the second element of a compound, and, in parallel to the first subtype treated above, we may distinguish two variants according to the composition of the second position, namely whether it is realised by a medial (PSx(x)#(x . . .)P) or final syllable (PS#x . . . P). As with the non-compound form #PXx#, the string #PSx# constitutes an overwhelming majority, 28

A similar explanation may extend to the comparable rarity of the sequence #Pxxx# as realisations of the last three positions of types D and D*. We find only a single example for each: 2689b he ni uuas iro cûð ênigumu (type D); 4593b Ni uuas thero thegno ênigumu (type D*).

MOSP-02.qxd

5/11/04

2:38 PM

Page 149

2.9. Type E

149

as shown in Table 2.89 below (cf. Kauffmann 1887: 341–5; Hofmann 1991, II: 201–9): (82) 1133a mancunneas mên 1448b niudlîco scal 3694b slîðmôde man The other string, #PS#, is used with low frequency (cf. Table 2.89), much as is the non-compound counterpart, #PX/x# (cf. Hofmann 1991, II: 200–1): (83) 167a kindiung giboran 2630a fisknet an flôd 3174b sorgspell ni forhal In a third subtype of type E (P#Px(x)#(x)P and P#P#x . . . P), the heavy drop is filled by an independent word, as exemplified below (cf. Kauffmann 1887: 341–5; Hofmann 1991, II: 200–1): (84) 1323a godes uuang forgeben 1774a uuîd strâta endi brêd 3902b thes godes barnes uuord Purely in terms of metrical configuration, type E heavy verses with single alliteration would appear scannable as subtype D2b as well, as noted in section 2.7 above. The catalogue given under Table 2.89 presupposes the syntax-based scansion that will be justified in section 3.2.2 below, after we have fully discussed alliterative patterning of heavy verses in general. By way of comparison, Beowulf offers four examples of the heavy variant of type E with double alliteration in which the first drop is occupied by an independent lexical word (Suzuki 1996a: 113): (85) Beo 147a Beo 545a Beo 658a Beo 1395a

twelf wintra tid fif nihta fyrst Hafa nu ond geheald Ðys dogor þu

Thus, the Heliand contains a comparatively large number of the variant of type E (fortyfive out of a total of 428, c. 11 per cent, as against four out of a total of 141, c. 3 per cent in Beowulf) in which the second position is associated with an independent word. The distribution pattern of the major variants of type E that we have treated above is shown in Table 2.89. With regard to the correlation between verse distribution and the variety of type E verses used in the Heliand, the sequences -Sx#, -S#x, #Px#, and #P#x show far stronger preference for the a-verse with double alliteration than the sequences -Xx# and -X/x#x. We may accordingly generalise that the association of the second position with a syllable without lexical stress (X/x) favours occurrence of b-verses; by contrast, the realisation of the same position by a lexical-stressed syllable (S or P)

MOSP-02.qxd

5/11/04

2:38 PM

150

Page 150

2. Metrical types and positions

Table 2.89. Distribution of major variants of type E variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

PX/x#x . . . P PXx(x)#(x)P P#Px(x)#(x)P PSx(x)#(x . . .)P PS#x . . . P P#P#x . . . P

1 25 10 109 12 14

3 39 6 53 3 1

4 59 12 69 2 2

12.5% 20% 36% 47% 71% 82%

37.5% 32% 21% 23% 18% 6%

50% 48% 43% 30% 12% 12%

PX/x#x . . . P a-verse with double alliteration: 279a; a-verse with single alliteration: 39a, 929a, 2330a; b-verse: 1526b, 2482b, 3570b, 4291b PXx(x)#(x)P a-verse with double alliteration: 575a, 610a, 689a, 1047a, 1357a, 1450a, 2196a, 2259a, 2331a, 2398a, 2804a, 3004a, 3789a, 4009a, 4588a, 4592a, 4814a, 4915a, 4931a, 5224a, 5488a, 5638a, 5744a, 5858a, 5914a; a-verse with single alliteration: 327a, 332a, 460a, 680a, 686a, 1049a, 1466a, 1965a, 2031a, 2179a, 2180a, 2278a, 2306a, 2456a, 3058a, 3061a, 3079a, 3102a, 3136a, 3180a, 3383a, 3416a, 3558a, 3729a, 3918a, 4031a, 4102a, 4238a, 4293a, 4609a, 4715a, 4887a, 5218a, 5541a, 5601a, 5815a, 5819a, 5848a, 5978a; b-verse: 11b, 21b, 90b, 179b, 277b, 335b, 381b, 469b, 701b, 962b, 1026b, 1050b, 1294b, 1403b, 1984b, 2004b, 2030b, 2237b, 2248b, 2494b, 2635b, 2678b, 2695b, 2803b, 2811b, 2912b, 3031b, 3057b, 3156b, 3606b, 3607b, 3618b, 3630b, 3638b, 3671b, 3705b, 3758b, 3780b, 3818b, 3863b, 3917b, 3921b, 3922b, 4027b, 4036b, 4168b, 4528b, 4708b, 4759b, 4803b, 4886b, 4965b, 5143b, 5265b, 5422b, 5441b, 5524b, 5684b, 5983b P#Px(x)#(x)P a-verse with double alliteration: 510a, 732a, 1159a, 1344a, 1471a, 1473a, 1931a, 2171a, 3668a, 3710a; a-verse with single alliteration: 395a, 700a, 1117a, 1774a, 3107a, 4199a; b-verse: 103b, 639b, 769b, 926b, 973b, 2882b, 3074b, 3144b, 3612b, 3902b, 4368b, 4789b PSx(x)#(x . . .)P a-verse with double alliteration: 8a, 224a, 237a, 252a, 336a, 421a, 483a, 537a, 563a, 567a, 590a, 622a, 631a, 635a, 744a, 745a, 1125a, 1133a, 1137a, 1150a, 1242a, 1305a, 1328a, 1411a, 1424a, 1563a, 1641a, 1658a, 1668a, 1716a, 1721a, 1775a, 1776a, 1790a, 1799a, 1840a, 1876a, 1901a, 1991a, 1997a, 2055a, 2120a, 2151a, 2168a, 2183a, 2191a, 2219a, 2236a, 2239a, 2301a, 2307a, 2352a, 2359a, 2378a, 2414a, 2472a, 2548a, 2591a, 2650a, 2677a, 2735a, 2797a, 2867a, 2890a, 2891a, 2911a, 2982a, 3019a, 3141a, 3146a, 3270a, 3399a, 3466a, 3540a, 3545a, 3578a, 3735a, 3831a, 3908a, 3931a, 4101a, 4122a, 4137a, 4178a, 4234a, 4264a, 4284a, 4326a, 4331a, 4427a, 4535a, 4548a, 4730a, 4737a, 4841a, 4853a, 4879a, 4935a, 4948a, 4977a, 5004a, 5040a, 5090a, 5096a, 5452a, 5457a, 5491a, 5559a, 5829a; a-verse with single alliteration: 159a, 317a, 345a, 352a, 476a, 619a, 624a, 670a, 698a, 736a, 750a, 947a, 1058a, 1182a, 1192a, 1202a, 1268a, 1272a, 1428a, 1433a, 2132a, 2161a, 2391a, 2432a, 2798a, 2819a, 2878a, 2893a, 3078a, 3421a, 3446a, 3449a, 3594a, 3814a, 4185a, 4384a, 4527a, 4721a, 4761a, 4801a, 4811a, 4893a, 5038a, 5081a, 5095a, 5139a, 5241a, 5247a, 5355a, 5363a, 5576a, 5938a, 5977a; b-verse: 71b, 176b, 188b, 294b, 297b, 376b, 416b, 549b, 577b, 614b, 616b, 660b, 801b, 861b, 878b, 976b, 997b, 1035b, 1083b, 1087b, 1251b, 1448b, 1461b, 1504b, 1556b, 1644b, 1691b, 1850b, 1885b, 2087b, 2155b, 2487b, 2568b, 2624b, 2700b, 2705b, 2763b, 2856b, 2860b, 2932b, 3305b, 3694b, 4018b, 4050b, 4085b, 4169b, 4186b, 4225b, 4254b, 4260b, 4294b, 4297b, 4358b, 4554b, 4597b, 4628b, 4674b, 4747b, 5060b, 5071b, 5132b, 5201b, 5268b, 5399b, 5427b, 5546b, 5692b, 5705b, 5783b PS#x . . . P a-verse with double alliteration: 1287a, 1397a, 1623a, 2220a, 2323a, 2630a, 2993a, 3001a, 3033a, 4295a, 4320a, 4990a; a-verse with single alliteration: 167a, 1304a, 3244a; b-verse: 2293b, 3174b P#P#x . . . P a-verse with double alliteration: 1323a, 1494a, 1597a, 1707a, 1994a, 2094a, 2138a, 2914a, 2945a, 3132a, 3257a, 3483a, 4002a, 4076a; a-verse with single alliteration: 3981a; b-verse: 3362b, 4035b

MOSP-02.qxd

5/11/04

2:38 PM

Page 151

2.9. Type E

151

brings about higher incidence of a-verses with double alliteration. As regards these sequences with a lexical stress falling on the second position, the presence of a word-initial syllable #x immediately after it favours the a-verse with double alliteration to a greater extent, a pattern which obtains for the first drop of type A1 as treated in section 2.1 above: the strings -S#x and #P#x exhibit higher incidence of the a-verse with double alliteration than -Sx# and #Px#, respectively. On the other hand, the association of the same position with a primary-stressed syllable in heavy verses may lead to a lower frequency of the a-verse with double alliteration than with a secondary-stressed counterpart, the now familiar pattern that we find recurring across different metrical types, observed in sections 2.7 and 2.8 above: the sequence #Px# occurs less frequently than -Sx# in the a-verse with double alliteration. This is because heavy verses have at their disposal a wider range of alliterative patterns contingent upon syntactic constituency than those with a secondary-stressed drop, as will be fully discussed in section 3.2.2 below. The string #P#x, however, does not follow suit: it appears more frequently in the a-verse with double alliteration than -S#x. The above pattern of correlation stands in marked contrast to what is observed in Beowulf. In it, the b-verse constitutes a typical manifestation of type E consistently regardless of the morphological status (that is, compound versus non-compound) of the word that realises its first three positions, as well as of the number of syllables associated with its normal drop (Suzuki 1996a: 118). Moreover, the overall distribution hardly seems to be subject to anything approaching patterned variation: the heavy verse (P#Px . . . P) shows an unexpected distribution of being virtually limited to the b-verse, while the other four variants, PXx#(x)P, PS#x . . . P, PSx#(x)P, and even PX/x#x . . . P, are attested in the a-verse with double alliteration more frequently than is the configuration P#Px . . . P. The dramatic change in distribution pattern that sharply differentiates the Heliand from Beowulf would be difficult to characterise as a phonologically induced reorganisation comparable to the one that affected the first drop of type A1 (Suzuki 2000). The reintroduction of feet to non-lexical elements in word-final position (section 1.3.1) should hardly have had any effect on the composition of type E verses. The second position of this type was nearly always occupied by a lexical-stressed syllable -S- or -S#, or a non-lexical long medial syllable -X-. These three syllables were invariably footed in Old English, as they were immune to Defooting, which applied to non-lexical word-final syllables alone (Suzuki 1996a: 138–9). We may therefore be led to invoke metrical reasons for the marked difference at issue between the Heliand and Beowulf, namely, the blurring of normal and heavy Table 2.90. Distribution of major variants of type E in Beowulf variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

PX/x#x . . . P P#Px . . . P PXx#(x)P PS#x . . . P PSx#(x)P

1 8% 4 3% 13 18% 12 24% 75 28%

2 17% 1 1% 6 8% 0 0% 21 8%

9 136 55 38 176

75% 96% 74% 76% 65%

MOSP-02.qxd

5/11/04

152

2:38 PM

Page 152

2. Metrical types and positions

drops, and the consequent reorganisation of types A1 and A2 in the Heliand. In Beowulf, the metrical types involved in the normal versus heavy drop opposition were sensitive to the verse distribution in relation to the size of normal drops. Thus, the varied manifestations of the first normal drop of type A1 exhibited differing preferences for the a-verse with double alliteration, while type E remained constant in terms of its verse distribution patterning irrespective of the ways its first drop was realised. Characteristic of the Heliand, by contrast, was the obliteration of the twoway categorical distinction in favour of the gradient differentiation of a single entity: with the exception of the relic of the traditional type A2b (now reconfigured as type A2), there was no metrical reason to distinguish between basic and increased metrical types for classes A and D (see sections 2.1 and 2.7 above). The Heliand would accordingly have lost the significant criterion for differentiating type E from type A with respect to the structural basis of their first drop (i.e., second position). Because type A2a (/ \ / ⫻) was no longer identifiable as categorially distinct from type A1 in the Heliand, types A1 and E turned out to be indistinguishable insofar as their first drop was not involved in binary opposition: in type A1 it constituted a normal drop, and in type E a heavy drop, both lacking their structural opposites, a heavy and a normal drop, respectively. As pointed out above, the strings -Sx# and -Xx# were prototypical occupants of the medial positions of type E (\ ⫻). In the traditional metre, these strings were categorically excluded from the first drop of type A1, because they were footed, and because the normal drop was unqualified for association with footed material in traditional practice. With the reorganisation of the first drop of types A1 and A2 in force, however, -Sx# and -Xx# became fully qualified as its occupants, much as was -S#x, which had marginally been allowed in the traditional metre as well. Thus, the complementary distribution of the strings -Sx# and -Xx# in regard to types A1 and E was now lost; accordingly, not only PSx#P and PXx#P (type E) but also PSx#Px and PXx#Px (type A1) were counted as metrical in the innovated metre. The new type A1 variants PSx#Px and PXx#Px were of highly marked status, appearing as they did in the a-verse with double alliteration with much preference (Suzuki 2000). Traditionally, however, the corresponding type E variants PSx#P and PXx#P occurred in the b-verse with high frequency, as did type E in general without notable difference among its variants (see Table 2.90 above). Thus, without further restructuring the strings #PSx# and #PXx# would have had to be treated in divergent ways by these two metrical types in regard to their distribution pattern. While there may have been no logical necessity for avoiding such divergent treatment at all costs, it would not be surprising to see that the composition of the first drop of type E was reorganised by analogy on the basis of the organisation of the corresponding position of type A1, provided that some measure of generalisation was deemed appropriate as a means of attaining greater harmony in the metrical system. Type A1 was by far the more frequent type of the two, and the specific is more likely to follow the general rather than vice versa. Consequently, it now became possible for type E to be treated in the same way as type A1: type E was rendered sensitive to the varying weight of its first and second drops, and thus became subject to the covarying verse distribution pattern in concert with the general gradation pattern valid for type A1.

MOSP-02.qxd

5/11/04

2:38 PM

Page 153

2.9. Type E

153

The reorganisation of the first drop of type E along the above lines of generalisation, however, could not have been carried to its logical conclusion in defiance of the traditional metrical practice. Specifically, the position at issue served as a heavy drop rather than a normal counterpart in the traditional versecraft, and this rule persisted in the metre of the Heliand. This observance of the traditional practice in the Heliand is evidenced among others by the following features: (i) the constant association of the second position with a monosyllable or a resolved disyllable (rather than a multiplicity of unstressed syllables); (ii) the complete absence of verses of the form Pxx#P (as opposed to PXx#P); (iii) the preference of trisyllabic words (#PSx#, #PXx#) to disyllabic counterparts plus another syllable (#PS#x, #PX#x) as an occupant of the first three positions; (iv) the occurrence of the configuration PSx#P (as opposed to PXx#P) as the most favoured realisation of the first three positions. While the normal drop in non-final position may be associated with a concatenation of unstressed syllables, the heavy counterpart is limited to a single (stressed) syllable or a resolved disyllable. Feature (i) accordingly shows that the position at issue cannot be identified with the normal drop in its typical form. Since the normal drop is prototypically realised by a syllable of minimal prominence, the association with the long syllable and the dissociation from the short counterpart (feature [ii]) further confirms that the position in question is far from a prototypical normal drop. As discussed in sections 2.1.7 and 2.4.2.3 above, the medial syllables -S- and -Xcount as more prominent than the word-final counterparts -S# and -X#, respectively. The preferred association of the second position of type E with the medial syllables rather than the word-final counterparts (feature [iii]) adds plausibility to the view that the position under discussion should better be characterised as heavy. Finally, the maximal preference for the lexical-stressed medial syllable (-S-) rather than the one without lexical stress (-X-) leads us to conclude that the second position of type E has to be identified as a heavy drop. Also contributing to the persistent identity of the heavy drop in type E would have been the structural instability of the otherwise resulting configuration / ⫻ ⫻ /. That is, as pointed out in Sievers (1893: 31; see also Suzuki 1996a: 129), this configuration suffers from structural ambiguity: by virtue of multiple association of the normal drop, whatever sequence of unstressed syllables standing between stressed syllables would be assigned with a single normal drop (⫻), rather than two. Thereupon, the unmetrical sequence / ⫻ / would result without further language expressions to follow. In other words, on perceiving whatever syllable sequence analysable as / ⫻ ⫻ /, the audience would be misled to expect something more to follow, rather than take it as a completed expression. More specifically, the audience would be most likely to be held in suspense, awaiting still a further unstressed syllable, so that the whole sequence then could be identified as an instance of type A1, the commonest occurring configuration. Thus, constrained by this traditional status of the second position of type E, the Heliand poet, highly innovative as he was, did not totally break away from the conventional practice: he did not compose verses like Pxx#P at all by fully identifying the position in question with the normal drop (see above); nor did he produce a greater number of verses of the form PXx#P than PSx#P (see Table 2.89 above), although the syllable with lexical stress S counts as marked and correspondingly less typical as a realisation of the normal drop. The Heliand poet accordingly had to

MOSP-02.qxd

5/11/04

154

2:38 PM

Page 154

2. Metrical types and positions

conform largely to the traditional practice by distinguishing the first drop of type E from the prototypical normal drop, despite his innovation of making it more harmonious with it through graded patterning. The composition of type E was thus largely different from that of type A1. Type E had access to a much narrower range of variation, as shown in Table 2.89 above, because the first drop was a heavy one, limited to a single foot for association, as demonstrated above. Consequently, as far as the realisation of the second and third positions is concerned, only three major variants were available for gradation: -Sx#, -S#x, -Xx#. Type E accordingly became subject to a lesser graded distinction along a smaller number of parameters, which primarily concerned long syllables immediately after the first lift. As may be recalled, these three strings were all counted as marked realisations of the first drop of type A1 by virtue of the long syllable standing immediately after the first lift (parameter (i); section 2.1.7.1). The first parameter having thus applied vacuously to all the major variants beginning with a long syllable, the supplementary parameter based on the opposition between lexical and non-lexical stresses came into play, whereby the presence of lexical stress was stipulated as marked (section 2.1.7.2). The lexical and non-lexical syllables were accordingly distinguished: -S- vs. -X-. Assigned thereby a marked value, the lexical-stressed syllable -S- came to be used with higher frequency in the a-verse with double alliteration than the unmarked counterpart -X-. Added to the opposition between lexical and non-lexical stresses was the opposition between the word-initial (#x) and word-final unstressed syllables (-x#), which served as another significant parameter for controlling the gradation of the first drop of type A1 among other positions (section 2.1.7.1). By this parameter, the string -S#x was counted as marked in contrast to -Sx#, and accordingly occurred more frequently in the a-verse with double alliteration than Sx#.29 The remaining major string -Xx# was thus characterised as an unmarked variant, which motivated the dominant occurrence of it in the b-verse. In this way, the more or less uniform distribution of type E in the traditional metre (Table 2.90) became susceptible to gradation in the Heliand, so that a more marked, prominent variant appeared with greater frequency in the a-verse with double alliteration. The poet would have designed this novel gradation pattern for type E by making analogical extension of the generalisations that he devised in control of the variation on the first drop of type A1. To conclude, in accordance with the full gradation of type A1 tuned to the varied realisation of its first drop (section 2.1.7), the major variants of type E underwent an analogous gradation on the basis of the variation on realisation of their first and second drops (-S#x, -Sx#, -Xx#). They were ranked in terms of markedness based on prominence (in the descending order of markedness as given) and in such a way that the most marked variant (-S#x) showed a maximum preference for the a-verse with double alliteration, and conversely the least marked counterpart (-Xx#) appeared in the b-verse most frequently. In determining varying degrees of markedness, the poet drew on the applicable subset of the parameters that he used for reorganising the gradation pattern of the first drop of type A1. The establishment of parallel graded patterns for types A1 and E may therefore warrant our

29

A similar contrast seems to apply to the pair of -Xx# and -X#x, although the string -X#x occurs with a minimal frequency.

MOSP-02.qxd

5/11/04

2:38 PM

Page 155

2.9. Type E

155

assuming that the Heliand poet had no less high awareness of structural organisation of the metre than of the narrative discourse (cf. Murphy 1989; 1992; see also Chapter 5 below). In addition to the variation on the first drop in combination with the second, type E is implicated in gradation in terms of the size of the second drop. The number of syllables occupying this position varies between one and four, as indicated in Table 2.91. As Table 2.91 makes evident, we may generalise the overall distribution pattern as follows: as the number of syllables occupying the second drop (⫻) increases, the relative frequency of the b-verse decreases in favour of the a-verse with double alliteration. This general pattern obtains equally for each of the subtypes identified earlier. This correlation is expressed in the clearest fashion when the position in question is filled by three syllables or more: with a single exception, only a-verses with double alliteration may be provided with such long sequences of syllables for the normal drop. Compared with the distribution pattern relating to the number of syllables involved in the second drop of type E in Beowulf, it becomes clear that the correlation determined above constitutes a distinct feature of the Heliand. By contrast, Beowulf shows a consistent preference for the b-verse regardless of the number of syllables occurring in the second drop of type E, as substantiated in Table 2.92. The preference for the b-verse is epitomised when the position is filled maximally, the distribution diametrically opposed to that obtained in the Heliand. Thus we find here again the recurring pattern characteristic of the Heliand, whereby the larger the number of syllables appearing in the drop in non-initial position, the greater the frequency of occurrence in the a-verse with double alliteration. Of particular interest is the similarity in patterning to the varying distribution of type B1 relative to the number of syllables involved in the second drop (section 2.4.2.1). Table 2.91. Distribution of the varying number of syllables in the second drop of type E number of syllables

a-verse with double alliteration

a-verse with single alliteration

b-verse

1 2 3 4

102 34% 64 52% 6 86% 1 100%

76 26% 29 24% 1 14% 0 0%

119 40% 30 24% 0 0% 0 0%

Table 2.92. Distribution of the varying number of syllables in the second drop of type E in Beowulf number of syllables

a-verse with double alliteration

a-verse with single alliteration

b-verse

1 2 3

94 19% 11 19% 0 0%

25 5% 5 8% 0 0%

365 75% 43 73% 4 100%

MOSP-02.qxd

5/11/04

156

2:38 PM

Page 156

2. Metrical types and positions

2.10. Type E* (/ ⫻ \ ⫻ /) This metrical type, consisting of five metrical positions, is distinguished from type E by an additional drop standing between the first lift and the heavy drop. Based on the morphological status of the language material that occupies the heavy drop (the third position), this type falls into two major variants. First, the position in question may be filled by a primary-stressed syllable, the whole configuration being Px . . . Px . . . P, as exemplified in (86) and listed under Table 2.93 below (cf. Kauffmann 1887: 346; Heusler 1956: §263; Hofmann 1991, II: 209–12): (86) 6a liudo barno lobon 978a dôpte allan dag 2805a gôdes mannes forgang In the absence of double alliteration, metrical ambiguity would appear to arise in regard to the following verses, for example, which may scan as type E* and subtype D*2b alike, as observed in section 2.8 above: (87) 1223a armoro manno filu 254b gôdes cunnies man 5663b harda stênos clubun As will be shown in section 3.2.2 below, however, difference in syntactic constituency crucially bears on scansion, and it is on this basis that the above three verses and others like them should properly be scanned as type E*. Second, the heavy drop may be associated with the second element of a compound word, resulting thereby in the configuration PxSx . . . P (cf. Kauffmann 1887: 346; Hofmann 1991, II: 210–12): (88) 1676a mêðomhordas mêst 2529a sô endilôsan uuelon 36b uundarlîcas filo The distribution pattern of type E* with two major realisation variants (on the significance of the third variant expanded with anacrusis x . . . Px . . . P . . . P/x . . . PxSx . . . P, see below) is shown in Table 2.93. With respect to the realisation of the normal drops, the first drop (second position) usually corresponds to a single syllable (forty-nine examples out of a total of fiftythree occurrences of type E*) and is nearly always occupied by a word-final or medial syllable. In verse 1255a Nemnida sie thô bi naman, in which the position at issue is filled by a string of three syllables, an independent word is also involved in association in addition to a combination of medial and word-final syllables. The third drop (penultimate position) is associated with a greater variety of language material than the first counterpart. First, the number of associated syllables is larger: only about a half of all examples are minimally filled (twenty-six instances out of the fifty-three); in another twenty-six examples the position corresponds with

MOSP-02.qxd

5/11/04

2:38 PM

Page 157

2.10. Type E*

157

Table 2.93. Distribution of type E* number of syllables

a-verse with double alliteration

a-verse with single alliteration

b-verse

Px . . . Px . . . P PxSx . . . P

27 68% 9 69%

6 15% 1 8%

7 3

18% 23%

2

2

18%

x . . . Px . . . P/Sx . . . P 7

64%

18%

Px . . . Px . . . P a-verse with double alliteration: 6a, 47a, 52a, 268a, 503a, 865a, 978a, 1255a, 1634a, 1710a, 1746a, 1766a, 2170a, 2192a, 2206a, 2232a, 2805a, 2882a, 3665a, 4231a, 4807a, 5029a, 5148a, 5426a, 5712a, 5773a, 5804a; a-verse with single alliteration: 1223a, 1555a, 3147a, 3524a, 4245a, 4799a; b-verse: 254b, 492b, 4084b, 4131b, 4547b, 5663b, 5664b PxSx . . . P a-verse with double alliteration: 468a, 735a, 1676a, 2085a, 2333a, 2652a, 3036a, 5622a, 5660a; a-verse with single alliteration: 2529a; b-verse: 36b, 1467b, 3006b

two syllables, and there is a single case in which the position in question is occupied by three syllables: 735a giâmarlîcara forgang. Second, when filled by a single syllable, the position at issue may be realised by not only a medial or word-final syllable but also a prefix or an independent (function) word: 1255a, 1555a, 1634a, 2192a, 2206a, 2882a, 3524a, 5622a. It should be noted that, when accompanied by anacrusis, type E* verses, with the resulting configurations x . . . Px . . . Px . . . P or x . . . PxSx . . . P, might be scanned as type B1 by demoting the medial lexical-stressed syllables P or S and reducing therewith these verses to those like 2615b than uueldi gerno gehue uuesan and 3581b that sie erðe endi himil. We count eleven such examples, as follows (cf. Kauffmann 1887: 346; Hofmann 1991, II: 209–11): (89) 268a thes uuîdon rîkeas giuuand 2882a sô grôte craft mid gode 3665a themu landes hirdie te lobe a-verse with double alliteration: 268a, 2333a, 2652a, 2882a, 3036a, 3665a, 5804a; a-verse with single alliteration: 1555a, 2529a; b-verse: 1467b, 5664b There are four grounds for scanning the above verses as type E* with anacrusis rather than type B1, however. First, as shown in Table 2.93 above, the predominance of a-verses with double alliteration, which is fully in line with the general distribution pattern of type E* characterised with a marked preference for the a-verse, would cast doubt on the scansion of these verses as type B1, which is attested with overwhelming frequency in the b-verse (section 2.4). Second, a vast majority of these examples, eight out of the total of eleven, have a single unstressed syllable at the beginning. As will be shown fully in the following section, anacrusis is most typically realised by a single syllable. By contrast, as substantiated in section 2.4.1 above, the first drop of type B1 is commonly occupied by a sequence of two or more syllables without a significant concentration on a particular number of syllables such as is observed in the group of verses under consideration. Third, as regards the configuration x . . . Px . . . Px . . . P, the association of class 1 words (nouns and adjectives; for

MOSP-02.qxd

5/11/04

2:38 PM

158

Page 158

2. Metrical types and positions

details, see section 3.2.2 below) such as craft and rîkeas with the normal drop, the kind of association that the scansion as type B1 would have to invoke, is most implausible: it would violate one of the fundamental principles of association otherwise unviolable. Accordingly, the configuration x . . . Px . . . Px . . . P defies scansion as type B1. Fourth, in regard to the configuration x . . . PxSx#(x)P (instantiated by verses 2333a, 2529a, 2652a, 3036a, 1467b), there are substantive reasons for identifying the minimally distinct configuration x . . . PSx#(x)P, lacking in the extra drop between P and S, as type E with anacrusis, rather than type B1, as argued in section 2.4.2.3 above. This may lend added plausibility to the analogous scansion of the sequence expanded with the medial drop as a metrical type minimally different from type E, namely type E*. In the traditional metre, type E* is not acknowledged as an authentic metrical type. To be sure, we find a number of verses of the configuration PxSx . . . P in Beowulf – one of the two major variants of type E* in the Heliand – as listed below (Fulk 1992: §§88, 95): (90) Beo 1136a Beo 1187a Beo 1681a Beo 1918a Beo 2894a Beo 104b Beo 667b Beo 995b Beo 998b Beo 1459b Beo 1782b Beo 2387b Beo 2757b Beo 3037b

wuldortorhtan weder umborwesendum æ¯r wundorsmiþa geweorc oncerbendum fæst morgenlongne dæg fifelcynnes eard sundornytte beheold wundorsiona fela irenbendum fæst atertanum fah symbelwynne dreoh Ongenðioes bearn maððumsigla fealo wundordeaðe swealt

Yet these verses all contain a weak vowel immediately before a sonorant, /r/, /l/, /n/, or /m/. And as treated fully in Fulk (1992: §§76–98), there is every reason to assume that the vowels in question arose derivatively as consequences of vowel parasiting, whereby weak vowels developed before the originally syllabic sonorants. These parasitic vowels therefore do not count in scansion, and as a convention for indicating their metrical insignificance they are underdotted in modern editorial practice: e.g., wuldo.rtorhtan. The apparent sequence PxSx . . . P therefore reduces to the configuration PSx . . . P as a minor realisation variant. In this connection, it may be as well to mention that the principle of four metrical positions per verse was so fully in force in the traditional metre that it would have excluded the possibility of creating a new type consisting of five positions such as type E*. Given the lack of authenticity of the sequence PxSx . . . P in the traditional metre, one might reasonably wonder whether the same may apply to the Heliand. In order to justify fully that the configuration PxSx . . . P constitutes an independent metrical type in the Heliand metre, we need to have a closer look at all instances of this

MOSP-02.qxd

5/11/04

2:38 PM

Page 159

2.10. Type E*

159

configuration, as listed below (cf. Kauffmann 1887: 346; Hofmann 1991, II: 210–12): (91) 468a 735a 1676a 2085a 2333a 2529a 2652a 3036a 5622a 5660a 36b 1467b 3006b

sâliglîcan sebon giâmarlîcara forgang mêðomhordas mêst gôdlîcnissea godes is bedgiuuâdi te baka sô endilôsan uuelon sô craftiglîco giqueðen mid sûðarliudiun giseten uuundarlîc giuuaraht uundartêcan giuuaraht uundarlîcas filo sô lango sô thu fîundskepies uuiht ungilôbiga sind

As it turns out, there are not a few examples that show a weak vowel before a sonorant, notably /r/, analogous to the verses in Beowulf listed in (90) above. One might assume then that parasitic vowels are involved here and accordingly that these verses are reducible to the configuration PSx . . . P, namely, type E. Not all instances, however, are explainable along the same line. Some verses contain a vowel appearing in contexts other than those conducive to vowel parasiting: -gi- (2333a, 3006b), -ig(468a, 2652a); -lîc- (2085a); -und- (1467b). Moreover, there are a larger number of verses with the configuration Px . . . Px . . . P (as exemplified in (86) above), which is virtually absent in Beowulf, with a single possible exception of Beo 1424b Feþa eal gesæt. And for these verses of the form Px . . . Px . . . P, vowel parasiting seems largely out of the question as a credible account. Further, the same extra drop may be realised by a string of syllables, a disyllabic one as in 2206a hêlandi Crist an hand, or even a trisyllabic one as in 1255a Nemnida sie thô bi naman. These instances would be most unlikely to have been affected by vowel parasiting. The foregoing consideration may thus justify us in postulating an independent metrical type, type E* (/ ⫻ \ ⫻ /), for the metre of the Heliand. In the traditional metre, on the other hand, one of the two configurations realising this type, PxSx . . . P, was demonstrably in existence, yet this was none other than a marginal realisation variant of type E. We may accordingly conclude that type E* was brought to being by innovation in the Heliand. A question facing us then is why and how this new type arose and came to be integrated into the Heliand metre. As observed above, as regards the configuration PxSx . . . P commonly found in Beowulf and the Heliand, slightly less than the half of all instances in the Heliand (six out of thirteen examples, more precisely) are closely parallel to those in Beowulf in their phonological structure: the first drop is realised by a medial vowel that stands immediately before a sonorant. Furthermore, the same lexical item is involved in several verses: uundar- ‘wonder-’ (5622a, 5660a, 36b) on the one hand and wundor(Beo 1681a, 995b, 3037b) on the other. Following Sievers’s (1893: §116.9) insight, we may accordingly surmise that the configuration PxSx . . . P attested in the traditional metre would have provided a structural basis for extended use in the Heliand.

MOSP-02.qxd

5/11/04

160

2:38 PM

Page 160

2. Metrical types and positions

Having identified a probable source of analogical extension, let us consider the motivation for the innovation of introducing type E* in the Heliand metre. To this end, we need to explore the issue against the general background of the metrical reorganisation at work in the Heliand, rather than regarding it as an isolated phenomenon. Specifically, in traditional versecraft, unstressed materials may be realised as attachments to the first lift in strictly circumscribed contexts. As a consequence, we obtain what appear to be extra weak elements in a verse in addition to its mandatory four metrical positions. The first drop of subtype D* is an instance of such additional drops, and anacrusis is another (for detail on anacrusis, see section 2.11 below). These apparently expanded verses, however, comprised four positions as did ordinary verses in general, and whatever language material seemingly occupying the extra drop was in fact associated with the immediately preceding (subtype D*) or following (anacrustic verses) first lift. This traditional basis of composition was radically reorganised in the Heliand, however. The Heliand metre reconfigured the apparent drop on the surface as an independent metrical position constitutive of a metrical type, one on an equal footing with the four others. Consequently, unstressed expressions that were originally attached by extensive association to the first lift came to occupy this new weak position on their own. The establishment of type E* as an autonomous type may be understood in similar fashion. The extra weak syllables, which arose secondarily from vowel parasiting, lacked independent metrical status in the traditional metre; rather they were literally parasitic on the preceding first lift. Such an inorganic status, however, received metrical recognition in the Heliand metre along with the first drop of subtype D* and anacrusis, thereupon giving rise to type E*. Once given a full recognition as an underlying metrical position, language materials other than those due to vowel parasiting became available for composition, and a wealth of verses traditionally disallowed came into being as authentic metrical expressions in the Heliand.

2.11. Anacrusis Anacrusis is a weak element that may optionally occur at the beginning of a verse. Because of its optionality, anacrusis counts as something extra to the four metrical positions constitutive of the normal verse. Yet the exact structural value of such an extrinsic verse element may vary from a metrical system to another even within the broadly same tradition. In this section, I shall show that, optional as it still was in metrical structure, anacrusis was reconfigured as a full metrical position in the Heliand, therewith being treated scarcely different from the obligatory metrical positions; it was thus no longer subject to the severe restrictions that had been imposed on its distribution and realisation in the traditional metre. By contrast, in Beowulf anacrusis was a derived element or a ‘pseudo position’ on the surface: it constituted an integrated part of the verse-initial lift, rather than an independent position on its own; its occurrence was accordingly strictly controlled, and its distribution clearly bounded; and it served a special marking function unknown to other positions.

MOSP-02.qxd

5/11/04

2:38 PM

Page 161

2.11. Anacrusis

161

2.11.1. Anacrusis as a full metrical position: a synchronic account Anacrusis is used much more freely and extensively in the Heliand than in Beowulf, which places rather strict restrictions on its occurrence. In Beowulf, anacrusis serves as a marker of increased metrical strength, as demonstrated in Suzuki (1995a; 1996a: 315–40; see also Duncan 1994: 10–11): it prototypically occurs at the beginning of those metrical configurations that show maximal preference for the a-verse with double alliteration. In other words, anacrusis is limited to type A2 (e.g., Beo 2252a in (92) below), and to heavy variants of type A1 in which the first drop is occupied by more than a single unstressed syllable (e.g., Beo 409a in (92) below); and to types D2a (e.g., Beo 2455a in (92) below) and D2b (e.g., Beo 1667a in (92) below). By contrast, anacrusis is disallowed from the b-verse in general; and even in the a-verse, it is not allowed to attend light variants of type A1 such as Beo 11a gomban gyldan and Beo 252a frumcyn witan, or type D1 verses like Beo 2409a wong wisian. Since type E fails to exhibit obligatory realisation of double alliteration, and since it appears in the b-verse more frequently (Table 2.90), it does not qualify to be expanded with anacrusis, despite the presence of the verse-initial lift. Thus, in Beowulf the distributional predilection for the b-verse is incompatible with use of anacrusis. Moreover, anacrusis in Beowulf is prototypically realised by a verbal prefix (e.g., for-, ge-, on-) or the negative particle ne; in consequence, anacrusis is, with rare exceptions, limited to monosyllables. Because type A2a begins with a compound noun for the most part, it may follow that this type lacks anacrusis despite its increased metrical strength. (In the following exemplification, anacrustic syllables are italicised.) (92) Beo 409a Beo 1667a Beo 2252a Beo 2455a

ongunnen on geogoþe forbarn brogdenmæ¯l gesawon seledream Gesyhð sorhcearig

By contrast, anacrusis in the Heliand lacks all the above prototypical properties that characterise anacrusis in Beowulf. First, light variants of type A1 may show anacrusis in much the same way as their heavy counterparts and type A2, as follows (cf. Kauffmann 1887: 299–308; Hofmann 1991: 67; 1991, II: 32–5, 54–9, 62): (93) 216a that barn an ire barme 296a is môd giuuorrid 641a them uuîsun mannun 3189a thero hôbidscatto 5124a there thiade thinghûs 40b mid ênu uuordo 421b them the god antkennead 1169b scoldun sâliglîco x . . . Px#Px a-verse with double alliteration: 137a, 271a, 492a, 545a, 656a, 835a, 1083a, 1215a, 1417a, 1506a, 1532a, 1892a, 1974a, 2399a, 2516a, 2663a, 2688a, 2875a, 2879a, 3321a, 3480a, 3605a, 3627a, 3648a, 3676a, 3715a, 3816a, 3867a, 3875a, 3887a, 3939a, 4473a, 4844a, 4860a, 4874a, 5075a, 5147a, 5156a, 5183a, 5206a, 5227a, 5472a, 5482a, 5578a, 5682a, 5793a, 5945a; a-verse with single

MOSP-02.qxd

5/11/04

162

2:38 PM

Page 162

2. Metrical types and positions alliteration: 114a, 291a, 561a, 641a, 808a, 844a, 879a, 1257a, 1706a, 1759a, 1802a, 1812a, 1883a, 1983a, 2318a, 2380a, 2452a, 2454a, 2686a, 2775a, 2895a, 2901a, 2920a, 2954a, 3020a, 3026a, 3077a, 3105a, 3347a, 3738a, 4230a, 4739a, 4907a, 4983a, 5142a, 5219a, 5304a, 5407a, 5533a, 5553a, 5581a, 5666a, 5668a, 5799a, 5800a, 5826a, 5959a; b-verse: 7b, 17b, 40b, 101b, 146b, 157b, 290b, 345b, 415b, 432b, 482b, 518b, 520b, 631b, 656b, 671b, 706b, 744b, 849b, 854b, 875b, 917b, 1006b, 1110b, 1129b, 1147b, 1176b, 1235b, 1255b, 1268b, 1284b, 1286b, 1295b, 1333b, 1361b, 1386b, 1389b, 1470b, 1471b, 1472b, 1507b, 1511b, 1609b, 1628b, 1641b, 1718b, 1808b, 1962b, 1974b, 1994b, 2068b, 2074b, 2128b, 2181b, 2232b, 2261b, 2297b, 2313b, 2330b, 2338b, 2380b, 2382b, 2388b, 2389b, 2396b, 2399b, 2400b, 2413b, 2456b, 2479b, 2483b, 2547b, 2548b, 2586b, 2595b, 2609b, 2611b, 2630b, 2690b, 2719b, 2761b, 2764b, 2845b, 2891b, 2899b, 2924b, 2997b, 3003b, 3016b, 3104b, 3124b, 3157b, 3192b, 3219b, 3246b, 3263b, 3326b, 3352b, 3359b, 3410b, 3439b, 3508b, 3518b, 3582b, 3635b, 3641b, 3676b, 3708b, 3849b, 3892b, 3912b, 3919b, 3923b, 3926b, 4002b, 4066b, 4125b, 4132b, 4136b, 4162b, 4165b, 4175b, 4179b, 4243b, 4367b, 4497b, 4566b, 4657b, 4667b, 4696b, 4714b, 4725b, 4776b, 4807b, 4830b, 4836b, 4843b, 4870b, 4897b, 4900b, 4909b, 4936b, 4950b, 4972b, 4988b, 4999b, 5004b, 5029b, 5051b, 5096b, 5114b, 5187b, 5233b, 5234b, 5249b, 5298b, 5316b, 5341b, 5366b, 5375b, 5450b, 5461b, 5487b, 5495b, 5516b, 5541b, 5543b, 5549b, 5550b, 5555b, 5564b, 5588b, 5649b, 5651b, 5661b, 5675b, 5736b, 5799b, 5807b, 5818b, 5820b, 5830b, 5832b, 5859b, 5871b, 5891b, 5894b, 5911b, 5935b, 5939b, 5941b, 5979b x . . . P#xPx (#x ⫽ prefix) a-verse with double alliteration: 1792a, 2087a, 2276a, 2518a, 2645a, 4348a, 4485a, 4785a, 5311a, 5528a; a-verse with single alliteration: 296a, 1152a, 1218a, 1253a, 1513a, 1917a, 2169a, 2347a, 2517a, 3333a, 3336a, 3628a, 3781a, 4081a; b-verse: 421b, 1090b, 1804b, 2522b, 2941b, 3140b, 3153b, 3387b, 4092b, 4222b, 5165b, 5742b x . . . PxSx a-verse with single alliteration: 333a, 448a, 1811a, 2661a, 3189a, 5066a; b-verse: 1169b, 2249b, 2890b, 3465b, 5687b, 5946b x . . . P#xxxPx a-verse with double alliteration: 216a, 292a, 320a, 379a, 382a, 394a, 437a, 449a, 555a, 592a, 713a, 741a, 777a, 805a, 822a, 831a, 933a, 1048a, 1576a, 1888a, 2067a, 2103a, 2203a, 2265a, 2427a, 2442a, 2770a, 2873a, 2884a, 3009a, 3156a, 3161a, 3213a, 3705a, 3841a, 3861a, 4061a, 4130a, 4158a, 4159a, 4355a, 4521a, 4583a, 4714a, 4775a, 4945a, 4967a, 4976a, 5002a, 5098a, 5115a, 5192a, 5362a, 5453a, 5460a, 5484a, 5632a, 5703a, 5756a, 5758a, 5779a, 5808a, 5824a, 5828a, 5880a, 5939a; a-verse with single alliteration: 995a, 3659a; b-verse: 1085b, 3687b, 3981b, 5449b x . . . Px#PS a-verse with double alliteration: 2585a, 3860a, 5124a, 5321a, 5400a

Among type A1 variants in which the two lifts are occupied by primary-stressed syllables, the prototypical variant Px#Px (section 2.1.5) shows a minimal degree of preference for the a-verse with double alliteration: as demonstrated above (see also

MOSP-02.qxd

5/11/04

2:38 PM

Page 163

2.11. Anacrusis

163

Hofmann 1991, II: 54, 60 and Duncan 1994: 6–7, 10), this configuration realises the a-verse with double alliteration in less than 20 per cent of all its occurrences (174 out of a total of 925 instances; cf. Table 2.23);30 conversely, the same variant exhibits a marked preference for the b-verse, accounting for 64 per cent (588 occurrences out of the total of 925). However, this variant is expandable with anacrusis, as indicated by verses 641a and 40b in (93) above: the anacrustic variant x . . . Px#Px has 286 examples in the corpus; and it should be emphasised that anacrusis is as common in the b-verse as in the a-verse, the pattern recurrently observed in other configurations as well. The variant P#xPx, where the first drop x embodies a prefix attached to the second lexical word in the verse (section 2.1.5), also shows a comparably strong preference for the b-verse: it occurs twenty-eight times (c. 25 per cent) in the a-verse with double alliteration, and seventy-two times (c. 65 per cent) in the b-verse (cf. Table 2.26). This configuration thus fails to demand maximal implementation of double alliteration, much as the verse form Px#Px. Anacrusis, however, applies to this variant as well, as testified by verses 296a and 421b given in (93) above. Furthermore, even the variant PxSx, which is attested in the a-verse with double alliteration only once (785a môdarmâgun; cf. Table 2.29), allows for anacrusis, as illustrated by verses 3189a and 1169b in (93) above. By way of comparison, heavy variants of type A1, as well as type A2, are characterised by their marked preference for the a-verse with double alliteration (sections 2.1.7 and 2.2.1). For example, one of the heavy variants of type A1, the configuration P#xxxPx, appears sixty-two times (c. 83 per cent) in the a-verse with double alliteration and twelve times (c. 16 per cent) in the b-verse (cf. Table 2.28); and type A2 in its minimal form Px#PS occurs seven times exclusively as a-verses with double alliteration (cf. Table 2.38). As expected, these heavy variants may be accompanied with anacrusis, as shown by verses 216a and 5124a given in (93) above. Second, types D and D*, even their light variants, subtypes D1 (e.g., 289a, as given in (94) below) and D*1 (e.g., 1460a), respectively, may be expanded with anacrusis, as are their heavy variants, that is, subtypes D2a (e.g., 2900a) and D2b (e.g., 3081a), on the one hand, and subtypes D*2a (e.g., 4354a) and D*2b (e.g., 5466a), on the other, as exemplified below (cf. Kauffmann 1887: 336; Hofmann 1991, II: 177–8, 194–5, 197): (94) 289a 1460a 2900a 3081a 4354a 5466a 30

that godes ârundi geflîhit uuiðar is fîunde tefôr folc mikil antloken liohto mêst thiu mikilo meginstrengi giblôðit briostgithâht

It should be noted that the tables provided in the preceding sections on the distribution of metrical types and their variants include anacrustic verses. For the sake of accuracy, the figures given in the present context concern the non-anacrustic verses alone. As substantiated in section 3.2.1.3 below, however, anacrusis hardly affects distribution patterning, and thus referring to the earlier tables would be illuminating all the same for our concerns here.

MOSP-02.qxd

5/11/04

164

2:38 PM

Page 164

2. Metrical types and positions x . . . P#PX/xx a-verse with double alliteration: 5410a; a-verse with single alliteration: 289a; b-verse: 18b, 751b, 992b, 2689b, 2854b, 3029b, 4590b, 4635b, 4790b, 5188b, 5702b, 5789b, 5848b, 5926b x . . . P#PS/sx, x . . . P#P#P/px a-verse with double alliteration: 297a, 620a, 915a, 1097a, 2900a, 4040a, 4052a, 5398a, 5443a, 5741a; a-verse with single alliteration: 2459a, 5481a; b-verse: 1692b, 3937b x . . . P#Px . . . S, x . . . P#Px . . . P a-verse with double alliteration: 197a, 599a, 671a, 693a, 3081a, 3163a, 3198a, 3882a, 4573a, 5140a, 5380a, 5392a, 5743a, 5929a; b-verse: 4241b x . . . Px . . . PX/x a-verse with double alliteration: 42a, 323a, 371a, 594a, 860a, 973a, 1460a, 1599a, 2125a, 2413a, 2546a, 2659a, 2810a, 3176a, 3224a, 3414a, 3441a, 3769a, 3883a, 4400a, 4414a, 4499a, 4572a, 4663a, 4881a, 5634a, 5887a, 5941a, 5963a, 5965a; a-verse with single alliteration: 940a; b-verse: 1309b, 2541b, 3526b, 4593b, 5118b x . . . Px . . . PS/sx, x . . . Px . . . P#P/px a-verse with double alliteration: 193a, 300a, 360a, 386a, 454a, 507a, 565a, 637a, 648a, 672a, 696a, 796a, 883a, 895a, 1084a, 1178a, 1352a, 1381a, 1388a, 1485a, 1504a, 1526a, 1642a, 1646a, 1661a, 1726a, 1896a, 1949a, 1960a, 1976a, 1989a, 1996a, 2009a, 2063a, 2081a, 2176a, 2181a, 2251a, 2269a, 2317a, 2371a, 2395a, 2403a, 2515a, 2583a, 2640a, 2653a, 2666a, 2667a, 2768a, 2774a, 2776a, 2847a, 2899a, 2924a, 2929a, 2942a, 2943a, 2946a, 3154a, 3248a, 3277a, 3297a, 3335a, 3476a, 3603a, 3637a, 3687a, 3698a, 3704a, 3797a, 3943a, 4011a, 4089a, 4115a, 4119a, 4123a, 4190a, 4242a, 4259a, 4347a, 4354a, 4374a, 4441a, 4451a, 4497a, 4662a, 4722a, 4743a, 4755a, 4913a, 4981a, 5068a, 5089a, 5101a, 5133a, 5171a, 5190a, 5195a, 5213a, 5339a, 5364a, 5367a, 5368a, 5375a, 5549a, 5646a, 5655a, 5708a, 5774a, 5803a, 5909a, 5935a, 5967a; a-verse with single alliteration: 518a, 2520a, 4973a; b-verse: 832b, 1165b, 2533b, 2957b, 3270b, 4753b, 4796b, 5093b, 5159b, 5974b, 5976b x . . . Px . . . Px . . . S, x . . . Px . . . Px . . . P a-verse with double alliteration: 340a, 414a, 571a, 578a, 657a, 717a, 946a, 985a, 1144a, 1536a, 1730a, 1752a, 1810a, 1904a, 2284a, 2357a, 2444a, 2553a, 2658a, 3129a, 3144a, 3344a, 3349a, 3359a, 3425a, 3426a, 3432a, 3646a, 3825a, 4090a, 4129a, 4194a, 4267a, 4285a, 4309a, 4319a, 4419a, 4552a, 4836a, 5158a, 5214a, 5298a, 5466a, 5690a, 5973a; b-verse: 2467b, 5000b, 5755b

Of one hundred occurrences of the prototypical variant of type D1 without anacrusis (P#Pxx), nineteen (19 per cent) are a-verses with double alliteration, and fifty-six (56 per cent) are b-verses (section 2.7; cf. Table 2.80). By contrast, the weighty variants of type D, subtypes D2a and D2b, show a preference for the a-verse with double alliteration: without anacrusis, subtype D2a occurs fifty-five times (c. 64 per cent) in the a-verse with double alliteration, and nineteen times (c. 22 per cent) in the b-verse; subtype D2b occurs thirty-nine times (c. 56 per cent) in the a-verse with double alliteration, and twenty-two times (c. 31 per cent) in the b-verse. Despite

MOSP-02.qxd

5/11/04

2:38 PM

Page 165

2.11. Anacrusis

165

the avoidance of the a-verse with double alliteration and preference for the b-verse, subtype D1 is qualified for anacrusis in the Heliand, as are subtypes D2a and D2b characterised with the contrary verse distribution pattern. Third, types E and E* may be attended by anacrusis, as exemplified below (cf. Kauffmann 1887: 343–5; Heusler 1956: §262; Hofmann 1991, II: 201, 205–6, 208–12): (95) 732a an tuêm gêrun atogan 2333a is bedgiuuâdi te baka 3665a themu landes hirdie te lobe 4122a sô môdstarke man x . . . PSx(x)#(x . . .)P, x . . . P#Px . . . P a-verse with double alliteration: 483a, 631a, 635a, 732a, 1344a, 1471a, 1658a, 1721a, 1876a, 1994a, 2094a, 2219a, 2548a, 2591a, 2650a, 3132a, 3257a, 3270a, 3466a, 4101a, 4122a, 4284a, 4427a, 4730a, 4841a, 5096a, 5452a; a-verse with single alliteration: 159a, 345a, 1058a, 3446a, 4185a, 4527a, 4801a, 5095a, 5139a; b-verse: 71b, 103b, 294b, 376b, 639b, 997b, 1035b, 1083b, 1556b, 1644b, 2568b, 2856b, 2860b, 2882b, 3902b, 4035b, 4260b, 5268b, 5399b, 5705b x . . . PxSx . . . P, x . . . Px . . . Px . . . P a-verse with double alliteration: 268a, 2333a, 2652a, 2882a, 3036a, 3665a, 5804a; a-verse with single alliteration: 1555a, 2529a; b-verse: 1467b, 5664b Of the five variants of type E that we identified in section 2.9 above, the configurations PSx(x)#(x . . .)P and P#Px . . . P are definitely eligible for anacrusis, as they are unambiguously identified as type E with anacrusis. As regards the two configurations PX/x#x . . . P and PS#x . . . P, however, what might appear their anacrustic variants x . . . PX/x#x . . . P and x . . . PS#x . . . P should rather be scanned as type B1 primarily on the evidence of the common distribution pattern involved, as discussed at length in section 2.4.2.3 above. The remaining configuration x . . . PXx#(x)P is structurally indeterminate: it is amenable to being scanned as type E with anacrusis or type B1 on competing grounds, as shown in section 2.4.2.3 above. In this way, the five configurations of type E fall into three groups on the basis of the varying degrees of plausibility with which they may be scanned as anacrustic variants of type E: from most plausible (x . . . PSx(x)#(x . . .)P, x . . . P#Px . . . P) to ambiguous (x . . . PXx#(x)P) to least plausible (x . . . PX/x#x . . . P and x . . . PS#x . . . P). The varied plausibility is determined by the extent to which the configurations concerned come into conflict with the extant comparables that are identifiable as type B1. Thus, nothing inherent in anacrusis may be held responsible for the apparently limited range of its occurrence in type E. Fourth, in the Heliand anacrusis may be manifested by long syllable sequences, so much so that even verses with decasyllabic anacrusis are attested (Hofmann 1991: 67; cf. Kauffmann 1887: 308; Sievers 1893: §113): (96) 5754b Hie sagda simnen, that hie scoldi fan dôðe astandan 5807b sô huem sô ina muosta undar is ôgon scauuon Fifth, even when monosyllabically filled, anacrusis may be embodied by independent words, as well as by prefixes or ne. In fact, independent words are used

MOSP-02.qxd

5/11/04

166

2:38 PM

Page 166

2. Metrical types and positions

more widely than prefixes: for example, the type A1 variant Px#Px is expanded with anacrustic independent words (forty-five instances; e.g., 641a them uuîsun mannun) more frequently than with anacrustic prefixes (thirty instances; e.g., 2318a gelôbon habdun).31 In view of such a distributional pattern, we may conclude that prefixes do not serve as a prototypical source for anacrusis in the Heliand. The realisation of anacrusis by a long sequence of syllables and by independent words indicates that anacrusis may be realised by a concatenation of sentence particles in the Heliand (as exemplified in the extreme form in (96) above), in sharp contrast to Beowulf, in which anacrusis is largely confined to bound forms. In summary, in the Heliand anacrusis may occur freely before the first lift of any verses that begin with a lift (cf. Heusler 1956: §264);32 accordingly, type E, as well as types A1, A2, and D, may initiate with anacrusis. Obviously dissociated from the a-verse with double alliteration, anacrusis is capable of occurring in the b-verse as well as in the a-verse without appreciable distinction; and in the a-verse, it is as compatible with single alliteration as with double alliteration. Thus, anacrusis in the Heliand, in distinction from the one in Beowulf, cannot be characterised as an exponent of increased metrical strength, an exclusive accompaniment to the a-verse with double alliteration. And no longer limited to prefixes, anacrusis may be realised by an independent word and even a sequence of such words. As the foregoing examination has demonstrated, the distribution of anacrusis has been expanded in the Heliand to such an extent that it may come closest to appearing freely before the verse-initial lift without regard to the nature of given configurations on the one hand and of language materials used as anacrusis on the other. Anacrusis is thus no longer the device that a selected group of metrical types have the privilege of applying to a limited set of language material, as in Beowulf; thus, it scarcely differs from an ordinary drop, as Hofmann (1991: 67) correctly observes. In accordance with the extension of use, the status of anacrusis underwent profound change within the metrical system. Contrary to anacrusis in Beowulf, anacrusis in the Heliand constitutes an autonomous full metrical position in the underlying verse structure, rather than a derived realisation on the surface that lacks a corresponding presence as an independent position at the underlying level.

31

32

Independent words: a-verse with double alliteration: 137a, 1215a, 1892a, 2516a, 2875a, 3648a, 3676a, 3875a, 3887a, 5206a, 5472a, 5482a, 5682a, 5793a, 5945a; a-verse with single alliteration: 641a, 4739a, 5553a, 5668a, 5799a, 5800a, 5959a; b-verse: 40b, 101b, 415b, 631b, 671b, 744b, 875b, 917b, 2181b, 2330b, 2380b, 2389b, 2399b, 2547b, 2611b, 2764b, 3326b, 3352b, 4175b, 4999b, 5461b, 5549b, 5564b. Prefixes: a-verse with double alliteration: 1506a, 4844a; a-verse with single alliteration: 1257a, 2318a, 3077a, 3105a, 5407a; b-verse: 17b, 146b, 520b, 1361b, 1389b, 2128b, 2388b, 2413b, 3003b, 3582b, 3641b, 3849b, 3912b, 3923b, 4162b, 4667b, 4696b, 4807b, 5316b, 5543b, 5661b, 5675b, 5979b. Martin (1907: 32, 54), however, assumed that anacrusis occurred even before the first drop of types B1 and C. Such an assumption has to be rejected, principally because, as seemed apparent to Martin himself, we would then encounter an intractable problem of distinguishing between anacrusis and the first drop in a principled way.

MOSP-02.qxd

5/11/04

2:38 PM

Page 167

2.11. Anacrusis

167

Anacrusis in the Heliand should therefore be characterised as an optional drop (⫻) in verse-initial position, represented as in (97) below: (97) Type A1: (⫻) / ⫻ / ⫻ Type A2: (⫻) / ⫻ / \ Type D: (⫻) / / ⫻ ⫻ Type D*: (⫻) / ⫻ / ⫻ ⫻ Type E: (⫻) / \ ⫻ / Type E*: (⫻) / ⫻ \ ⫻ / Table 2.94 provides an overview of the distribution of anacrustic and nonanacrustic verses of the major variants of types A1, A2, D, D*, E, and E* in the Heliand. We may readily recognise that anacrusis may occur freely on any variants (apart from minor variants of type E) without regard to their particular configurational properties. Of particular interest is that incidence of anacrusis shows no significant correlation with the differing degrees of preference for the a-verse with double alliteration characteristic of particular realisation variants. Anacrusis is not only capable of occurring on the variants with a minimal preference for the a-verse with double alliteration, as observed above, but it may affect them as frequently as, or even more frequently than, the configurations with a stronger preference for the a-verse with double alliteration. A case in point is the treatment of the three variants of type A1, each with its first drop associated with two syllables, Pxx#Px, Px#xPx, and P#xxPx. As substantiated in section 2.1.3 above, the configuration Pxx#Px is distinguished from the other two by its notably low incidence of the a-verse with double alliteration. This variant, however, is expanded with anacrusis with higher frequency than the two others, the frequency that is nearly as high as that for the variant P#xxxxPx, which has a markedly strong preference for the a-verse with double alliteration. A comparison with Beowulf will be illuminating in this connection, as shown in Table 2.95. In accordance with the restriction of anacrusis largely to the configurations with marked preference for the a-verse with double alliteration, the variant P#xxPx, with its high incidence of the a-verse with double alliteration at the rate of approximately seven out of every ten, is accompanied with anacrusis with greater frequency than the configurations Pxx#Px and Px#xPx, both attested in the b-verse around 50 per cent of all occurrences (cf. Suzuki 1996a: 155). Another illustration of the insensitivity of anacrusis to verse distribution will be that subtypes D1, D2a, and D2b are undifferentiated as regards their susceptibility to anacrusis, although they occur in the a-verse with double alliteration with widely varying frequency. Specifically, subtype D1, notably the configuration P#PX/xx, is attended by anacrusis as frequently as subtypes D2a and D2b, particularly the configurations P#PS/sx and P#Px . . . S (Table 2.94), despite their diametrically opposed verse distribution patterning observed in section 2.7 above. In sharp contrast, the same configuration with a comparable propensity to the b-verse is scarcely compatible with anacrusis in Beowulf, while the other two may be used with anacrusis (Table 2.95), again in full compliance with the generalisation that anacrusis largely applies to configurations with marked preference for the a-verse with double alliteration.

MOSP-02.qxd

5/11/04

2:38 PM

Page 168

168

2. Metrical types and positions

Table 2.94. Distribution of major variants of types A1, A2, D, D*, E, and E* with and without anacrusis with anacrusis

without anacrusis

total

Type A1 PxSx Px#Px Psx#Px Pxx#Px P#xPx PX#Px PS#Px PXx#Px Px#xPx P#xxPx P#xxxPx P#xxxxPx PSx#Px PS#xPx

12 287 1 40 72 1 4 4 222 70 72 32 5 0

13% 24% 9% 35% 25% 4% 25% 17% 26% 28% 49% 37% 26% 0%

77 87% 925 76% 10 91% 73 65% 211 75% 24 96% 12 75% 20 83% 619 74% 183 72% 75 51% 54 63% 14 74% 16 100%

89 1212 11 113 283 25 16 24 841 253 147 86 19 16

100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%

Type A2 Px . . . PS

15

23%

51

77%

66

100%

Type D PS/sX/xx P#PX/xx P#PS/sx P#P#P/px P#Px . . . S P#Px . . . P

4 16 11 3 2 13

10% 14% 13% 21% 14% 18%

37 100 75 11 12 59

90% 86% 87% 79% 86% 82%

41 116 86 14 14 72

100% 100% 100% 100% 100% 100%

Type D* Px . . . PX/xx Px . . . PS/sx Px . . . P#P/px Px . . . Px . . . S Px . . . Px . . . P

34 31% 82 27% 46 44% 7 23% 41 32%

74 219 58 24 88

69% 73% 56% 77% 68%

108 100% 301 100% 104 100% 31 100% 129 100%

Type E PXx(x)#(x)P PSx(x)#(x . . .)P

58 32% 44 19%

121 187

68% 80%

179 100% 231 100%

8 34

62% 85%

Type E* Px . . . Sx . . . P Px . . . Px . . . P

5 6

38% 15%

13 40

100% 100%

MOSP-02.qxd

5/11/04

2:38 PM

Page 169

2.11. Anacrusis

169

Table 2.95. Distribution of major variants of types A1, A2a, A2b, D1, D2a, and D2b, and subtypes D*1, D*2a, and D*2b, with and without anacrusis in Beowulf with anacrusis

without anacrusis

total

Type A1 PxSx Px#Px P#xPx Pxx#Px Px#xPx P#xxPx P#xxxPx P#xxxxPx

0 2 15 1 6 13 3 1

202 100% 869 nearly 100% 305 95% 30 97% 528 99% 107 89% 6 67% 1 50%

202 871 320 31 534 120 9 2

Type A2a PS#Px

0

0%

144 100%

144 100%

Type A2b Px . . . PS

5

5%

103

95%

108 100%

Type D1 PSxx PPxx

0 0% 1 less than 1%

98 100% 243 nearly 100%

98 100% 244 100%

Type D2a PPS/sx PPP/px

8 9% 1 33%

79 2

91% 67%

87 100% 3 100%

Type D2b PPx . . . S PPx . . . P

2 6

5% 18%

40 27

95% 82%

42 100% 33 100%

Subtype D*1 Px . . . Pxx

1

2%

49

98%

50 100%

56 84% 2 100%

67 100% 2 100%

19 15

21 100% 18 100%

Subtype D*2a Px . . . PS/sx Px . . . PP/px Subtype D*2b Px . . . PxS Px . . . Px . . . P

0% less than 1% 5% 3% 1% 11% 33% 50%

11 16% 0 0% 2 10% 3 17%

90% 83%

100% 100% 100% 100% 100% 100% 100% 100%

Because anacrusis may thus involve any variants of the metrical types that begin with a lift, we have the proposed analysis confirmed whereby anacrusis in the Heliand constitutes an independent metrical position fully integrated into the structure of relevant metrical types, rather than a surface increment derivatively supplied to a limited set of realisation variants as in Beowulf. In addition to the overall applicability of anacrusis

MOSP-02.qxd

5/11/04

2:38 PM

170

Page 170

2. Metrical types and positions

Table 2.96. Distribution of types A1, A2, D, E, and E* with or without anacrusis

A1 aA1 A2 aA2

a-verse with double alliteration

a-verse with single alliteration

b-verse

total

1444 615

358 12% 94 9%

1199 40% 396 36%

3001 100% 1105 100%

48% 56%

56 98% 20 100%

1 0

3% 0%

D aD

118 25

40% 51%

54 18% 4 8%

D* aD*

411 189

88% 89%

18 4

E aE

120 27

50% 48%

E* aE*

29 7

69% 64%

4% 2%

0 0

0% 0%

57 100% 20 100%

123 42% 20 41%

295 100% 49 100%

36 19

8% 9%

465 100% 212 100%

55 23% 9 16%

66 27% 20 36%

241 100% 56 100%

5 12% 2 18%

8 19% 2 18%

42 100% 11 100%

scarcely depending on particular properties of variant configurations, we may adduce the following pieces of evidence in further support of the proposed analysis. First, as Table 2.96 makes evident,33 the distribution of anacrusis in the three verse categories (the a-verse with double alliteration, the a-verse with single alliteration, and the b-verse) is analogous to that for the corresponding types without anacrusis: anacrusis thus shows no significant correlation to distinct verse distribution. This fact seems to suggest that use of anacrusis scarcely affects the particularities of metrical structures involved. Anacrusis may accordingly be characterised as a purely optional element that may be mechanically added to the beginning of underlying metrical types. Second, as often pointed out (e.g., Heusler 1956: §260; Hofmann 1991: 67), and further substantiated in Table 2.97 below, longer forms of anacrusis occur with greater frequency in the b-verse than in the other two verse categories.34 The preference of the b-verse for longer sequences of unstressed syllables at the beginning of a verse is also exhibited by types B1 and C (for details, see Table 2.47, section 2.4.1, and Table 2.76, section 2.6.1). The sharing of this feature indicates that anacrusis is treated no differently from the initial drop in the Heliand; it accordingly lends further support to the interpretation that anacrusis constitutes a fully integrated underlying position in the Heliand much as do other metrical positions.

33

34

As far as type E is concerned, only the variants PSx(x)#(x . . .)P and P#Px . . . P are definitely expanded with anacrusis for the reasons adduced above. Accordingly, the remaining variants are excluded from the tabulation here. This fact empirically disproves Hinderschiedt’s (1979: 11) remark based on a limited sample. See also her unwarranted generalisation on the relation between size of drops and verse distribution, which will be refuted in section 2.12.3.2 below.

MOSP-02.qxd

5/11/04

2:38 PM

Page 171

2.11. Anacrusis

171

Table 2.97. Verse distribution according to the number of syllables used as anacrusis number of syllables

a-verse with double alliteration

a-verse with single alliteration

b-verse

Type A1 1 2 3 4 5 6 or more

471 97 27 16 3 1

72% 55% 26% 27% 7% 1%

32 5% 16 9% 31 30% 6 10% 3 7% 6 9%

148 64 46 38 40 60

Type D 1 2 3 4 5 6 or more

18 5 2 0 0 0

58% 56% 67% 0% 0% 0%

3 10% 1 11% 0 0% 0 0% 0 0% 0 0%

Type D* 1 2 3 4 5 6 or more Type E 1 2 3 4 5 6 or more Type E* 1 2 3 or more

141 92% 29 81% 15 100% 2 67% 1 33% 1 50% 17 7 2 0 0 1

3 1 0 0 0 0

2% 3% 0% 0% 0% 0%

total 23% 36% 44% 63% 87% 90%

651 177 104 60 46 67

100% 100% 100% 100% 100% 100%

10 32% 3 33% 1 33% 2 100% 1 100% 3 100%

31 9 3 2 1 3

100% 100% 100% 100% 100% 100%

6% 17% 0% 33% 67% 50%

153 36 15 3 3 2

100% 100% 100% 100% 100% 100%

26 16 8 2 0 4

100% 100% 100% 100% 0% 100%

9 6 0 1 2 1

65% 44% 25% 0% 0% 25%

3 12% 1 6% 3 38% 0 0% 0 0% 2 50%

6 23% 8 50% 3 38% 2 100% 0 0% 1 25%

6 75% 1 100% 0 0%

2 25% 0 0% 0 0%

0 0% 0 0% 2 100%

8 100% 1 100% 2 100%

The independent status of anacrusis as an optional but full metrical position constitutive of a metrical type sharply differentiates the Heliand from Beowulf. In Beowulf, anacrusis is a consequence of a prosodically motivated derivation, that is, procliticisation, whereby the first lift of maximally prominent verses (increased metrical types) is realised on the surface by a sequence of unstressed syllables and a

MOSP-02.qxd

5/11/04

172

2:38 PM

Page 172

2. Metrical types and positions

primary-stressed syllable (Suzuki 1996a: 337–40). In other words, anacrusis does not constitute a verse-initial drop on its own in an underlying metrical structure; rather, they are parasitic as it were on the verse-initial lift, arising as derived entities by virtue of an extended linguistic-metrical association of the first lift occasioned by procliticisation. 2.11.2. Anacrusis reconfigured: a diachronic account Having characterised anacrusis as a full metrical position in the Heliand metre from a synchronic perspective, we should now turn to its diachronic dimension. The question to be addressed then has to do with the dysfunctioning of anacrusis as an exponent of increased metrical types: why did anacrusis cease to be the privilege of increased metrical types and was consequently generalised as a full position qualified for attaching to any verses that begin with a lift? I would argue that the distinction between basic and increased metrical types that was fully in force in Beowulf became obscured in the Heliand to such an extent that it was no longer capable of providing a structural motivation for restricting occurrence of anacrusis in the way it did in Beowulf. As noted above, the traditional distinction at issue rested on the clustering of verse distribution and alliterative patterning: increased metrical types – types A2a, A2b, D2a, and D2b – were used for the most part in the a-verse with double alliteration in contrast to their basic counterparts, which failed to exhibit such a maximal preference for the a-verse with double alliteration. In Beowulf the increased metrical types occurred with overwhelming frequency in the a-verse and manifested double alliteration with rare exceptions (Suzuki 1996a: 278). In the Heliand, however, the distinction became blurred in large measure and was relegated to the periphery in its traditional form: of the four increased types, only type A2b survived as an independent configuration (type A2) in the Heliand, and the remainder was reorganised as graded variants of the basic types, types A1 and D (sections 2.1.7 and 2.7). Put another way, the traditional sharp categorisation between basic and increased metrical types was largely reconfigured and reintegrated as a gradient distinction within the unified metrical types, types A1 and D. Further evidence for the disintegration of the traditional distinction between basic and increased types is the changed distribution in the Heliand of types D and D*, with reference to their corresponding subtypes, subtypes D1 and D*1, D2a and D*2a, and D2b and D*2b, respectively, in comparison with the distribution of their structural analogues in Beowulf (that is, the basic and expanded variants D1 and D*1 of type D1, the basic and expanded variants D2a and D*2a of type D2a, and the basic and expanded variants D2b and D*2b of type D2b, respectively). In Beowulf, where the traditional categorisation fully obtained, the proportion of the expanded variant D*1 to the basic counterpart D1 was markedly lower than those of D*2a and D*2b to D2a and D2b, respectively: fifty-two occurrences (13 per cent) of the expanded variant (D*1) as against 342 of the basic one (D1), sixty-nine occurrences (43 per cent) of the expanded variant (D*2a) as against ninety for the basic one (D2a), thirtynine occurrences (34 per cent) of the expanded variant (D*2b) as against seventy-five of the basic one (D2b) (Suzuki 1996a: 104). As discussed in Suzuki (1996a: 103), the extremely low incidence of the variant D*1 may be ascribed to the absence of a heavy drop in type D1, the underlying configuration that served as a basis for the expanded variant at issue. By contrast, the proportion of subtype D*1 to subtype D1

MOSP-02.qxd

5/11/04

2:38 PM

Page 173

2.11. Anacrusis

173

increased dramatically in the Heliand, approaching that of subtype D*2b to subtype D2b, as shown in Table 2.83, section 2.8 above. Such a levelling in the incidence of subtype D*1 relative to subtypes D*2a and D*2b would seem to indicate that the traditional categorisation between types D1 and D2a/D2b came to be disintegrated and accordingly was rendered hardly capable of providing a viable basis for deriving expanded variants nearly exclusively from the increased types, that is, types D2a and D2b. The largely lost demarcation between basic and increased metrical types would then have undermined the structural basis for associating anacrusis with enhanced metrical strength: upon being reorganised as an extensively graded patterning in the Heliand, the traditional distinction in question lost the capacity of serving as a categorical basis for regulation. Anacrusis accordingly ceased to be the privilege of the increased metrical types; it was deprived of its distinctive function of marking increased metrical strength. As a consequence of the loss of the basis of restriction, anacrusis became accessible to all variants of types A1 and D, not only to the heavy variants of types A1 and D (types A2 and D2 in the traditional metre), but also to the light variants of types A1 and D. The marginalisation of the traditional opposition between basic and marked types further made extension of anacrusis to type E possible, a metrical type that had originally been exempt to that opposition and accordingly disallowed from its use. The extension of anacrusis to type E was facilitated by a further metrical development characteristic of the Heliand. As discussed at length in section 2.4.2.3 above, the Heliand exhibits a far larger number than Beowulf of the verse form x . . . PXx#(x)P, a configuration that is structurally ambivalent, equally scannable as a variant of type B1 as well as type E with anacrusis (Sievers 1893: 153; for details, see section 2.4.2.3 above). Some examples are reproduced below for ease of reference: (98) 3587a endi ôk uualdandes uuerk 4949a themu hêroston cûð 52b uuið fîundo nîð By contrast, there are at most only four comparable examples in Beowulf: Beo 501b, 932b, 949b, 1830b (Suzuki 1996a: 93). The Heliand thus obviously generalised the use of the originally rare disyllabic sequence -Xx# far beyond the limit allowed in Beowulf. (For the motivation and mechanism of the generalisation involved, see section 2.4.2.3 above.) It may be recalled that, since Beowulf did not permit type E to be expanded with anacrusis, it did not suffer the structural ambiguity that affected the Heliand: the traditional metre had no other choice than to scan the rare configuration x . . . PXx#(x)P as type B1. I would now like to claim that the structural ambiguity in scansion that inherently characterised the sequence x . . . PXx#(x)P in the Heliand would have promoted the derivation of type E with anacrusis in general and of the configuration x . . . PSx(x)#(x . . .)P in particular. As argued above, the obscuration of the distinction between basic and increased types made it no longer viable for anacrusis to mark increased metrical strength by procliticisation: it was reanalysed as an autonomous underlying metrical position that may optionally appear before the verse-initial lift

MOSP-02.qxd

5/11/04

174

2:38 PM

Page 174

2. Metrical types and positions

at the beginning of any verse. Accordingly, it became possible (but not mandatory as yet) for anacrusis to occur on type E, a metrical type that had traditionally been unaffected by the opposition between basic and increased types (Suzuki 1996a: 128–31, 317). Given such a restructuring of the underlying metrical system, the extensive (if still relatively marginal) use of the configuration x . . . PXx#(x)P as type B1 would have served as a source of analogy for further development. More specifically, by virtue of the structural ambivalence the configuration in question was now subject to an alternative scansion, whereby it was recharacterised as an anacrustic type E variant in accordance with the reconfiguration of anacrusis in the Heliand metre, although the rescansion did not fully replace the earlier scansion as type B1. It should be noted in this connection that type B1 as well as type E occurred predominantly in the b-verse as observed in Beowulf. Given such rescansion, it would have been only a short step to composing the sequence x . . . PSx(x)#(x . . .)P by further analogical extension, that is, by substituting the lexical-stressed syllable S for the non-lexical counterpart X, a mode of generalisation practised elsewhere. As may be recalled (section 2.4.2.3), the configuration x . . . PSx(x)#(x . . .)P was found hardly compatible with scansion as type B1. Accordingly, unless one had invoked extension of anacrusis to type E, this configuration could not have been accepted as metrical. The emergence of this configuration thus owed to the structural ambiguity of the configuration x . . . PXx#(x)P, which served as a catalyst as it were in making anacrusis reconfigured as an autonomous position accessible to the prototypical variant of type E (PSx(x)#(x . . .)P). Moreover, the hybrid nature of the configuration x . . . PXx#(x)P may provide an account of the fact that monosyllabic anacrusis did not occur with the greatest frequency for type E in the b-verse in contrast to the a-verse, as well as to the other types regardless of the distinction between the a-verse and the b-verse (see Table 2.97 above): because type E with anacrusis constituted a partial syncretism with type B1, and because type B1 was much preferred in the b-verse and least likely to begin with a monosyllabic drop (section 2.4.1), it is only natural that the frequency of monosyllabic upbeat failed to be maximal for the particular configurations at issue. The increased use of independent words as an anacrustic element may also be regarded as a consequence of the reanalysis of anacrusis as an underlying full position. Drops other than the verse-final one, especially those that appeared toward the beginning of a verse, were not subject to restriction as regards the quantity and quality of language materials they were permitted to be associated with. Accordingly, once reanalysed as a verse-initial drop and concomitantly dissociated from procliticisation, anacrusis lost its original limitation to prefixes as prototypical targets of the original process, and thereupon became capable of being filled by independent words, as well as by prefixes, and consequently of serving as a locus for sentence particles, as was the initial drop of types B1 and C. In conclusion, the foregoing discussion has shown that the extensive use of anacrusis in the Heliand, in terms of its distribution and language materials, is ascribed to the reconfiguration whereby it constituted a full metrical position that may be added at the head of any metrical types beginning with a lift; this reanalysis was made possible by the blurred categorisation between basic and increased metrical types in the Heliand metre, and also assisted by the increased use of the configuration x . . . PXx#(x)P in the same metre, originally as a variant of type B1 and subsequently as an anacrustic

MOSP-02.qxd

5/11/04

2:38 PM

Page 175

2.12. Paradigmatic and syntagmatic organisation of the metre

175

variant of type E by alternative scansion; and in the final analysis, the obscuration of the distinction in question, as well as the frequent occurrence of x . . . PXx#(x)P, may ultimately be attributed to the weakening of stress in Old Saxon (section 1.3).

2.12. The paradigmatic and syntagmatic organisation of the metre: a recapitulation 2.12.1. The system of metrical types On the basis of the foregoing investigations, we may identify the following ten configurations as significant metrical types for the Heliand metre: (99) Inventory of metrical types in the Heliand A1: / ⫻ / ⫻ A2: / ⫻ / \ A3: ⫻ ⫻ / ⫻ B1: ⫻ / ⫻ / B3: ⫻ ⫻ ⫻ / C: ⫻ / / ⫻ D: / / ⫻ ⫻ D*: / ⫻ / ⫻ ⫻ E: / \ ⫻ / E*: / ⫻ \ ⫻ / The metrical types beginning with a lift, that is, types A1, A2, D, D*, E and E*, may optionally be expanded with an extra drop or anacrusis at their opening, as follows (a ⫽ expanded with anacrusis): (100) aA1: ⫻ / ⫻ / ⫻ aA2: ⫻ / ⫻ / \ aD: ⫻ / / ⫻ ⫻ aD*: ⫻ / ⫻ / ⫻ ⫻ aE: ⫻ / \ ⫻ / aE*: ⫻ / ⫻ \ ⫻ / Since anacrusis is an optional constituent, and since the anacrustic verse forms occur far less frequently than the non-anacrustic counterparts, these configurations supplied with anacrusis are characterised as marked variants of the relevant metrical types. Taking the optional status of anacrusis into account, we may provide the following generalised representations, in which (⫻) designates an optional position (anacrusis): (101) A1: (⫻) / ⫻ / ⫻ A2: (⫻) / ⫻ / \ D: (⫻) / / ⫻ ⫻ D*: (⫻) / ⫻ / ⫻ ⫻

MOSP-02.qxd

5/11/04

2:38 PM

Page 176

176

2. Metrical types and positions E: (⫻) / \ ⫻ / E*: (⫻) / ⫻ \ ⫻ /

The other configurations consisting of more than four metrical positions, namely, types D* and E*, are minimally distinct from types D and E, respectively, by virtue of the presence of an additional drop in the second position (disregarding the optional anacrustic element mentioned in the last paragraph) immediately after the first lift; ‘*’ designates the presence of such an extra drop. In view of the close parallelism in structure, we may group types D and D* together on the one hand, and types E and E* together on the other, and represent each class (D and E) in generalised form by using (⫻) as a symbol for the extra drop in question, much in the same way as for anacrusis. We accordingly obtain the following common representations: (102) D: (⫻) / (⫻) / ⫻ ⫻ E: (⫻) / (⫻) \ ⫻ / By way of further reduction, types A1, A2, and A3 may be assigned to the same class (A) on the basis of the common configurational properties, as may types B1 and B3 to class B. The whole inventory of metrical types therefore reduces to the five metrical classes, A, B, C, D, and E, as follows: (103) Five metrical classes A: (⫻) / ⫻ / ⫻ B: ⫻ / ⫻ / C: ⫻ / / ⫻ D: (⫻) / (⫻) / ⫻ ⫻ E: (⫻) / (⫻) \ ⫻ / Representative of the five classes thus identified are their respective unmarked members, that is, types A1, B1, C, D, and E, which may be called basic metrical types. The marked members, that is, types A2, A3, B3, D*, and E*, as well as the anacrustic variants of relevant types, are derived by adjusting their unmarked configurations through additional operations, promotion, demotion, or expansion. Promotion and demotion change the status of a metrical position present in the basic types. By promotion, a normal drop is replaced by a heavy one. This process is responsible for deriving type A2 from type A1. The contrary process, demotion, replaces the first lift with a drop, thereby deriving types A3 and B3 from types A1 and B1, respectively. A third operation, expansion, supplies an optional drop in the ways described in the following paragraph, and concerns derivation of types D* and E*, as well as anacrustic variants of types A1, A2, D, D*, E, and E*. The entirety of metrical types including both basic and derived ones is thus integrated into the metrical system or paradigm as represented in Table 2.98 below. As mentioned above, an extra drop is supplied by expansion. Any metrical type opening with a lift may thereby be provided with an additional drop either before the first lift at the beginning of a verse or after the first lift at the second position: types A1, A2, D, E —⬎ aA1, aA2, aD, aE; types D, E —⬎ types D*, E*. The derivation by expansion thus falls into two processes, anacrusis formation and drop insertion at

MOSP-02.qxd

5/11/04

2:38 PM

Page 177

2.12. Paradigmatic and syntagmatic organisation of the metre

177

Table 2.98. The system of metrical types in the Heliand metre class

metrical type

status: additional operation:

unmarked

A B C D E

A1 B1 C D E

marked promotion A2

demotion

expansion

A3 B3 D* E*

the second place. In addition to supplying a drop at two different places in a verse, the two processes differ in their scope of operation, applying as they do to two different sets of metrical types. Anacrusis formation occurs on the basic and derived metrical types alike, while drop insertion involves exclusively the basic metrical types. In the strictest sense, then, only the second process concerns derivation of distinct metrical types, whereas anacrusis formation is responsible for generating variants of metrical types. Accordingly, the configurations expanded with anacrusis do not figure in the system of metrical types as represented in Table 2.98 above. Yet the two operations seem closely paralleled, as they both involve expansion of a configuration by an extra drop at two complementary places relative to the first lift. Such structural parallelism may therefore justify their common treatment here and in the following paragraphs. With two optional drops available, the maximal number of metrical positions allowed in a normal verse amounts to six, as embodied by types D* (⫻ / ⫻ / ⫻ ⫻) and E* (⫻ / ⫻ \ ⫻ /), each extended with anacrusis. This provides a structural basis for composing relatively longer verses than attested in Beowulf; and such an expansion of normal verses has also an important bearing on the composition of hypermetric verses, as will be considered in Chapter 4 below. On the other hand, since both types B1 and C begin with a drop, neither meets the condition for the occurrence of an optional drop and thus is affected by the expanding process at issue. Although optional, the formal status of these extra drops in the Heliand is entirely different from that of their counterparts in Beowulf. Both anacrusis and the extra drop of type D* in Beowulf are derivative (Suzuki 1996a: 24–35, 337–40): their occurrence is prosodically conditioned, by procliticisation (anacrusis) on the one hand and encliticisation (type D*) on the other. And the cliticisation in question is occasioned by the maximal strength of the strong position concerned, namely the first lift of increased metrical types (that is, types A2a, A2b, D2a, and D2b). The anacrustic elements and the extra drop are thus optional increments of the first lift of the verses involved; they are dependent on it, rather than aligned to separate positions on their own as in the Heliand. The absence of the configuration / ⫻ \ / ⫻ in the Heliand, which would come to being by supplying an extra drop to the configuration / \ / ⫻, is obviously attributed to the non-existence of type A2a (/ \ / ⫻; to use the term for the Beowulf metre) as

MOSP-02.qxd

5/11/04

178

2:38 PM

Page 178

2. Metrical types and positions

an independent type in contrast to type A2b (/ ⫻ / \; again in terms of the Beowulf metre), as discussed in section 2.1.1 above. The establishment of these optional elements as independent metrical positions in the Heliand brought in its train a radical increase of their employment in terms of their distribution and the range of language material put to use. The distribution of an optional drop after the first lift was thus maximised: now it was permitted to appear freely as a second position after the first lift, whereas earlier it had been limited to types D2a and D2b (subtypes D2a and D2b in terms of the Heliand metre), to the exclusion of type E. Furthermore, even those optional drops that had been allowed in Beowulf were subjected to generalisation in the Heliand: the light variant of type D, subtype D1, became as well-qualified for the augmentation at the second place as were the heavy ones, resulting in subtype D*1; proclitics, as well as enclitics, were permitted to occupy the additional drop in question; and anacrusis was allowed to involve a sequence of independent words, rather than limited to prefixes. Having determined the location in which these optional drops may occur, as well as the relaxation of requirements on their qualification, we need to address the question of their functional basis. The limitation of optional drops to the immediate proximity to the first lift, either before or after it, may remind us of similar conditions on the occurrence of unstressed elements in a verse, generally known as Kuhn’s Laws, a set of two rules on placement of unstressed syntactic elements in a verse (Kuhn 1933: 4–5, 15–23; Kendall 1991: 17; Suzuki 1996a: 55; Momma 1997: 56–75; see further section 3.2.2 below). Of primary interest here is the first law, which stipulates that as far as clause-initial verses are concerned, sentence particles – i.e., unstressed words that function as constituents of a sentence as a whole, such as substantive pronouns, many adverbs and finite verbs, and conjunctions (Kuhn 1933: 4–5; Kendall 1991: 1; Momma 1997: 56–64, 94–5) – have to be grouped together either before or after the first lift (exclusively). We thus see the same kind of privilege granted to the weak positions located at the immediate proximity to the first lift. The distribution of optional drops in the Heliand, however, cannot be fully reduced to the operation of Kuhn’s Laws. While anacrusis is often realised by sentence particles (section 2.11.1), the extra position after the first lift of types D* and E*, unlike the obligatory first drop of type A1, scarcely houses these syntactic elements; rather, it is largely occupied by word-internal bound forms (sections 2.8 and 2.10). In this light, the extensive use of an optional drop after the first lift should better be understood primarily in phonological and metrical terms: the neutralisation of the opposition between basic and increased types disrupted the prosodic foundation – cliticisation – of deriving the extra drop either before or after the first lift (cf. Suzuki 1996a: 24–35, 337–40), and therewith led to its occurrence by analogical extension before and after the first lift of whatever kind without any restriction on its qualification. While the expansion at the beginning of a verse by anacrusis may apply to the a-verse as much as to the b-verse, the occurrence of an extra drop at the second position after the first lift has notable consequences for verse distribution. That is, a vast majority of verses expanded at the second position occur in the a-verse with double alliteration. Such a strong preference for the a-verse with double alliteration is most clearly illustrated by the variant of type D*, subtype D*1 (Px . . . Pxx), in which the penultimate position is occupied by an unstressed syllable, such as 304a ôdan arbides.

MOSP-02.qxd

5/11/04

2:38 PM

Page 179

2.12. Paradigmatic and syntagmatic organisation of the metre

179

Table 2.99. Frequency of types A2, A3, and B3 in the Heliand and Beowulf Heliand A2 (A2b for Beowulf) A3 B3 total of a-verses

77 1% 397 7% 9 less than 1% 5985

Beowulf 132 4% 309 10% 5 less than 1% 3182

While the unexpanded variant, subtype D1 (P#Pxx), overwhelmingly occurs in the b-verse, the expanded counterpart is attested in the a-verse with double alliteration with high frequency (section 2.8). And much the same can be said of the other types that begin with a lift. Of related significance seems the strong preference for the a-verse with double alliteration that is exhibited by heavy variants of type A1 in which the first drop is realised by relatively long sequences of syllables, such as 302a halon imo te hîuuon (section 2.1.5). In this light, we can hardly avoid the conclusion that expansion of the drop immediately after the first lift has a significant correlation with the a-verse with double alliteration. The exact mode of expansion, however, is immaterial: while type A1 is expanded by associating the mandatory first drop with weighty language material, types D* and E* are expanded by inserting an optional drop after the first lift. As represented in (99) above, the Heliand has only two derived metrical types, apart from the extremely marginal configuration (type B3): one increased type, that is, type A2 on the one hand, and one reduced type, that is, type A3, on the other. By contrast, Beowulf has at its disposal a greater variety of increased metrical types, namely types A2a, A2b, D2a, and D2b. Thus, the Heliand is distinguished from Beowulf by the decreased variety of derived metrical types available. The limited role of derived metrical types in the Heliand concerns not only the range of the variety put to use, but also the frequency in occurrence of those types that are still in use. As shown in Table 2.99, the Heliand uses types A2 and A3, the increased type in particular, with markedly lower frequency than Beowulf. Furthermore, corresponding to the decrease of type A2, the variant of type A3 ending in the second element of a compound is much diminished in frequency (section 2.3). Without seeing the above difference in distribution in its proper perspective, we must not hasten to conclude that the Heliand obscured the structural opposition between the a-verse and the b-verse by largely disfavouring use of the metrical types that were traditionally limited to the a-verse in occurrence. Specifically, we have to recall at this point that the Heliand subsumed by reorganisation the expanded variants (i.e., subtypes D*1, D*2a, and D*2b) of types D1, D2a, and D2b in the traditional metre under a single independent metrical type, type D* (/ ⫻ / ⫻ ⫻; section 2.8); and furthermore that it introduced a novel expanded metrical type, type E* (/ ⫻ \ ⫻ /; section 2.10) to the system. These two expanded metrical types are characterised by the presence of an additional drop immediately after the first lift. Table 2.100 shows the frequency of these types in the Heliand and their corresponding configurations in Beowulf.

MOSP-02.qxd

5/11/04

2:38 PM

180

Page 180

2. Metrical types and positions

Table 2.100. Frequency of types D* and E* in the Heliand and their corresponding configurations in Beowulf Heliand D* (/ ⫻ / ⫻ ⫻) E* (/ ⫻ \ ⫻ /) total of a-verses

677 11% 53 1% 5985

Beowulf 160 5% 0 0% 3182

As should be clear from the above table, the incidence of the expanded configurations is increased in the Heliand. Given the preponderant occurrence of these expanded types in the a-verse (sections 2.8 and 2.10), we have to assume that the Heliand still continues to differentiate the a-verse from the b-verse by giving it access to special configurations scarcely available to the other. The Heliand, however, differs from Beowulf in its strategy for distinguishing between the a-verse and the b-verse: it marks out the a-verse primarily in quantitative rather than qualitative terms by allowing it to be expanded by an extra weak position, rather than by associating it with a heavier syllable. This extended use of quantitative means makes sense when we place it in the wider contexts of the reorganisation of the metre: for realisation of the drop in the Heliand, the association with a long sequence of unstressed syllables came to figure centrally as an exponent of added metrical strength at the expense of the association with a stressed syllable (see sections 2.12.2 and 2.12.3.1 below), as with the greater expansion of the first drop of type A1 in the a-verse and the significant increase in occurrence in the a-verse of subtype D*1. In regard to the obligatory first drop, we may note that the a-verse with double alliteration and the b-verse stand in contrast in terms of the preferred location of the first drop, as recognised in part by Rieger (1876: 62): the a-verse shows preference for the occurrence of the first drop after the first lift; by contrast, the b-verse exhibits marked preference for the verse-initial drop. Thus, types A1, D, and E are more frequent in the a-verse at the approximate rate of six or seven instances out of every ten, whereas the pattern reverses for types B1 and C, for which about seven verses out of every ten occur in the b-verse. Of further interest here is the correlation between the varying size of anacrusis and verse distribution (section 2.11.1). The longer the anacrustic elements are, the greater the frequency is with which they appear in the b-verse. We may accordingly generalise that the long verse-initial drops, regardless of their optionality, are favoured in the b-verse. The drop is realised through association with varied language material: it may be occupied by a number of syllables ranging between one and more than ten; and it may be filled by a syllable with varying degrees of stress. Such a wide range of variability in the realisation of the drop, however, is sensitive to its position within the verse. This syntagmatically conditioned variability may be generalised as follows: the earlier the drop occurs in a verse, the greater the range of material used for its realisation. Accordingly, the verse-initial drop is maximal in its potential for linguistic realisation. The maximality of the verse-initial drop is demonstrated among other things by the fact that the commonest number of syllables involved in this position is two rather than one, as elsewhere. At the other end of the extreme,

MOSP-02.qxd

5/11/04

2:38 PM

Page 181

2.12. Paradigmatic and syntagmatic organisation of the metre

181

the minimality of the verse final drop finds the fullest expression in the requirement that it be filled only by a single syllable, with the occasional exception of a disyllabic ending -xx#. Between these two polarities, the verse internal drops may be realised by a limited range of material to choose from. Such a gradation pattern of the realisation of the drop may be explained by assuming that the Linearity-based Metrical Strength Scale (section 1.2) was inherited and maintained as a valid generalisation in the Heliand metre (see further section 2.12.3.1 below). 2.12.2. The blurred distinction between normal and heavy drops The traditional opposition between normal and heavy drops was neutralised in the metre of the Heliand to such an extent that it remains in force only for the distinction between types A1 (/ ⫻ / ⫻) and A2 (/ ⫻ / \). By contrast, the traditional types A2a (/ \ / ⫻), D2a (/ / \ ⫻), and D2b (/ / ⫻ \) no longer constitute independent metrical types: the corresponding configurations are now reorganised as realisation variants of types A1 (/ ⫻ / ⫻) and D (/ / ⫻ ⫻), respectively. These three configurations fail to show their original distinctive feature in full, namely maximal preference for the a-verse with double alliteration. In the reorganised metre, this originally categorical feature of the heavy metrical types figures as no more than an optional (though frequently occurring) marker of relatively heavier variants that are involved in gradient distinction from lighter counterparts. The dysfunctioning of the opposition between the two kinds of drop was brought about by the interplay of phonological and metrical factors. The weakening of stress and concomitant relative levelling of stressed and unstressed syllables (section 1.3) in the phonology would have obliterated much of the perceptual basis for the traditional distinction involved. Complementary with this general phonological development was a complex of metrical innovations, which were primarily concerned with the reorganisation of the traditional metre into a more integrated system permeated with finely and harmoniously graded patterning. This complex of phonological and metrical factors would then have significant bearing on diction by presumably discouraging the Heliand poet from making extensive use of poetic compounds; and the disfavoured use of such compounds in the Heliand would in turn have further disrupted the tradition of maintaining the stressed yet metrically weak position (i.e., the heavy drop) as an autonomous unit. As a consequence, the opposition between normal and heavy drops was relegated to the periphery, involving as it did only the differentiation between types A1 and A2; and the resultant neutralisation was implemented in such ways that, in accordance with general expectation, the unmarked member, the normal drop, was realised at the expense of the marked counterpart, the heavy drop. The loss of the opposition between normal and heavy drops radically expanded the availability of language materials to choose from, thereby bringing about increased incidence of originally marginal instances, on the one hand, and making possible the creation of metrical forms that had not been allowed in the traditional metre, on the other. Most important, having abandoned the original restriction of the heavy drop to a stressed syllable as well as the limitation of the normal drop to unstressed syllables, the Heliand metre permits the strings -Sx# and -Xx# to appear in contexts that were originally inaccessible: -Sx# and -Xx# are now qualified to constitute the first drop of type A1; and -Xx# is made increasingly compatible with constituting the second drop of type B1, while -S# and -sx# occur as less marginal

MOSP-02.qxd

5/11/04

182

2:38 PM

Page 182

2. Metrical types and positions

variants; -Sx# and -Xx# may realise the first drop of type D*, as may -S#. Furthermore, -S# is occasionally allowed to constitute the second drop of type C. Despite the much expanded variety of materials to be used for the normal drop in general, the second drop of types B1 and C is not occupied by a stressed syllable as often as is the first drop of type A1, for example. While conforming fully to the generalisation based on the Linearity-based Metrical Strength Scale (section 1.2; see also section 2.12.3 below), such limited increase in association also stems from the fact that originally there were no configurations like ⫻ / \ / and ⫻ / / \, that is, heavy variants of types B1 and C in which the second drop constitutes a heavy drop. Without a stock of proper verse expressions to model after, the poet would have encountered greater difficulty in composing such type B1 and C verses as have their second drop realised by a stressed syllable. 2.12.3. The distinction between verse-initial and non-initial normal drops As demonstrated above, the opposition between normal and heavy drops was largely replaced by the gradient distinction of a single category, namely, the (normal) drop. Unlike the heavy drop, the normal drop may be realised by a varying number of syllables. Yet the range of variability permitted depends on a particular position in which a given drop appears within the verse, in accordance with the Linearity-based Metrical Strength Scale (section 1.2). Put another way, the normal drop is subject to context-sensitive differentiation primarily in terms of the number of syllables it may be associated with, and secondarily in terms of the prosodic properties of the syllables involved. By far the most outstanding context-dependent distinction is the one between initial and non-initial drops. 2.12.3.1. Quantitative and qualitative distinctions While the most frequently occurring number of syllables associated with the noninitial drop is one, the verse-initial drop is realised by a disyllabic or trisyllabic sequence with the greatest frequency. The distinction is demonstrated by comparing the first drop of types B1 and C with that of type A1, the non-initial drop that may be occupied by a larger number of syllables than any others. Partly drawing on the data presented earlier (Tables 2.47 and 2.76), we may determine the range of variability and the average value for each drop on the basis of the varying numbers of syllables it is realised by, as indicated in Table 2.101. Thus, the verse-initial drop (types B1 and C) is realised on average by a larger number of syllables than is the non-initial drop (type A1) that exceeds other non-initial ones in terms of the number of syllables associated. A question arising then is: why is the commonest number of syllables involved in the verse-initial drop larger than elsewhere? By invoking the Linearity-based Metrical Strength Scale (section 1.2), we may claim that the verse-initial drop is inherently stronger than any other drop, and therefore that maximal strength requires a greater amount of prominence in realisation, which is guaranteed by a larger number of syllables in association. While the verse-initial drop is distinguished from non-initial ones primarily in quantitative terms, that is, by the highest average number of syllables involved in association, the non-initial drops located in verse-internal positions are partly differentiated in qualitative terms as well, that is, by the quality of syllables concerned. A case in point concerns the distinction between the first drop of type A1 (second

MOSP-02.qxd

5/11/04

2:38 PM

Page 183

2.12. Paradigmatic and syntagmatic organisation of the metre

183

Table 2.101. Distribution of types A1, B1, and C according to the numbers of syllables in the first drop number of syllables type A1 1 2 3 4 5 6 7 8 9 10 11

1628 (40%) 1405 (34%) 522 (13%) 334 (8%) 160 (4%) 48 (1%) 9 (less than 1%) 0 (0%) 0 (0%) 0 (0%) 0 (0%)

total

4106 (100%)

average

2.07 per verse

type B1 139 (4%) 609 (19%) 769 (24%) 661 (21%) 485 (15%) 278 (9%) 111 (4%) 66 (2%) 19 (1%) 4 (less than 1%) 1 (less than 1%) 3142 (100%) 3.80 per verse

type C 187 (10%) 450 (25%) 415 (23%) 303 (17%) 200 (11%) 115 (6%) 67 (4%) 30 (2%) 11 (1%) 10 (1%) 2 (less than 1%) 1790 (100%) 3.44 per verse

position) and the second drop of type B1 (third position) in regard to the treatment of medial and word-final syllables. As demonstrated in sections 2.1.2 and 2.4.2.3 above, the disyllable -Sx# is allowed as a realisation of the first drop of type A1, but disallowed from the second drop of type B1. Further, while differentiated from -S#, the disyllable -sx# (beginning with a short stressed syllable) is hardly distinguishable from the unstressed counterpart -xx# as an occupant of the first drop of type A1; by contrast, the same sequence -sx# is treated as analogous to -S# as a marginal occupant of the second drop of type B1, as opposed to the commoner variants -xx# and -Xx# without lexical stress. Thus, while determining the central membership of the second drop of type B1, the presence versus absence of lexical stress hardly figures as a significant factor for the characterisation of the first drop of type A1, for which syllable quantity (that is, syllable length and the number of syllables) serves as an overarching principle of organisation. This means that the first drop of type A1 is capable of being realised by syllables with lexical stress as well as those without, and the presence of lexical stress hardly affects the acceptability of particular realisations of the position in question. The functioning of the number of syllables as a primary parameter for determining the gradation of the first drop of type A1 (second position) shows parallelism with the patterning of the verse-initial drop, for which it serves as an exclusive parameter. It would seem no coincidence that lexical stress becomes increasingly significant in conditioning the differentiation of the drop towards the end of a verse. The second drop of type B1 is located in the third position, for which stress figures more centrally, as treated in the above paragraph. The role of lexical stress culminates at the end of a verse, where it provides the crucial criterion for binary distinction, namely for distinguishing between normal and heavy drops. Conversely, the role of

MOSP-02.qxd

5/11/04

184

2:38 PM

Page 184

2. Metrical types and positions

the number of syllables diminishes towards the end of a verse, as the range of language material available for use becomes correspondingly restricted. Thus, the number of syllables and lexical stress constitute two complementary parameters that are involved in a trading-off relation predicated on linearity. 2.12.3.2. Correlation with verse distribution Although the initial drop is in general distinguished from others by its greatest prominence realised in terms of the highest average number of syllables involved as discussed above, finer-grained distinctions of drops are induced by the opposition between the a-verse and the b-verse. That is, the a-verse and the b-verse exhibit distinct preference patterns for the size of the initial and non-initial drops realised in terms of the differing number of syllables, as indicated in Table 2.102. As Table 2.102 makes evident, the average number of syllables occupying the normal drop differs along two parameters: (i) the location in the verse, most notably the distinction between verse-initial and non-initial drops; (ii) the distinction of the a-verse and the b-verse. The generalisation predicated on the first parameter, that is, the Linearity-based Metrical Strength Scale (section 2.12.3.1), remains unaffected by the second parameter of the verse distinction: the verse-initial drop, whether in the a-verse or the b-verse, is occupied by the largest number of syllables on average; and similarly but less conspicuously, the drop located in the second position is realised by the higher average number of syllables than is the one in the third position. Of immediate interest in the present context, however, is the role of the second parameter. Specifically, as far as the verse-initial drop is concerned, the b-verse contains a larger number of syllables than the a-verse (as in types B1 and C). By contrast, the converse obtains for the non-initial drops: the a-verse has it realised by a larger number of syllables than the b-verse, as shown most evidently in the first drop of type A1, but in the second drop of type B1, the first drop of type D*, and the second drop of type E to a limited extent as well. Accordingly, the verse-initial and non-initial drops are involved in complementary distribution as to the size and prominence of their realisations: in the a-verse, the relatively prominent non-initial drop is favoured at the expense of the initial counterpart; in the b-verse, on the other hand, the verse-initial drop is given the privilege of carrying a relatively great amount of prominence at the expense of the non-initial one. This divergence between the initial and non-initial drops invalidates Hinderschiedt’s (1979: 11, 26) undifferentiated treatment of drops (and anacrusis; see section 2.11 above) and her subsequent sweeping claim that drops are occupied by a smaller number of syllables in the b-verse for types A, D, and E and in the a-verse for types B and C. The generalisation put forward above seems to apply to an optional drop in part as well. The extra drop of types D* and E* (that is, the non-initial drop standing between the lifts) may be realised by a relatively prominent syllable or a syllable sequence in the a-verse; by contrast, the position at issue is normally avoided in the b-verse. In other words, following most strictly the stipulation that the b-verse prefers the initial drop to the non-initial, the optional drop in question is largely kept to a minimum, that is to say, unrealised in the b-verse. In contrast to the extra drop of types D* and E*, anacrusis, the other optional drop, may appear in the a-verse and the b-verse, and with higher frequency in the a-verse at that (section 2.11.1). In view of the sharply differentiated treatment of

MOSP-02.qxd

5/11/04

2:38 PM

Page 185

2.12. Paradigmatic and syntagmatic organisation of the metre Table 2.102. Distribution of the varying number of syllables in the non-final normal drop in relation to the distinction between the a-verse and the b-verse number of syllables

a-verse

b-verse

type A1 (second position) 1 2 3 4 5 6 7 8 total average

651 26% 883 35% 450 18% 316 13% 155 6% 47 2% 9 less than 1% 0 0% 2511 100% 2.45 per verse

977 61% 522 33% 72 5% 18 1% 5 less than 1% 1 less than 1% 0 0% 0 0% 1595 100% 1.47 per verse

type B1 (first position) 1 2 3 4 5 6 7 8 9 10 11 total average

87 12% 219 29% 209 28% 139 18% 56 7% 19 3% 17 2% 5 1% 1 less than 1% 0 0% 0 0% 752 100% 3.02 per verse

52 2% 390 16% 560 23% 522 22% 429 18% 259 11% 94 4% 61 3% 18 1% 4 less than 1% 1 less than 1% 2390 100% 4.04 per verse

type B1 (third position) 1 2 3 4 5 total average

540 72% 194 26% 12 2% 5 1% 1 less than 1% 752 100% 1.32 per verse

1689 71% 658 28% 41 2% 2 less than 1% 0 0% 2390 100% 1.31 per verse

type C (first position) 1 2 3

63 192 167

11% 34% 29%

124 258 248

10% 21% 20%

185

MOSP-02.qxd

5/11/04

2:38 PM

186

Page 186

2. Metrical types and positions

number of syllables

a-verse

b-verse

4 5 6 7 8 9 10 11 total average

84 15% 31 5% 18 3% 7 1% 6 1% 0 0% 0 0% 0 0% 568 100% 2.89 per verse

219 18% 169 14% 97 8% 60 5% 24 2% 11 1% 10 1% 2 less than 1% 1222 100% 3.70 per verse

type D (third position) 1 2 total average

186 93% 15 7% 201 100% 1.07 per verse

139 97% 4 3% 143 100% 1.03 per verse

type E (third position) 1 2 3 4 total average

178 64% 93 33% 7 3% 1 less than 1% 279 100% 1.39 per verse

119 80% 30 20% 0 0% 0 0% 149 100% 1.20 per verse

the extra drops of types D* and E* in terms of their verse distribution, however, one might expect that anacrusis should tend to be avoided in the a-verse: according to the strict interpretation of the above generalisation, the initial extra drop ought to be kept to a minimum in the a-verse, precisely as is the non-initial counterpart in the b-verse. This apparent exceptionality of anacrusis would prove to be less deviant, however, when we are reminded of its near-exclusive association with the a-verse in the traditional metre, as discussed in section 2.11.1 above: anacrusis was nearly an exclusive attribute of the a-verse. The still higher incidence of anacrusis in the a-verse in the Heliand metre may thus reasonably be understood as a vestige of the earlier practice, not as a structural contradiction to the current metrical system, in which anacrusis is increasingly used in the b-verse. Furthermore, the distribution of the varying number of syllables used as anacrustic elements is found fully compatible with the generalisation at issue: the relatively large size of anacrusis shows preference for the b-verse, as substantiated in Table 2.97, section 2.11.1. This trading relation between the a-verse and the b-verse in regard to the size of the normal drop may also be held responsible for the near complementarity in the distribution of certain metrical types in relation to the distinction between the a-verse

MOSP-02.qxd

5/11/04

2:38 PM

Page 187

2.12. Paradigmatic and syntagmatic organisation of the metre

187

and the b-verse. Notably at stake are the two most frequently occurring metrical types, types A1 and B1, more specifically the preponderance of type A1 in the a-verse on the one hand and that of type B1 in the b-verse on the other. These two types are complementary in terms of the location of the first drop: it appears in non-initial (second) position in type A1 and in initial position in type B1. Thus, type A1 may be characterised by the minimum of the verse-initial drop (anacrusis excepted), type B1 by that of the non-initial. Given such a characterisation, it would follow as a matter of course that the a-verse shows a marked preference for type A1 at the cost of type B1, and conversely that the b-verse chooses type B1 over type A1 as an optimal metrical type. 2.12.4. On the treatment of the disyllabic ending -xx# The disyllabic ending -xx# is distinguished from the monosyllabic one -x# on the one hand, and the regular disyllabic string -xx# on the other. It counts as more prominent than the latter two by virtue of its marked morphological status of disyllabicity. This accounts for the stronger preference for the a-verse with double alliteration of the type A1 and type B1 variants, Pxx#Px and x . . . Pxx#P, than the variants Px#Px and Pxx#Px, and x . . . Px#P and x . . . Pxx#P, respectively (sections 2.1.7.1 and 2.4.1). The disyllabic ending -xx# is always treated as a unit by being aligned to a single metrical position, with rare exceptions treated below. In this respect, it is differentiated from other disyllabic strings consisting of a medial and a final syllable -xx#, which are separated from each other in verse-final and penultimate positions to constitute different metrical positions, as in types C (e.g., 68a that im uuârun sô gihôriga) and D (e.g., 2084a drôm drohtines). In other words, the ordinary sequence -xx# may not fill the verse-final drop. By contrast, the disyllabic ending is unqualified to occupy separate drops by virtue of the morphological integrity involved: it is incapable of constituting the final two positions of type C (⫻ / / ⫻); nor can it realise the last two positions of type D (/ / ⫻ ⫻). Accordingly, the sequence Px . . . Pxx (e.g., 917a hêrron mînumu) instantiates type A1 (/ ⫻ / ⫻), rather than type D* (/ ⫻ / ⫻ ⫻); by the same token, the sequence x . . . Pxx (e.g., 2471a talod imu, huô he ôdrana) scans as type A3 (⫻ ⫻ / ⫻), rather than type C (⫻ / / ⫻). Insofar as the disyllabic ending is unqualified to make two positions and consequently has to behave as a unit as an occupant of a single drop, it is characterised as equal to the word-final unstressed syllable -x#. When preceded by a syllable lacking in lexical stress, however, the disyllabic ending receives special treatment, and accordingly ceases to be fully parallel with the word-final syllable -x#. Specifically, the disyllabic ending is treated divergently depending on the length of the immediately preceding syllable, -X- or -x-. First on the treatment after a long syllable: there are a number of verses of the form PXxx#P, as exemplified below: (104) 3061a hêleandero bezt 5858a hettiandero hand 2635b menniscono barn a-verse with double alliteration: 4915a, 5858a; a-verse with single alliteration: 2031a, 2180a, 3061a, 3558a, 4031a, 5218a, 5601a; b-verse: 2635b, 2678b, 2811b, 3156b, 3606b, 3630b, 4036b

MOSP-02.qxd

5/11/04

2:38 PM

188

Page 188

2. Metrical types and positions

The configuration PXxx#P, as exemplified above, scans as / \ ⫻ /, type E, rather than / ⫻ \ /, an otherwise unattested verse form. The scansion accordingly leads us to infer that the disyllabic ending -xx# is less prominent than the preceding long monosyllable X. In more precise terms, the word form PXxx# has the following foot structure: (105) F F g

g

P X x_x_ Thus, the medial long syllable, constituting a foot on its own by virtue of being bimoraic in non-final position, is qualified to realise a heavy drop. By contrast, though bimoraic, the disyllabic ending is incapable of forming a foot, much as is the monosyllabic counterpart. The sequence PXxx#P thus proves to be no different from PXx#P in its essential prosodic properties. In sharp contrast to the sequence PXxx#P is the minimally distinct configuration Pxxx#P, as follows: (106) 2804a craftigana Crist As discussed in section 2.9 above, this example should best be scanned by associating with the first drop the two syllables -tiga-, that is, by splitting the morphologically integrated disyllable -gana#, in line with other instances of Pxxx#P which do not involve a disyllabic ending, such as 610a craftagoron cuning. The sequence Pxxx is therefore assumed to be organised in terms of foot constituency as follows: (107) F g P x

F 56 x_ x_

The second foot is thus formed over the second syllable (x) and the following one (x) by dissociating the latter from the word final. This is actually one of the few examples attested that are subject to dissociation of a disyllabic ending. At this point, we may note what seem to be further instances of the splitting of the disyllabic ending. Of particular interest are the following two verses of the form P(x)#Pxxx: (108) 2689b he ni uuas iro cûð ênigumu 4593b Ni uuas thero thegno ênigumu Since the word ênigumu cannot be regarded as a compound (true or quasi), it does not carry a lexically assigned, secondary stress. Hence it may follow that the above two verses scan as subtypes D1 (P#PX/xx) and D*1 (Px . . . PX/xx), respectively,

MOSP-02.qxd

5/11/04

2:38 PM

Page 189

2.12. Paradigmatic and syntagmatic organisation of the metre

189

rather than as subtypes D2a (P#PS/sx) or D2b (P#Px . . . S) and D*2a (Px . . . PS/sx) or D*2b (Px . . . Px . . . S). The first drop would accordingly be occupied by the first two syllables -nigu- of the trisyllabic sequence -nigumu, rather than -ni- alone. Otherwise, the last two syllables would be left occupying the final position, which would then carry greater prominence than the penultimate position. Although at issue here is the subcategorisation of a metrical type, not a categorisation in terms of metrical types, we witness here again the splitting of the disyllabic ending, whereby the first constituent serves as the right branch of the foot that is headed by the immediately preceding short syllable. One might still claim that the disyllabic ending as a whole has such a low degree of prominence as to be suitable for occupying a normal drop. On this view, the disyllabic endings in verses 2689b and 4593b given in (108) above are associated with the last drop as a unit, yet these verses constitute subtypes D1 and D*1, respectively. Following a similar line of reasoning, verse 2804a in (106) would be analysed as constituting type E (/ \ ⫻ /), rather than an unmetrical configuration / ⫻ \ /, although the disyllabic ending involved is allegedly associated as a unit with the third position. The above alternative view has to be rejected, however. Of particular significance in this regard are verses 610a craftagoron cuning and 929a ênhuuilic ni bist treated in section 2.9 above. These verses contain an unstressed short trisyllabic sequence xxx. Yet what is mapped onto the first drop of type E has to be a disyllable xx, with the final one x aligned with the second drop; otherwise we would be led to the scansion whereby the first drop is occupied by less prominent material (x) than the second (xx) in violation of the metrical strength pattern of type E (/ \ ⫻ /). Moreover, the complete absence of a disyllable -xx# as a self-sufficient realisation of the two drops of type E (section 2.9) may demonstrate that the two consecutive drops have to be sharply distinguished in terms of prominence as determined by foot construction: the first and second drops have to be occupied by footed and unfooted material, respectively, unless suspension of resolution is invoked in exceptional circumstances as will be specified in section 3.1.1.6 below. All considered then, to align the disyllabic ending with the second drop of type E would lead to the undesirable consequence that the disyllabic ending -xx# has to be regarded as definitely less prominent than the preceding unstressed short syllable -x-. More specifically, we would have to postulate the following implausible representation in terms of foot structure for the sequence Pxxx: (109) F F g g P x x_x_ The above representation is anomalous due to the monomoraicity of the second foot, standing at variance with the bimoraicity requirement for the foot. Alternatively, in order to satisfy the bimoraicity requirement, one might assume the second foot to be branching, thereby dominating the disyllabic ending on the right. This analysis, however, would seem no more viable, because the resulting complex foot consisting of

MOSP-02.qxd

5/11/04

190

2:38 PM

Page 190

2. Metrical types and positions

three syllables -xxx# would occupy the first drop as a whole, with the devastating consequence of the second drop remaining unoccupied. Thus, only in extremely limited contexts is the disyllabic ending -xx# decomposed for prosodic reasons: it is split so that the first member may constitute the right branch of the foot that is headed by the immediately preceding, lexically unstressed short syllable. In such marked circumstances, then, the sequence Pxxx# is categorically distinguished from Pxx#: the string Pxxx# consists of two feet and an unfooted syllable as represented in (107) above, whereas the string Pxx# comprises two feet exhaustively without any unfooted material. The prosodic distinction is clearly shown by the presence of Pxxx#P versus the absence of Pxx#P as a type E verse. Since the word form Pxxx# is infrequent in occurrence, however, the dissociation of the morphologically integrated unit is a highly marked process, and consequently the verse form Pxxx#P constitutes an extremely marginal variant of type E. 2.12.5. On the treatment of quasi compounds Standing at the two extremes among medial and word-final syllables are the second element of a compound on the one hand and a part of a simple (noncompound) word on the other. The two entities are sharply distinguished by the presence versus absence of lexical stress, and accordingly they are metrically differentiated by the categories that they are prototypically associated with: the heavy drop on the one hand and the normal drop on the other. Somewhat ambivalent against this dichotomy, however, is the second element of a group of words designated here as quasi compounds (section 2.2.1), exemplified by unreht ‘injustice’ and gôdlîk ‘goodly’. These words are obviously decomposable into two constituents with appreciable meanings for each (un⫹reht; gôd⫹lîk); yet they are distinguished from true compounds due to the less than full autonomy of the constituents involved: they may not appear independently (e.g., un- ‘un-’) or their meanings are markedly different from those of their independent counterparts (e.g., -lîk ‘-ly’ versus lîk ‘body’). In corresponding fashion, these words are subject to varying metrical treatment: in some contexts they behave analogous to true compounds, in others they seem scarcely distinguishable from non-compound words. Throughout the preceding discussions, however, we have tacitly treated quasi compounds as equivalent to true compounds. We are accordingly required at this point to see that our identification is confirmed in large measure and to provide qualifications where needed. To this end, in this section we focus on such positiondependent varying treatments of the second elements of quasi compounds and provide an explanatory account for the mechanism of their metrical variation. Of central concern are the metrical positions for which the presence of lexical stress (rather than the number of syllables) serves as a differentiating feature (see section 2.12.1 above). For subtype A1s (PS#px), the quasi compound, like the true compound, may occupy the first two positions, as exemplified below (section 2.1.8). The non-compound word (PX# or Px#), however, is categorically excluded from these positions. Thus, the quasi compound is treated equivalent to the true compound PS#, in distinction from the non-compound PX#/Px#.

MOSP-02.qxd

5/11/04

2:38 PM

Page 191

2.12. Paradigmatic and syntagmatic organisation of the metre

191

(110) 1698a mênful maco 410b unrîm cuman a-verse with double alliteration: 1475a, 1698a, 1801a; b-verse: 410b, 1468b, 5281b With respect to the composition of type A2 (/ ⫻ / \), the second element of a quasi compound is as capable of serving as the heavy drop as is that of a true compound, and accordingly associated with the a-verse with double alliteration: all thirteen verses ending in a quasi compound and therefore scannable as type A2 are a-verses with double alliteration. In this respect, the quasi compound PS# is sharply differentiated from the non-compound of the form PX#, which fails to induce maximal occurrence of the a-verse with double alliteration (section 2.2.1). (111) 1024a sâlig sinlîf 3897a ubiles anmôd a-verse with double alliteration: 848a, 1024a, 1393a, 1638a, 1846a, 1891a, 2005a, 2892a, 3897a, 4343a, 4492a, 5209a, 5252a As regards the association with the second drop of type B1, too, the second element of a quasi compound (-Sx#) is found sharply distinguished from the medial and final syllables of a non-compound word (-Xx#). While the sequence -Xx# qualifies for association with the position in question (with the attendant structural ambiguity characteristic of this configuration; section 2.4.2.3), the string -Sx# may not occupy it; as a consequence, in contrast to the configuration x . . . PXx#(x)P, the configuration x . . . PSx#(x)P is identified as type E with anacrusis, rather than as type B1. The distinction is substantiated by the diametrically opposed patterns of distribution characteristic of the two configurations, as shown in Table 2.103. A radically different situation obtains for the composition of type C (section 2.6.2). The quasi compound (PSx#, Psx#) is hardly distinguished from the non-compound word (PXx#, Pxx#): type C verses containing a quasi compound in the last three positions appear in the a-verse with single alliteration with a frequency as high as those having a non-compound in the same positions. By contrast, type C verses with a true compound are slightly favoured in the b-verse, a characteristic which is Table 2.103. Distribution of x . . . PSx#(x)P (PSx# ⫽ quasi compound; type E with anacrusis) and x . . . PXx#(x)P (PXx# ⫽ non-compound; scannable as type B1)

quasi compound non-compound

a-verse with double alliteration

a-verse with single alliteration

b-verse

5 62.5% 7 12%

1 12.5% 8 14%

2 25% 43 74%

x . . . PSx#(x)P a-verse with double alliteration: 1876a, 2650a, 3270a, 4284a, 4841a; a-verse with single alliteration: 4527a; b-verse: 1035b, 2856b x . . . PXx#(x)P (see under Table 2.66, section 2.4.2.3 above)

MOSP-02.qxd

5/11/04

2:38 PM

Page 192

192

2. Metrical types and positions

Table 2.104. Distribution of type C verses containing true compounds, quasi compounds, and non-compounds

true compound quasi compound non-compound

a-verse with double alliteration

a-verse with single alliteration

b-verse

2 1% 2 2% 4 2%

141 43% 87 78% 120 72%

182 56% 23 21% 42 25%

Table 2.105. Distribution of type D verses containing true compounds, quasi compounds, and non-compounds

true compound quasi compound non-compound

a-verse with double alliteration

a-verse with single alliteration

b-verse

49 67% 22 81% 20 17%

8 11% 3 11% 26 22%

16 22% 2 7% 70 60%

manifested most saliently when the last two positions are occupied by an independent word. Thus, we witness here that the quasi compound is treated akin to the noncompound in marked contrast to the true compound. The appearance of a quasi compound in the last three positions of type D (/ / ⫻ ⫻) shows more pronounced preference for the a-verse with double alliteration than that of a true compound (section 2.7), so much so that more than eight out of every ten constitute a-verses with double alliteration. When a non-compound word is used instead, however, the pattern is reversed with the predominant incidence of the b-verse. An analogous generalisation applies to the composition of type E verses: the quasi compound is diametrically opposed to the non-compound. More specifically, the occurrence of the second element of a quasi compound in the second position (heavy drop) strongly favours use of the a-verse with double alliteration, while that of a constituent of a non-compound is observed in the b-verse with a comparable frequency. As with type D, but in a more conspicuous way, the quasi compound exhibits a stronger correlation with the a-verse with double alliteration than the true compound. As pointed out in section 2.9 above, the true compound induces the a-verse with double alliteration with the rate of slightly less than one out of every two. Thus, the quasi compound, rather than the true compound, is characterised by the maximal occurrence of the a-verse with double alliteration. In summary, the second element of a quasi compound is generally treated on a par with the second element of a true compound, in categorical distinction from the medial and final syllables of a non-compound word. There is a single exception to this generalisation, however: for the composition of the second lift of type C, the element in question behaves indistinguishable from the constituent of a non-compound. The question arising then is what motivates such divergent metrical treatment. The second lift of type C stands out from the other metrical positions under consideration here: it is a lift, while the remaining positions are drops (heavy or normal). The

MOSP-02.qxd

5/11/04

2:38 PM

Page 193

2.12. Paradigmatic and syntagmatic organisation of the metre

193

Table 2.106. Distribution of type E verses containing true compounds, quasi compounds, and non-compounds

true compound quasi compound non-compound

a-verse with double alliteration

a-verse with single alliteration

b-verse

87 44% 36 69% 26 20%

49 25% 7 13% 42 32%

63 32% 9 17% 63 48%

distinction between these two types of position is based on metrical strength or prominence: the lift is distinguished from the drop by virtue of a greater degree of metrical strength. Accordingly, the lift is prototypically realised by a primarystressed syllable, the heavy drop by a secondary-stressed syllable, and the normal drop by an unstressed counterpart. Given such mapping relations predicated on the presence of lexical stress, the second element of a quasi compound is not considered prominent enough to realise the maximally strong position, the lift, because it is less autonomous in morphological status than the second element of a true compound, which is fully lexical. Thus, the lack of a full lexical stress makes it infeasible to use a quasi compound as a proper realisation of the last three positions of type C involving two lifts (/ / ⫻), resulting thereby in the metrical indistinguishability from a noncompound. Insofar as the association with the less salient positions is concerned, however, the relative independence of the second element of a quasi compound is treated as analogous to that of the corresponding constituent of a true compound, and accordingly regarded as having a comparable amount of prominence.

MOSP-03.qxd

5/11/04

2:41 PM

Page 194

3 Resolution and alliteration: repatterning and reconstitution

This chapter treats resolution (section 3.1) and alliteration (section 3.2), the two major metrical devices for adding prominence to the lift in the traditional metre. I shall show that these devices underwent profound change in the Heliand metre in the specifics of their operation, and investigate underlying motivations for the metrical reconstitutions that the Heliand poet accomplished as he reorganised the inherited metre.

3.1. Resolution A first device for heightening prominence is resolution, an association of a strong position with a disyllabic foot (px or sx). The term ‘resolution’ stems from the observation that a strong position may be regarded as ‘resolved’ by being simultaneously linked with two different syllables, rather than being mapped in a one-to-one correspondence to a single stressed syllable as in ordinary cases (cf. Hofmann 1991: 61). 3.1.1. Suspension of resolution Given a disyllabic foot, not every instance of a strong position is resolved through association with the disyllable as a unit; rather it may be left unresolved, so that it is occupied exclusively by the first constituent syllable (i.e., the short stressed syllable) of a disyllabic foot. This inoperation of resolution is called suspension of resolution. Suspension of resolution is a context-dependent phenomenon and seems to concern the following configurations (the bold-faced p and s indicate that they occupy a single position on their own through suspension of resolution): PS#px (subtype A1s); x . . . P#px (type C); x . . . Psx (type C); Psxx (subtype D1); P#P#px (subtype D2a); P#Psx (subtype D2a); PsxP (type E). The following sections (3.1.1.1 through 3.1.1.5) examine each of these configurations in comparison with their minimally distinct ones, with special reference to the word forms involved in suspension or operation of resolution. The collected data will in turn be marshalled to explore the formal and functional basis of the opposition between operation and suspension of resolution in sections 3.1.1.6 and 3.1.2 further below. 3.1.1.1. The second lift of subtype A1s (PS#px) As shown in section 2.1.8 above, the second lift of type A1, when preceded by the second element of a compound, remains unresolved. Such a configuration, in distinction

MOSP-03.qxd

5/11/04

2:41 PM

Page 195

3.1. Resolution

195

from the sequence PS#Px, may be called subtype A1s (a short variant of type A1). There are thirty-nine examples of the configuration PS#px, as listed under Table 2.37, section 2.1.8 above (cf. Kauffmann 1887: 297; Hofmann 1991, II: 59–60), and the word forms used as the second constituent px in these verses are as follows: (1) 703a slîðmôd cuning 2382a friðubarn godes 519b uuilspel mikil Word forms realising px of PSpx cuman (410b); cuning (703a; 630b); frumid (5194b); geban (5281b); godes (2870a; 2382a, 5776a; 983b, 3883b, 4525b); hugis (1468b); lahad (1851b); maco (1698a); manag (1703a; 413a, 5366a, 5565a; 876b, 1288b, 1367b, 1842b); mikil (792a, 1220a, 1827a, 2734a, 4322a; 759a, 5836a; 519b, 978b, 1619b, 3918b); sehan (1475a, 1801a); sprekan (4590a); tholon (2604a, 3392a); uueron (5942a) By contrast, there are only four instances of the configuration PS#pxx, in which the second lift is resolved through association with the disyllable px: (2) 3122a berthlîc biliði 1077b That friðubarn tholode 1802b uuârlîc biliði (also 2427b) The vanishing rarity of the resolved configuration PS#pxx (four instances as against thirty-nine instances of the unresolved counterpart PS#px) sharply differentiates the Heliand from Beowulf, in which the corresponding ratio is thirty-three (PS#pxx) to sixty-seven (PS#px), as listed below: (3) PS#pxx in Beowulf 215a, 328a, 330a, 467a, 522a, 715a, 767a, 819a, 994a, 1047a, 1171a, 1198a, 1205a, 1239a, 1267a, 1476a, 1534a, 1558a, 1602a, 1607a, 1670a, 1802a, 1852a, 2043a, 2072a, 2154a, 2607a, 2616a, 2979a, 3006a, 3041a; 1317b, 1906b PS#px in Beowulf 67a, 69a, 120a, 252a, 284a, 406a, 629a, 643a, 657a, 776a, 786a, 817a, 881a, 973a, 1015a, 1180a, 1256a, 1432a, 1672a, 1682a, 2035a, 2191a, 2588a, 2947a, 2959a, 3000a, 3172a; 64b, 303b, 619b, 838b, 994b, 1112b, 1275b, 1287b, 1288b, 1289b, 1310b, 1457b, 1510b, 1731b, 1807b, 1828b, 1834b, 1896b, 1914b, 1925b, 1964b, 2007b, 2060b, 2110b, 2158b, 2256b, 2265b, 2334b, 2417b, 2430b, 2457b, 2460b, 2613b, 2754b, 2906b, 2969b, 2972b, 3019b, 3081b, 3135b Thus, while in Beowulf the configurations PS#px and PS#pxx are fully opposed, in the Heliand the opposition comes closest to being neutralised so that only the unresolved string px is allowed to occur after PS in type A1. We may generalise accordingly that suspension of resolution is obligatory for the second lift of type A1 when the first drop is realised by the second element of a compound. This is a radical reorganisation of the rule that is in force in Beowulf: in it, operation of resolution is obligatory for the same position, and subtype A1s is derived by a special operation without invoking suspension of resolution (Suzuki 1996a: 232).

MOSP-03.qxd

5/11/04

196

2:41 PM

Page 196

3. Resolution and alliteration

There is good reason for the near loss of the opposition between the two configurations PS#px and PS#pxx in the Heliand. Earlier in the traditional metre, as in Beowulf, the shorter string PS#px was identified as type A1 (/ ⫻ / ⫻), while the resolved counterpart PS#pxx was categorised as type A2a (/ \ / ⫻). The opposition between px and pxx after PS was thus associated with the categorical distinction in metrical types. As discussed at length in section 2.1.7 above, however, the traditional distinction between types A1 and A2a was lost in the Heliand: type A1 covered a wide spectrum of graded variation in its realisation on the surface, whereas with loss of type A2a, type A2 in the Heliand constituted only a reflex of type A2b in the traditional metre. Upon disruption of the categorical distinction, the traditional opposition between PS#px and PS#pxx would have been deprived of its formal basis: since there was only one metrical type involved, the variation between px and pxx would hardly have lent itself to a rule-governed differentiation. It would thus have been chaotic to allow for the two distinct forms in the same context without encoding them with proper conditions for differentiation. The question arising then is: why did the resolved string, rather than the unresolved counterpart, come to be avoided after the sequence PS? Why was px rather than pxx favoured as a normal string after PS? The retention of px after PS seems well-motivated on structural grounds: the difference in markedness involved in the traditional opposition may account for the divergent treatment of these two strings. In traditional practice, type A2a was a marked entity in opposition to type A1 as indicated by its significantly stronger preference for the a-verse with double alliteration (Suzuki 1996a: 375). Given such a difference in markedness, we may expect in view of the mechanism of neutralisation in general that the unmarked member persists while the marked counterpart falls into disuse. Accordingly, the shorter string px, which was a constituent of the unmarked opposite, type A1, was chosen as a legitimate string after PS.1 Of related interest is the total absence of the unresolved configuration Px#px (section 2.1.8): we find no example like Beo 1828b hwilum dydon (Suzuki 1996a: 83–6) in the Heliand. By contrast, the resolved configuration Px#pxx is attested frequently: we count seventy-six such verses. Similarly, the unresolved configuration P#xpx is absent, while the resolved counterpart P#xpxx is counted thirty-eight in occurrence. In (4) below, these resolved configurations Px#pxx and P#xpxx are exemplified, and all word forms used as the second constituent pxx are listed: (4) 62a gicoran te kuninge 2482a nâhor mikilu 3551b liohtes tholodun Word forms realising pxx of Pxpxx

1

Projecting the Heliand metre back to the traditional counterpart, we may be justified in making the following inference: the differential treatment of the two configurations PS#pxx and PS#px in the Heliand metre may testify that they were originally kept distinct, that is, they constituted different metrical types in the traditional metre. This reasoning may then falsify Sievers’s (1893: 33) characterisation of the configuration PS#px as a short variant of type A2a (A2ak) by subsuming it under type A2a (rather than type A1), to which the configuration PS#pxx also belongs as a long variant (A2al).

MOSP-03.qxd

5/11/04

2:41 PM

Page 197

3.1. Resolution

197

berege (5534a); betara (1486b); biboda (5662b); biliði (4647b); cumana (1265a, 2028a, 2729a, 3427b, 5610b); cumanne (3298b); dunida (5799a); faranne (4782a, 5656a); feridun (2915b); firihon (5863a); firihun (1392a); fratahun (380a, 3331a, 3763a, 4543a); fratohon (1738a); frumana (4802b); frumidon (732b); frumidun (180a); gareuuea (1857a); gebanne (3739a); huerigin (591b); gehuuilikas (1592b); gehuuilikes (1218a, 1253a, 1917a, 2169a, 2347a, 3333a, 3336a, 3628a, 3781a, 1670b, 2284b, 3913b); himila (213b, 864b); himile (656a, 3288a, 4367a, 661b, 1489b); huilicumu (4842a, 1964b); idisi (5828b); kumana (4458b); kumane (3632b, 4466b); kuninge (62a); lacanes (5497b); lobodun (2875a, 417b); macode (241a); managa (737a, 441b, 743b, 903b, 1006b, 1111b, 1866b, 2736b, 2818b); manage (966b, 1214b); managun (1764b); menegi (3715a, 4473a, 2751b, 3711b); menigi (1974a, 2833a, 5482a); mikilu (182a, 2482a, 3770a, 4584a, 5825a); sehanne (3128a); soleri (4542a); spilode (2764b); tuelibi (2820b); thenidun (1155b); -tholoda (5290b, 5301b); tholode (3346a, 3379b, 5050b, 5171b); tholodun (1346b, 3551b, 3590b, 4522b); -thologian (1534b); tholoian (5015a, 1351b, 5216b); ubile (1525a); -uuarahta (5417a); uuatare (1001a); uuatares (2040b, 4536b); uuatere (979a, 3047a); uuateres (3369a); uueroldi (246b); uulitigost (271a); uuonodun (3959a); uuritanan (5789a) As shown by comparison of the two configurations Pxpxx and PS#px, attested in comparable numbers, the same lexical items are involved in resolution as well as in suspension of resolution: e.g., cuman ‘come’, cuning ‘king’, frummian ‘carry out’, geban ‘give’, manag ‘many’, mikil ‘great’, tholon ‘endure’. We are therefore led to suspect that operation of resolution is not conditioned by whatever inheres in individual lexical items. Hence it may be reasonably assumed that resolution is determined by metrical structure: more specifically, we may generalise that resolution has to be implemented on the second lift of type A1 when it is preceded by an unstressed syllable. 3.1.1.2. The second lift of type C (x . . . P#px; x . . . Psx) A vast majority of type C verses leave their second lift unresolved when it is associated with a short stressed syllable (cf. Hofmann 1991: 147). The configurations x . . . P#px (512 examples; exemplified in (5) below, and listed under Table 2.77, section 2.6 above) and x . . . Psx (183 examples; exemplified in (6) below, and listed under Table 2.77, section 2.6 above) thus occur with by far greater frequency than their resolved counterparts x . . . P#pxx (fifteen examples, as listed in (7) below) and x . . . Psxx (ten examples, as listed in (8) below). In other words, the unresolved configurations x . . . P#px and x . . . Psx account for c. 97 per cent and 95 per cent, respectively, of the type C verses with the second lift occupied by a primary-stressed short syllable on the one hand, and by a secondary-stressed short syllable on the other. We may accordingly generalise that resolution is normally suspended from the second lift of type C. (5) x . . . P#px 94a Thô uuarð thiu tîd cuman 2143a thar sie iro torn manag 1374b ef he im than lâtid is môd tuuehon

MOSP-03.qxd

5/11/04

198

2:41 PM

Page 198

3. Resolution and alliteration

(6) x . . . Psx 2727a habdun ina for uuârsagon 4157a than al thit liuduuerod 100b that sie hebancuning (7) x . . . P#pxx (Underlined are examples with the resolved first lift, i.e., x . . . px#pxx.) 2701a hêt that sie frume fremidin 1732b endi spel managu a-verse with double alliteration: 2701a, 5371a; b-verse: 1197b, 1530b, 1732b, 1813b, 2517b, 2712b, 2752b, 2945b, 3483b, 3586b, 4268b, 5873b, 5959b Word forms realising pxx of x . . . P#pxx beteran (3483b); cumana (5873b); fremidin (2701a); frumidi (2712b); gibiðig (3586b, 4268b); huilican (1530b); managa (1197b, 5959b); managu (1732b); niðana (1813b); sorogot (2517b); tugiðos (2752b); tuehode (2945b); uuerodes (5371a) (8) x . . . Psxx (Underlined are examples with the resolved first lift, i.e., x . . . pxsxx.) 1893a fora uueroldcuningun 928b ef thu tharo forasagono a-verse with single alliteration: 1893a; b-verse: 568b, 928b, 1422b, 2154b, 2473b, 2511b, 2518b, 4114b, 5280b Word forms realising sxx of x . . . Psxx -cuninge (568b, 2154b, 2473b, 2511b, 5280b); -cuninges (2518b); -kuninges (4114b); -cuningun (1893a); -sagono (928b, 1422b) As with the pair of type A1 variants, Px#pxx and PS#px, treated in the last section, a number of lexical items used with resolution in x . . . P#pxx and x . . . Psxx are also attested in their unresolved forms in the configurations x . . . P#px and x . . . Psx, such as cuman (94a), -cuning (100b), frumid (1825b), manag (2143a), -sagon (2727a), tuuehon (1374b), -uuerod (4157a), as given in (5) and (6) above. This again indicates that implementation and suspension of resolution does not reside in lexical choice. The distribution pattern of the resolved and unresolved variants of type C in the Heliand is no different from the one obtained in Beowulf: in it, too, the unresolved strings px and sx occur as the third and fourth positions of type C at the expense of the resolved counterpart pxx and sxx, with comparable ratios of distribution (Suzuki 1996a: 180). In addition to being extremely rare (twenty-five instances altogether), the resolved configurations x . . . P#pxx and x . . . Psxx show another notable feature: 80 per cent of the relevant examples receive resolution on the first lift as well, as catalogued in (7) and (8) above. Thus, much as in Beowulf (Suzuki 1996a: 200), the sequences x . . . px#pxx and x . . . pxsxx are much more favoured to the sequences x . . . P#pxx and x . . . Psxx in which the first lift is realised by a long stressed syllable. With respect to implementation and suspension of resolution on the second lift of type C, we may accordingly rank the three variants under consideration (x . . . P#px/

MOSP-03.qxd

5/11/04

2:41 PM

Page 199

3.1. Resolution

199

x . . . Psx, x . . . px#pxx/x . . . pxsxx, x . . . P#pxx/x . . . Psxx) in terms of their preference as follows: (9) x . . . P#px/x . . . Psx 傻 x . . . px#pxx/x . . . pxsxx 傻 x . . . P#pxx/x . . . Psxx 3.1.1.3. The second lift of subtype D1 (Psxx) With subtype D1, the unresolved configuration Psxx (as given in (10) below; cf. Kauffmann 1887: 334) stands in contrast to the resolved counterpart P#pxxx (given in (11) below): (10) 464b aðalboranan 2443b hebencuninges Word forms realising sxx of Psxx -berandero (765b); -boranan (464b); -boranes (222b); -cuninge (1472a, 82b, 278b, 1120b, 2767b, 2981b), -cuninges (130a, 2114a, 91b, 362b, 781b, 1939b, 2344b, 2377b, 2443b, 4258b, 5575b) (11) 3029b that uuîf fagonode 2244b uueros sorogodun Word forms realising pxxx of P#pxxx fagonoda (526b); fagonode (3029b); soragodun (5789b); sorogodun (2244b); uuacogeandi (384a) As it turns out, the lexical status of the second lift seems of vital importance in determining operation and suspension of resolution: when the position at issue is occupied by the second element of a compound, resolution is suspended, and the configuration Psxx is accordingly obtained; on the other hand, when the position in question is filled by a primary-stressed syllable of an independent word, resolution is implemented, resulting in the configuration P#pxxx. In this way, the resolved and unresolved configurations are involved in complementary distribution with respect to the stress value of the second lift: the resolved configuration P#pxxx and the unresolved configuration Psxx are attested to the exclusion of their unresolved and resolved counterparts, P#pxx and Psxxx, respectively. Such a distribution pattern obtained in the Heliand differs from the one observed in Beowulf. In it, there are instances in which the second element of a compound is subject to resolution: e.g., Beo 744a unlyfigendes (for further examples, see Suzuki 1996a: 225). A further notable feature in the Heliand is that, when the second lift is suspended from resolution, the first one is nearly always resolved: except for verses 765b and 2767b, all the examples given in (10) above contain the resolved first lift (aðal-, heben-). The heavier variants P#pSx and P#p#Px, with the second lift suspended from resolution, are non-existent as expected, because a lexical word may not end in a short stressed syllable (Suzuki 1996a: 239–40). Instead, the resolved configuration P#pxSx is attested numerously, as exemplified below (for the statistics, see Table 3.4, section 3.1.4 below): (12) 1041a hôh himilrîki 3059a Crist cuning êuuig 5443a thera idis opanlîco

MOSP-03.qxd

5/11/04

200

2:41 PM

Page 200

3. Resolution and alliteration

As regards subtype D2b, neither of the configurations P#pxS (unresolved) and P#pxxS (resolved) occurs in the corpus; nor are their expanded counterparts Px . . . pxS and Px . . . pxxS attested. Turning to the expanded counterpart of subtype D1, subtype D*1 has its second lift invariably resolved: the configurations Px . . . pxx is absent in favour of the configuration Px . . . pxxx, as shown in (13) below. In the light of the correlation between resolution and primary stress pointed out above, the second lift of subtype D*1 has to be resolved, because it is always occupied by an independent word. (13) 760a flôdo fagorosta Word forms realising pxxx of Pxpxxx faganondi (5982a); fagorosta (760a); gadulingos (1266a) The resolved heavier configuration Px . . . pxSx, subtype D*2a, occurs frequently, as exemplified in (14) below (for the statistics, see Table 3.4, section 3.1.4 below), while the unresolved counterpart Px . . . pSx is excluded for the same reason as noted above concerning the absence of the configuration P#pSx. (14) 659a folgodun ferahtlîco 2115a holde heririncos 3277a fulgangan godes lêrun 3.1.1.4. The first drop of subtype D2a (P#P#px; P#Psx) At issue here is subtype D2a, in which the first drop is associated with an independent word or the second element of a compound. While the configurations P#P#px and P#Psx appear frequently as shown in (15) below (cf. Kauffmann 1887: 334–5), the resolved counterparts P#P#pxx and P#Psxx are totally absent. Thus, as in Beowulf, the third position of subtype D2a would appear categorically immune to resolution in the Heliand. (15) 1163a faran folc manag 3044a uuîs uuârsago 4703a hêr hebancuning Closer consideration will convince us, however, that the first drop of type D has nothing to do with suspension of resolution. As will be shown in section 3.1.1.6 below, resolution by its nature applies to the positions that are prototypically occupied by a lexical foot (Suzuki 1996a: 239–47). Accordingly, the lift and the heavy drop are qualified for resolution. On the other hand, the normal drop, which is prototypically realised by nonlexical material, is unaffected by resolution. Such a categorical exclusion of the normal drop from resolution is thus held responsible for the complete absence of it in the third position of type D (/ / ⫻ ⫻; section 3.1.1.4).2 As discussed in section 2.7 above, type D 2

At first sight, verses 2722a unsundigane and 2689b he ni uuas iro cûð ênigumu would appear to have the first drop resolved through association with the disyllabic sequence -diga- and -nigu-, respectively. This cannot be true, however. Since these syllable sequences lack lexical stress, and since resolution operates on strong positions (the lift and the heavy drop), these examples should be regarded as instances of a multiple association of unstressed syllables. See discussion in section 3.1.1.6 below.

MOSP-03.qxd

5/11/04

2:41 PM

Page 201

3.1. Resolution

201

in the Heliand metre has a normal rather than a heavy drop in its third position: the heavy drop constitutive of type D2a (/ / \ ⫻) in the traditional metre was integrated with the normal counterpart of type D1 (/ / ⫻ ⫻) through reconfiguration in the Heliand. Thus, while the configurations P#P#pxx and P#Psxx are non-existent in the Heliand (subtype D2a) as well as in Beowulf (type D2a), the structural reasons for their exclusion are different: in the Heliand resolution is inapplicable because the position in question is a normal drop; in Beowulf, on the other hand, implementation of resolution is ruled out because suspension of resolution is mandatory in the context at issue, i.e., the heavy drop of type D2a (Suzuki 1996a: 176–7). 3.1.1.5. The first drop of type E (PsxP) The first drop of type E may be involved in resolution or suspended from resolution: attested in the corpus are the resolved configurations Psxx#(x . . .)P, Psx#xP, and P#px#xP, as given in (16) below, as well as the unresolved counterpart Psx#P, as given in (17) below (cf. Kauffmann 1887: 343). Also occurring are the configurations Pxxx#P and Pxx#xP as given in (18) below, which are characterised as resolved on the second position, as argued in section 2.9 above. By contrast, the minimally distinct counterpart Pxx#P is absent, also as observed in section 2.9 above. (16) 537a hebencuninges hugi 4935a uuârsagono uuord 2700b langsamane râd Word forms realising Psxx of Psxx#(x . . .)P, Psx of Psx#xP, or px of P#px#xP folcskepi (3001a); friðugumono (619a); hebancuninges (317a); hebencuninges (159a, 537a, 997b, 1461b, 2087b); heritogono (2735a); langsamane (2700b); langsamoron (1202a); mênscaðono (5491a); uuamscaðon (2993a); uuamscaðono (5427b); uuamscaðun (3033a); uuârsagono (631a, 4935a); uuârsaguno (3399a); uueder (2914a); uueroldkuninges (2893a) (17) 1098a uuonodsaman uuelon 1742a uuînberi uuesan 3697b orlegas uuord (18) 610a craftagoron cuning 929a ênhuuilic ni bist 2804a craftigana Crist As in Beowulf (Suzuki 1996a: 177–8), the second position of type E would appear ambivalent with respect to implementation of resolution: sometimes it is resolved, other times it is not. Unlike Beowulf, however, the Heliand shows outstanding preference for resolution: while we count twenty-three examples of the resolved configurations as listed in (16) and (18) above, there are only three verses that are suspended from resolution, as shown in (17) above (cf. Russom 1998: 166).3 By contrast, 3

We may recall at this point that the configuration x . . . Psx#P has to be scanned as type B1 rather than type E with anacrusis, as discussed at length in section 2.4.2.3 above. At stake here are verses 3968a, 5827a, 3940b, 5369b, 5727b, 5793b, 5875b, 5886b.

MOSP-03.qxd

5/11/04

202

2:41 PM

Page 202

3. Resolution and alliteration

in Beowulf the proportion of the unresolved configuration Psx#P to the resolved counterparts Psxx#P and Psx#xP is nearly even, with the rate of six to seven (Suzuki 1996a: 177).4 Thus, the Heliand seems to be characterised by the distribution pattern that favours the resolved configurations at the expense of the unresolved counterpart. The configuration Psx#P thus has to be regarded as marked, and suspension of resolution on the second position of type E is accordingly a highly exceptional phenomenon. Because Hofmann (1991: 69) allows for the heavy drop to be resolved only for the string Pxsx such as iungarskepi, he does not regard the twenty-three examples listed in (16) and (18) above as resolved on the second position of type E. Accordingly, he has no principled means for distinguishing the configurations PsxP and PsxxP as far as the realisation of the first drop of type E is concerned: in both, the position in question is identified as being occupied by the short stressed syllable alone. The difference between the two configurations, then, would reside in the way the following position (the second drop) is realised, monosyllabic or disyllabic. Such a scansion, however, would entail implications empirically questionable. Since the normal drop is prototypically realised by a monosyllable, one would expect that the configuration PsxP is more natural than PsxxP, and should thus occur more frequently, a prediction that proves to be contrary to the fact and therefore disproves Hofmann’s analysis. The questions arising then are whether there is any structural regularity involved in the seemingly ambivalent treatment of resolution for the second position of type E in the Heliand, or whether the choice between implementation and suspension of resolution is a matter of pure accident; and what was the motivation and mechanism for the reorganisation whereby the Heliand came to favour resolution for the position obviously in defiance of traditional practice. Before treating these questions fully in the following sections, we should note a few further points of detail concerning the distinction between implementation and suspension of resolution for the first drop of type E. First, there is no instance of the unresolved verse in which the second position is occupied by an independent word: the configuration P#pxP is absent in contrast to the resolved one P#pxxP (2914a hôh uueder afhaben). We may suspect then that suspension of resolution is correlated in part with relatively weak prominence on the affected syllable, and that a primary-stressed syllable is less susceptible to suspension of resolution. In Beowulf, too, we find an instance of the configuration P#pxxP, but none of the unresolved counterpart P#pxP. A second point concerns implementation of resolution on different positions. Resolution on the first drop of type E seems to occur largely when at least one of the lifts is also resolved. Out of the twenty verses involving resolution on a lexicalstressed syllable as given in (16) above, fourteen (c. 70 per cent) are subject to resolution at least on one lift, and there are six exceptions to this generalisation: 631a, 1202a, 3399a, 4935a, 2700b, 5427b. Also notable is that all but one (1202a) of these exceptional verses contain a disyllabic ending such as -ono. It would accordingly seem far from gratuitous to generalise that operation of resolution on the second position of type E has to do with the existence elsewhere in the verse of a disyllable functioning as a morpholexical unit, be it a stem or an ending. We may draw 4

Psx#P: Beo 1584a, 463b, 623b, 783b, 2779b, 2792b; Psxx#P and Psx#xP: Beo 911a, 1187a, 1681a, 2650a, 2583b, 2661b, 2780b.

MOSP-03.qxd

5/11/04

2:41 PM

Page 203

3.1. Resolution

203

attention here to a limited but somewhat analogous situation in the treatment of the second lift of type C, whereby resolution on the second lift is found to be conditioned by implementation of resolution on the first lift (sections 3.1.1.2 and 3.1.1.6). By contrast, the observed dependence on the presence of a resolvable disyllable in another position seems to have no analogue in Beowulf: of the seven PsxxP verses noted above, only one (c. 14 per cent; Beo 911a fæderæþelum onfon) is affected by resolution elsewhere. 3.1.1.6. Implementation versus suspension of resolution: a synchronic and diachronic account As observed above from time to time, the same lexical items are used differently depending on metrical contexts. As far as type A1 is concerned, cuning ‘king’, for example, is resolved in the configuration Pxpxx (e.g., 62a gicoran te kuninge; section 3.1.1.1), but remains unresolved in the configuration PS#px (e.g., 703a slîðmôd cuning; section 3.1.1.1). In a similar vein, when used as the second element of a compound, the same lexical item -cuning- is resolved when associated with the second position of type E (e.g., 537a hebencuninges hugi), but remains unresolved for the second and third positions of type D (e.g., 2443b hebencuninges; section 3.1.1.3; 4703a hêr hebancuning; section 3.1.1.4). Thus, implementation of resolution is not a matter of lexical choice; rather it is a structurally regulated metrical practice. In more specific terms, implementation and suspension of resolution depends heavily on the metrical position concerned in its relation to the metrical type in which it appears, and, to a lesser extent, on the degree of stress associated with it, as shown in the foregoing examination. Of particular significance, implementation and suspension of resolution is largely complementary in their scope, as summarised in Table 3.1 below. There are three parameters involved that interact to determine susceptibility to suspension of resolution: (i) the prominence on the immediately preceding position; (ii) the location of the metrical position concerned in a verse; (iii) the prominence of the language expression associated with the position concerned. By the first parameter, the appearance of a stressed syllable as an occupant of the immediately preceding position makes possible suspension of resolution on the

Table 3.1. Implementation versus suspension of resolution metrical type

metrical position concerned

configuration affected by resolution

configuration suspended from resolution

A1 A1 C C D D E E

third position third position third position third position second position second position second position second position

PS#pxx (rare) Pxpxx x . . . P#pxx (rare) x . . . Psxx (rare) P#pxxx Psxxx (non-existent) P#pxxP PsxxP

PS#px Pxpx (non-existent) x . . . P#px x . . . Psx P#pxx (non-existent) Psxx P#pxP (non-existent) PsxP (rare)

MOSP-03.qxd

5/11/04

204

2:41 PM

Page 204

3. Resolution and alliteration

following position. As shown in the differential treatment of the two minimally distinct configurations PS#px and Pxpxx (section 3.1.1.1), the third position of type A1 receives resolution only when it is immediately preceded by the second position that is realised by an unstressed syllable (x); by contrast, when preceded by the second position that is associated with a stressed syllable (S), resolution is precluded from operating on the third position, thereby giving rise to subtype A1s. It should be noted that the difference between these two configurations PS#px and Pxpxx has to do with the prominence concretely realised on the second position, rather than with the categorical status of it in underlying metrical terms (as implied in Kauffmann 1887: 285 and Hofmann 1991: 61, for example), because both positions are a normal drop in the Heliand. Accordingly, the association of the immediately preceding position with a lexical-stressed syllable, regardless of its metrical status, makes the following position susceptible to suspension of resolution. Much the same explanation readily extends to the absence of the configuration Px . . . pxx (unresolved; subtype D*1) in contrast to the existence of Px . . . pxxx (resolved; subtype D*1): the presence of an additional drop (x . . .) between the lifts renders operation of resolution on the second lift obligatory. Parameter (i) thus delimits the context in which suspension of resolution may take place, and by virtue of this general framing function it ranks highest among all the three parameters. According to the second parameter, suspension of resolution affects the third position with greater preference than the second. This is testified by the contrast between the configurations x . . . P#px (type C; section 3.1.1.2) and P#pxxx (type D; section 3.1.1.3): despite the same metrical status of the position (lift) and the same immediate context involved (P_x), the third position is suspended from resolution (x . . . P#px) with overwhelming incidence, whereas the second counterpart undergoes resolution (P#pxxx). In other words, the occurrence in penultimate position constitutes a strong inducer for suspension of resolution (Suzuki 1996a: 188–9, 232), while the second position shows a relatively weak correlation with suspension of resolution. In similar fashion, the same parameter accounts for the prevalence of resolution over suspension of resolution for the second position of type E: as observed in section 3.1.1.5 above, the configuration PsxxP is by far the commoner variant than the unresolved counterpart PsxP. The third parameter has to do with the relative prominence of the language expression that realises the position involved. Specifically, resolution is more likely to be suspended from operating on the position that is attended by a secondary stress. Thus, the configurations P#pxxx and Psxx, which are identical except for the difference in prominence on the second position (the primary-stressed syllable versus the secondary-stressed syllable), are subject to divergent treatment (section 3.1.1.3): the configuration P#pxxx is resolved while the configuration Psxx is suspended from resolution. Being the second lift in both configurations, the positions in question are indistinguishable from each other in their metrical status. Therefore, we need to take into account the difference in prominence on their linguistic realisations. Similarly explainable would be the contrast between primary and secondary stresses on the second position of type E. While the unresolved configuration Psx#P is attested, the primary-stressed counterpart P#px#P is not, in contrast to the presence of P#px#xP, as pointed out in section 3.1.1.5 above. The difference in prominence

MOSP-03.qxd

5/11/04

2:41 PM

Page 205

3.1. Resolution

205

realised on the first drop may therefore be held responsible for the divergent treatment in question. A comparable distinction, however, is not shared by the pair of type C configurations x . . . P#px and x . . . Psx. As observed in section 3.1.1.2 above, and will be discussed more fully in section 3.1.3 below, these two variants seem hardly distinguishable as regards the consistency with which they are suspended from resolution: both are only exceptionally affected by resolution with hardly appreciable distinction in frequency. In this light, and on the evidence of the divergence observed in the above two pairs of configurations (P#pxxx versus Psxx; P#pxxP versus PsxP), we may accordingly generalise that the parameter of prominence, concerned as it is with the second position alone, is subordinate in scope to the second parameter of location in a verse: parameter (iii) applies exclusively to the cases that are determined as less amenable to suspension of resolution by parameter (ii). The three parameters thus identified are ranked according to their varying scope of application in the descending order of generality: parameter (i) 傻 parameter (ii) 傻 parameter (iii). The rule for suspension of resolution in the Heliand may accordingly be generalised as a complex of ranked subrules as follows: (19) Rule for Suspension of Resolution in the Heliand a. Suspension of resolution may occur on the strong metrical position (lift or heavy drop) that is immediately preceded by a foot with lexical stress. b. Suspension of resolution occurs more frequently on the third (penultimate) position than on the second. c. Suspension of resolution occurs more frequently on the secondary-stressed syllable. Ranking: subrule (a) 傻 subrule (b) 傻 subrule (c) Given the pairs of candidate configurations, one resolved, the other unresolved, these three ranked subrules brought together determine susceptibility to suspension of resolution as summarised in Table 3.2. The first subrule (19a) serves as a general qualifying condition on suspension of resolution: resolution may, but need not, be suspended when a candidate position is Table 3.2. Determining susceptibility to suspension of resolution by the three subrules type

candidate configurations unresolved resolved

subrule (19a) (19b) (19c)

configuration selected

A1 A1 C C D D E E

Pxpx PS#px x . . . P#px x . . . Psx P#pxx Psxx P#pxP PsxP

⫺ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹

Pxpxx PS#px x . . . P#px x . . . Psx P#pxxx Psxx P#pxxP PsxP/PsxxP

Pxpxx PS#pxx x . . . P#pxx x . . . Psxx P#pxxx Psxxx P#pxxP PsxxP

⫹ ⫹ ⫹ ⫺ ⫺ ⫺ ⫺

Key: ⫹ ⫽ counted as susceptible to suspension of resolution

⫺ ⫹ ⫺ ⫹

MOSP-03.qxd

5/11/04

206

2:41 PM

Page 206

3. Resolution and alliteration

immediately preceded by a lexical foot. Conversely, implementation of resolution is nearly obligatory on the position that is immediately preceded by material lacking lexical stress. As a consequence, the resolved configuration Pxpxx (type A1) is selected over the unresolved counterpart Pxpx. By contrast, the remaining configurations in Table 3.2, in which the target positions are all preceded by a foot with lexical stress, are counted as qualified for suspension of resolution. Subrule (19a), despite its overall binding force, seems to admit of some exception: there are a small number of verses that leave their first lift unresolved although the position in question is not preceded by a lexical foot. These exceptional verses fall into the following four configurations: (i) pxSx (type A1; also a questionable instance of px#Px); (ii) x . . . px (type A3); (iii) x . . . px#P (type B1); (iv) pXx#P (type E). As regards the configurations pxSx and px#Px, we find four examples as provided below (cf. Hofmann 1991: 150–1),5 disregarding thirteen apparently similar instances containing foreign names:6 (20) 1600a Fadar ûsa (emended; originally Fadar is ûsa (M) and Fader ist ûsa (C), a common type A1 verse) 5937a alauualdan 31b alomahtig 1143b hebenrîki If resolution were invoked, the above four verses would contain only three positions in violation of the fundamental principle of four positions per verse (section 1.2). Given the absence of the configuration P#Px, it may be reasonable to scan them as type A1 with the first lift unresolved. There is a single example of the configuration x . . . px, as follows: (21) 5605a mid mi samad Since there are only two weak syllables before samad, this verse cannot be scanned as type B3, as can verse 1533a that gi sô ni uurecan, for example (section 2.5). We may accordingly scan this verse as type A3 with the unresolved first lift. The configuration x . . . px#P is attested in the following five instances, apart from an array of verses that contain foreign names at the relevant positions:7 (22) 2001b 2014b 2780b 5730b 5738b 5

6

7

thar the heri dranc that for thene heri forð endi hêt thene godes man thar hie uuissa that godes barn Thar sia that godes barn (emended; originally Thar sia that barn godes)

Also noteworthy are verses 1603a Cuma thîn and 3455a godes êu, which would appear to have only two positions (cf. Hofmann 1991: 152). 3054a (Simon Petrus), 3196a, 3201a, 3210a, 3306a, 4508a, 4598a, 4629a, 5160a, 3093b, 3108b, 4866b, 5898b. Of our interest here are verses 326a that is Iesu Krist, 965a, 3557a, 3716a, 4478a, 5126a, 5176a, 5376a, 3257b, and 4720b, which will be treated in Appendix 1 when considering the prosodic structure of Iesu, Iohannes, Iudas, and Rumu.

MOSP-03.qxd

5/11/04

2:41 PM

Page 207

3.1. Resolution

207

Resolution on the first lift would have to be excluded, because the resulting verses would contain only three positions. Since there is no example of the sequence x . . . P#P, the above five verses should best be scanned as type B1 with the first lift occupied exceptionally by a short stressed syllable alone through suspension of resolution. Accordingly, these verses may be regarded as a violation of subrule (19a), rather than as too short verses consisting of three positions with the first lift resolved as Hofmann (1991: 152) and Russom (1998: 186) assume. Finally, we encounter the following two verses of the configuration pXx#P: (23) 2619b alungan tîr (aldarlangan tîr; C) 5544b derebia mann In parallel to the three groups of exceptional verses treated so far, the first lift of the above two verses may be interpreted as suspended from resolution exceptionally, and the whole verses may accordingly be identified as type E. All considered then, the above four groups of unusual verses should be analysed as containing the unresolved first lift. In other words, they are exceptional, not because they consist of only three positions with resolution implemented as Hofmann (1991, II: 213) scans them, but because their first lift is suspended from resolution despite the requirement to the contrary as prescribed by subrule (19a). Since the otherwise maximally prominent, first lift is suspended from resolution, we may assume that the verses under consideration are equipped with a correspondingly diminished level of prominence. This weakened prominence may in turn account for the notable fact that all these exceptional verses are attended by single alliteration, given the correlation between resolution and alliteration (see further below). Furthermore, the predominant occurrences of b-verses may be ascribed to the distribution patterns of the corresponding configurations with the first lift occupied by a foot (except for type A3): the configurations PxSx (type A1; Table 2.29), x . . . Px#P (type B1; Table 2.67), PXx#P (type E; Table 2.89) all exhibit demonstrable preference for the b-verse. Therefore, the unusual verses under discussion may be characterised as excessively reduced tokens of the otherwise minimal variants of types A1, B1, and E.8 Within the general contexts circumscribed by the first subrule, the other two subrules (19b) and (19c) come into play to determine specifically susceptibility to suspension of resolution in complementary ways. The second and third subrules accordingly apply vacuously to the sequence Pxpxx (type A1), because it is already disqualified from suspension of resolution by the first subrule, as indicated by shaded background in Table 3.2. Like the first subrule, subrules (19b) and (19c) are characterised as preference conditions, rather than absolute categorical rules: they favour suspension of resolution in their own different ways with less than mandatory force. Resolution may be 8

Since only lexical-stressed syllables are resolvable, and since there is no lexical word that ends in a short stressed syllable, that is, *-p#/-s#, it follows as a matter of course that suspension of resolution is categorically excluded from the first lift of types C and D: the sequences *p#P and *pS are ruled out. Precisely for the same reason, the second lift of types B1, B3, and E, as well as the final position of subtype D2b, are unqualified for suspension of resolution.

MOSP-03.qxd

5/11/04

208

2:41 PM

Page 208

3. Resolution and alliteration

suspended when one of these two subrules is in effect, but with varying degrees of incidence derived from the difference in their ranking. Thus, the unresolved configurations P#pxx (type D) and P#pxP (type E) are excluded in favour of the resolved counterparts P#pxxx and P#pxxP, respectively, because neither of the subrules is satisfied: the second p carries a primary (rather than secondary) stress and occupies the second (rather than the third) position. Even when one of the conditions is met, however, suspension of resolution is far from obligatory. By the second subrule, the sequence x . . . P#px (type C) is selected over x . . . P#pxx (type C) as normal and natural by virtue of subrule (19b), yet the sequence x . . . P#pxx may be realised if only marginally. In similar fashion, the second position of type E may be suspended from resolution when it is realised by a non-primary-stressed syllable (PsxP), yet resolution is also allowed to occur (PsxxP). Moreover, as a reflection of the lowest rank, subrule (19c) seems to have weaker force so that the resolved sequence PsxxP occurs more frequently than the unresolved counterpart PsxP. After having identified the set of structural constraints that conditions suspension of resolution as generalised in (19) above, we have to address the functional basis that motivates the setting of the parameters or subrules in question: why is resolution suspended under the conditions as specified by these subrules? Given a concatenation of two prominent positions, the second position, which counts as weaker on independent grounds on account of its proximity to the end of a verse (by virtue of the Linearity-based Metrical Strength Scale; section 1.2), may undergo a further lowering of prominence in perceptual terms by contrast with the preceding prominent position in the absence of a markedly lesser salience there, as observed by Rieger (1876: 49, 56) long ago. Such a syntagmatically motivated weakening effect may accordingly deter the second of the two consecutive prominent positions from realising its full potential that would be manifested in other contexts by association with a foot. The position in question may thus be required to be actualised in a less prominent shape. Accordingly, resolution, a means of maximising prominence, is found at variance with the position at issue on syntagmatic grounds, and therefore may be blocked from affecting it (Suzuki 1996a: 187). Suspension of resolution may therefore be characterised as a formal adjustment for decreasing prominence in correspondence with the syntagmatically motivated weakening that affects the second of two consecutive prominent positions. In more technical terms, suspension of resolution comes into effect through association of a strong metrical position (lift or heavy drop) with the level of the syllable rather than with the level of the foot as required by implementation of resolution. By being mapped onto the level lower down in the prosodic hierarchy, the position that is suspended from resolution receives a correspondingly lesser amount of prominence in accord with the syntagmatic requirement (Suzuki 1996a: 246). The foregoing consideration thus motivates the setting of parameter (i), subrule (19a), that distinguishes the concatenation of two prominent positions (PP, SP, PS) from the sequence of a minimally prominent position and a more prominent one (i.e., xP, xS). For the latter sequence, the second, more prominent position is capable of realising its full potential because of a lesser salience on the preceding position. Restated in more general terms, the second of the two consecutive prominent positions, which is susceptible to a relative weakening of metrical strength, is more likely to be realised in a correspondingly less salient form. Underlying subrule (19a)

MOSP-03.qxd

5/11/04

2:41 PM

Page 209

3.1. Resolution

209

is thus recognised the principle that suspension of resolution inheres in lesser prominence. The same principle provides a basis for both subrules (19b) and (19c). Subrule (19b) relates to a relative prominence that is determined by the location of a given position within a verse: the third position is weaker in metrical strength and accordingly less prominent than the second position as generalised in the Linearitybased Metrical Strength Scale (section 1.2). Subrule (19c), on the other hand, concerns an amount of prominence that is realised on the surface through association with a particular language material: the realisation as a non-primary-stressed syllable is characterised as less prominent than the one as a primary-stressed counterpart. At this point, we should recall the subregularity observed in the composition of x . . . P#pxx, a configuration which is rare in occurrence as noted earlier. As observed in section 3.1.1.2 above, the sequence x . . . P#pxx has a strong preference for resolution on the first lift. Put another way, implementation of resolution on the second lift of type C occurs with notably greater frequency when the first lift itself is resolved (x . . . px#pxx) than when it is occupied by a long stressed syllable (x . . . P#pxx), as observed in Sievers (1893: §114.4). How does such a tendency fit in with the above set of conditions for resolution? As discussed above, suspension of resolution is motivated by the clashing of prominent positions on the surface and the consequent weakening of prominence on the second one in perceptual terms. Subrule (19b) along with subrule (19c) is a preference condition: it does not exclude implementation of resolution on categorical grounds. When the second lift of type C is thus exceptionally allowed to be realised by a foot in the form of a resolved disyllable, however, a further strategy is taken to see that the resultant prominence contour deviates minimally from the underlying metrical configuration. If resolution were free to apply exclusively to the second of two consecutive prominent positions, the resulting realisation would strongly contradict the weakening that that position is otherwise expected to receive: the second position would be given its maximal prominence on the surface at the cost of the underlying metrical contour whereby the first should stand out as stronger. By giving maximal prominence to the first lift as well by means of resolution, the exceptionally sanctioned full realisation of the second lift as a foot rather than as a single stressed syllable is less likely to overshadow the preceding stronger position by its maximal prominence. Thus, the exceptional occurrence of the configuration x . . . px#pxx is made possible exactly through the same mechanism that applied to the traditional metre (cf. Suzuki 1996a: 200–1). The weakening of the second position due to the clashing of two prominent positions makes itself felt independently as the significantly high incidence of resolution on the first lift of type C when the second one is realised by a long syllable, as will be treated in depth in section 3.1.4 below. We may also recall a further manifestation of the syntagmatic conditioning for implementation of resolution, mentioned in section 3.1.1.5 above: the operation of resolution on the second position of type E is favoured by the presence of a disyllable in another position, the first or second lift for the most part. Thus, the string sx is limited to the configurations pxsxxP, Psxxpx, and PsxxxP. Noteworthy here, however, is that the resolvable disyllable in question does not have to be found in the immediately preceding position. With a view to gaining a deeper understanding of the rule for suspension of resolution in the Heliand as characterised on synchronic grounds, we shall put them in

MOSP-03.qxd

5/11/04

210

2:41 PM

Page 210

3. Resolution and alliteration

Table 3.3. Implementation versus suspension of resolution in Beowulf metrical type

metrical position concerned

configuration affected by resolution

configuration suspended from resolution

A1 A1 C C D D E E

third position third position third position third position second position second position second position second position

PS#pxx Pxpxx x . . . P#pxx (rare) x . . . Psxx (rare) P#pxxx Psxxx P#pxxP PsxxP

PS#px Pxpx (rare) x . . . P#px x . . . Psx P#pxx Psxx P#pxP (non-existent) PsxP

their diachronic and comparative perspective by making a comparison with the situation in Beowulf. Table 3.3 provides an overview of the distinction between operation and suspension of resolution in Beowulf in the metrical contexts corresponding to those covered in Table 3.1 above. With respect to the opposition between implementation and suspension of resolution, the Heliand differs from Beowulf primarily in their sharper contrast in distribution. This fundamental characteristic of the Heliand finds two major manifestations: (i) the third position of type A1, and (ii) the second position of types D and E. The first manifestation concerns the contrast between the configurations PS#pxx and PS#px. While Beowulf allows both configurations to occur frequently (see (3) above), the Heliand favours the unresolved configuration PS#px at the expense of the resolved one PS#pxx, which is all but non-existent. Thus, as pointed out in section 3.1.1.1 above, suspension of resolution is closest to being obligatory in this context in the Heliand. As for the second manifestation, while in Beowulf implementation and suspension of resolution occur without appreciable difference in incidence, they are subject to a fully or nearly complementary distribution in the Heliand: for type D, the resolved configuration P#pxxx and the unresolved one Psxx are attested to the exclusion of the unresolved and resolved counterparts P#pxx and Psxxx, respectively (section 3.1.1.3); and for type E, the resolved configuration PsxxP is highly preferred to the unresolved counterpart PsxP (section 3.1.1.5). As proposed in Suzuki (1996a: 231–2), suspension of resolution is governed by the following complex of rules in Beowulf: (24) Rule for Suspension of Resolution in Beowulf (Suzuki 1996a: 231–2) Resolution may be suspended if the following conditions are met: a. The metrical position with which a resolvable disyllable gets associated is of level 2. b. The position in question is immediately preceded by a stronger position. c. The disyllable involved is of non-minimal prominence. Resolution must be suspended if, in addition to the above conditions, the following extra condition is met: d. The metrical position with which a resolvable disyllable gets associated is penultimate.

MOSP-03.qxd

5/11/04

2:41 PM

Page 211

3.1. Resolution

211

Subrule (24a) is equal to saying that operation and suspension of resolution involves the lift and heavy drop, as opposed to the normal drop. It thus finds a reflection in the earlier part of subrule (19a) for the Heliand. Subrule (24d), corresponding to the second subrule (19b) postulated earlier for the Heliand, prescribes that suspension of resolution is (nearly) obligatory for the third metrical position. Consequently, the second lift of type C is normally subject to suspension of resolution in the Heliand as well as in Beowulf. These two subrules are thus essentially maintained in their traditional forms in the Heliand. Differentiating between Beowulf and the Heliand in regard to suspension of resolution are therefore subrules (24b) and (24c) for Beowulf on the one hand, and subrules (19a) and (19c) for the Heliand on the other. Subrule (24b) and its transform, subrule (19a), have to do with the opposition between the configurations PS#pxx and PS#px (type A1). Through the motivation and mechanism elucidated in sections 2.1.8 and 3.1.1.1 above, the configuration PS#px nearly ousted its structural opposite PS#pxx in the Heliand. This loss of opposition would then have disrupted the basis for composing the verse form Pxpx by analogy, and thereupon have brought about a reanalysis of the configuration PS#px, whereby lack of resolution on the second lift came to be treated as a fully integrated case of suspension of resolution. This integration consequently replaced the traditional stipulation, subrule (24b), with subrule (19a), which was generalised sufficiently to subsume the reanalysed configuration PS#px under suspension of resolution in the Heliand. Subrule (24c) draws on Kaluza’s Law, a regulation on the working of resolution in Beowulf. The following section then examines from a synchronic perspective whatever bearing this rule has on the Heliand metre, and then considers from a diachronic perspective the emergence of subrule (19c) as its replacement. 3.1.2. Loss of Kaluza’s Law and its consequences: a further dimension on resolution In the metre of Beowulf, operation of resolution is regulated by a generalisation referred to as Kaluza’s Law (Suzuki 1996a: 205–33). This law specifies the conditions for implementation of resolution in the contexts where a resolvable disyllable is immediately preceded by a syllable that is associated with a stronger position, as stipulated in (24c). Restated in light of the reorganisation that the metre underwent in the Heliand notably in regard to the configuration PS#px (sections 3.1.1.1 and 3.1.1.6), Kaluza’s Law, a constraint on operation and suspension of resolution, would come to effect when a resolvable disyllable immediately follows a foot. More specifically, should Kaluza’s Law apply in the Heliand, resolution would be suspended from operation if and only if the second syllable of the disyllable in question ends in a consonant or a vowel that is not a reflex of pre-OS *i or *u; conversely, resolution would have to be implemented if and only if the second syllable of the disyllable ends in a vowel that derives from pre-OS *-i or *-u. The function of Kaluza’s Law would thus be to exclude the minimally sonorous disyllable *-VCi/u from suspension of resolution and to guarantee its full involvement in resolution. The question to be addressed then is whether Kaluza’s Law actually applies to the Heliand. An examination of the contexts in which suspension of resolution is

MOSP-03.qxd

5/11/04

212

2:41 PM

Page 212

3. Resolution and alliteration

mandatory or nearly so shows that the short disyllables -VCi/u as well as the long counterparts are equally used: (25) Subtype A1s (PS#px; see (1) above), type C (x . . . P#px; see under Table 2.77, section 2.6 above), subtype D1 (Psxx; see (10) above) short disyllables: filo (u-stem, adv.: 5686b); garo (wa-stem, adv.: 825b); garu/o (wa-stem masc. nom. sg.: 5646b; fem. nom. sg.: 5940b; neut. acc. sg.: 273b, 929b, 2023b, 2324b, 2831b, 2998b, 3440b, 4608b, 5206b); hugi (i-stem nom. sg.: 1357b, 2253b, 2677b, 2916b, 2928b, 3178b, 3291b, 4491b, 4588b, 4671b, 4727b, 4771b, 5464b); sunu/o (u-stem nom. sg.: 1084b, 2192b, 3132b, 3214b, 3547b, 3906b, 4181b, 4203b, 4549b: acc. sg.: 4845b, 4946b, 5238b); long disyllables: -bora- (past participle: 222b, 464b); dedi (3sg. pret. subj.: 4883b); maco (3sg. pres. subj.: 1698a); uuero (a-stem gen. pl.: 4145a, 5071a, 5270b); and numerous examples of disyllables ending in a consonant, such as (-)cuning, godes (for further examples, see (1) and (10) above) Conversely, resolution affects long as well as short disyllables which stand immediately after a foot:9 (26) Subtype D1 (P#pxxx; see (11) above); type E (PsxxP; see (16) above) short disyllable: -skepi (i-stem dat. sg.: 3001a) long disyllables: fago- (526b, 3029b); -gumo- (619a); -sago/u- (631a, 3399a, 4935a); -sma- (2700b); -samo- (1202a); -scaðo- (5491a, 5427b); sora- (5789b); soro- (2244b); -togo- (2735a); uuaco- (384a); and many instances of disyllables ending in a consonant, such as -cuninges and uueder (for further examples, see (16) above) Thus, implementation and suspension of resolution in the Heliand does not rest on the traditional distinction between short and long disyllables, and therefore we must conclude that Kaluza’s Law is not in effect in the Heliand. The total dysfunctioning of Kaluza’s Law raises the question of why the law lost its force, to which we turn in the following. The posing of such a question of course presupposes that Kaluza’s Law constituted an integral part of the traditional metre, rather than an innovation that exclusively affected the metre of Beowulf. Before exploring the loss of the law from the metre of the Heliand, then, we have to justify that the law was part of the traditional versecraft and accordingly should not be dismissed out of hand simply as foreign to Old Saxon. Two points are worth considering in this connection. First, as remarked above, the phonological basis of the law, namely the distinction between short and long disyllables, derives from the special property of pre-OE *-i and *-u, that is, the mora-based susceptibility to High Vowel Deletion (Suzuki 1996a: 241–5). This phonological change and its synchronic consequences make themselves felt in Old Saxon as well, most notably in the distinction between short and long forms 9

The configuration Pxpxx (type A1) has to be left out of consideration here, because in traditional versecraft it falls outside the domain of Kaluza’s Law by virtue of the association of the preceding position with unfooted material x.

MOSP-03.qxd

5/11/04

2:41 PM

Page 213

3.1. Resolution

213

of i- and u-stem nouns: the short stems retain -i or -u, while the long counterparts lose these endings, as indicated below (cf. Holthausen 1921: §§287–306; Prokosch 1939: §§82–3; Krahe and Meid 1969b: §§12–18; Gallée 1993: §§313–29): (27) i-stem nom. sg. hugi ⬍ PGmc. *hug-iz ‘thought’ vs. gast ⬍ PGmc. *gast-iz ‘guest’ u-stem nom. sg. sunu ⬍ PGmc *sun-uz ‘son’ vs. hand ⬍ PGmc. *hand-uz ‘hand’ It is thus justified to postulate the unique feature – the moraic susceptibility to High Vowel Deletion – of unstressed short vowels *i and *u for pre-Old Saxon as well. Given the existence of the phonological foundation of Kaluza’s Law, it would seem natural to assume in the absence of evidence to the contrary that pre-Old Saxon, too, would have been compatible with the working of Kaluza’s Law. Second, even if Old Saxon poetry were denied of its own native tradition and consequently reduced to an import from Old English verse (see section 1.1 above), Kaluza’s Law as a traditional element of Old English metre would have an intrinsic bearing on Old Saxon metre as realised in the extant works. We could hardly avoid investigating the status of the traditional law on new soil, whether it fell into disuse there or it had long ceased to function in the old territory before introduction onto the continent. With the legitimacy established of posing the question that we did, we enter into a substantive consideration. The full functioning of Kaluza’s Law as a metrical rule was made possible by rendering the metre sensitive to the morphological consequences of High Vowel Deletion. As a result of the phonological change at issue, the short vocalic endings *-i and *-u came to be involved in the morphophonological alternation with zero, as exemplified in (27) above. The syllable quantity of the preceding stressed syllable chooses one of the overt and covert endings in such a manner that the overt variant appears only after a short stressed syllable. Such an allomorphic relation would have led to a reanalysis whereby the overt ending was identified as part of the short stem by analogy to the long counterpart that lacked an explicit ending through the mechanism discussed in Suzuki (1996a: 259–61): the short form sunu, for example, was accordingly reinterpreted as a stem pure and simple comparable to hand. Put another way, the original ending came to be incorporated as an integral part of the stem, and this morphological incorporation would have provided a motivation for treating the short disyllable as an inseparable unit that was found incompatible with being associated with separate metrical positions through suspension of resolution. One might be led to assume therefore that the Heliand metre would have shown no respect for morphological integrity, and hence the short disyllable was allowed to be associated with two distinct metrical positions, thereby rendering Kaluza’s Law invalid. Yet we find evidence to show that morphological integrity constituted a significant constraint in the metre of the Heliand. As discussed at length in section 2.7 above, the disyllabic endings (-xx#) are incapable of being split to occupy two different positions in verse-final position, so that we encounter no instance of the sequence *P#Pxx, while the sequence P#Pxx is normally attested.

MOSP-03.qxd

5/11/04

214

2:41 PM

Page 214

3. Resolution and alliteration

Despite the demonstrable metrical significance of morphological integrity in the Heliand, Kaluza’s Law failed to be at work. Obviously, the respect for morphological integrity cannot be identified as a sufficient condition for the sustained working of Kaluza’s Law, and accordingly it is not justified to attribute the dysfunction of the law in the Heliand metre to the absence of this morphological property. On an alternative conceptualisation, Kaluza’s Law was originally conditioned by the moraic value of the second syllable of a disyllable (Fulk 1992; Suzuki 1995b). In the contexts specified above, resolution was suspended from operation if and only if the second syllable of the disyllable was long, that is, bimoraic or more. Conversely, resolution had to apply in the presence of the short or monomoraic second syllable of the disyllable. Thus, on such a view the metre would have been sensitive to moraic distinction of word-final unstressed syllables if Kaluza’s Law had been applicable to it. Thus, moraic sensitivity would have served as an enabling condition for Kaluza’s Law. One may accordingly wish to attribute the inapplicability of Kaluza’s Law in the Heliand metre to its presumable insensitivity to moraic value of unstressed syllables. However, there is an indication that the Heliand was capable of making a moraic distinction for unstressed syllables. As treated in section 2.1.5 above, the unstressed short syllable (-x#; monomoraic) and the long counterpart (-X#; bimoraic) were largely distinguished: the configuration PX#Px exhibited a notably stronger preference for the a-verse with double alliteration than the sequence Px#Px; and this metrical distinction constituted one of the many innovations in the Heliand metre. It would appear surprising that Kaluza’s Law, sensitive as it originally was to a similar moraic distinction, did not find its way into the Heliand despite the availability of such a condition. Again, lack of sensitivity to moraic distinction cannot be held responsible for the failure of Kaluza’s Law to operate in the Heliand. Thus, neither indifference to morphological integrity nor to moraic distinction of unstressed syllables would have had subverted Kaluza’s Law in the Heliand metre, because it was actually sensitive to both. Of decisive significance in undermining the force of Kaluza’s Law seem rather the following two phenomena, one metrical, the other phonological in nature. A primary factor, I would claim, was constituted by the loss of type A2a as an independent metrical type in the Heliand (section 2.1.1). As shown in detail in Suzuki (1996a: 223), type A2a provided a principal context in which normally the short disyllable (-VCi/u) was exclusively qualified for resolution. With demise of this particularly favourable context, the privilege that the short disyllable enjoyed lost much of its structural foundation, and therewith the short disyllable was largely deprived of its basis for distinction from other disyllables (e.g., -VCVC). Upon loss of type A2a, there was no metrical reason why resolution had to be implemented on the disyllable in the configuration Psx#Px: suspension of resolution from this context would simply result in a sequence comparable to the one that became independently available in the Heliand, namely PSxPx (section 2.1.2). By contrast, in the virtual absence of the configuration PSxPx in Beowulf, the sequence Psx#Px had no other choice but to be scanned as resolved on the second position so as to be accepted as metrical. A second disturbing factor came from the phonology. Old Saxon is characterised by an abundance of svarabhakti vowels, those secondarily developed unstressed

MOSP-03.qxd

5/11/04

2:41 PM

Page 215

3.1. Resolution

215

vowels that came into being due to vowel epenthesis (see section 1.3.2 above). This process may be reproduced here for ease of reference: (28) ø —⬎ V / VR__C Key: R ⫽ /r/ or /l/ Epenthesised vowels are varied in quality, depending at least in part on the vowels of the immediately preceding syllables. This means that many instances of the newly developed resolvable sequence -VCV- are not identified as the privileged disyllable -VCi/u. As a result of vowel epenthesis, then, a whole variety of resolvable word forms found their way into the lexicon as amenable to resolution. Their original shape before operation of vowel epenthesis, however, was -VCC, a long syllable, and as such they had been outside the scope of resolution in traditional metrical practice. By virtue of their derived status, Kaluza’s Law as an archaic constraint on resolution would not have been capable of exercising any control over the new disyllables. Put another way, these novel word forms were free to be used without regard for the traditional opposition between implementation and suspension of resolution: these new long disyllables -VCV- (e.g., ferahtlîco, sorogodun; see (26) above) were accordingly allowed to be resolved alongside the short counterparts even in the remaining minor contexts that required use of the short disyllables for resolution, particularly the second position of type D. The purely phonologically motivated transformation of these monosyllabic stems into resolvable forms thus contributed further to disrupting Kaluza’s Law by obscuring the traditional distinguishability of the short disyllables through the introduction of these new long disyllables in the contexts originally prohibited from occurrence. These two factors thus crucially diminished the identifiability of the short disyllables by destroying the most privileged context of their use as well as by creating the long disyllables that were immune to the traditional distinction encoded in Kaluza’s Law. Things went further, however, and finally put an end to the working of Kaluza’s Law. The weakened identity of the short disyllables led in turn to endanger the identity of the long disyllables that depended on the categorical opposition to the short counterparts. It should be pointed out here that the long disyllables constituted by far the majority and therefore they counted as unmarked. It is the short disyllables that required positive evidence for identification and consequent differentiation. In the absence of positive grounds for identifying the short disyllables, then, the long counterparts were no longer recognisable as such. The consequences of the lost opposition subsequently made itself felt in the traditional contexts in which suspension of resolution was obligatory and hence the short disyllables were disallowed: the original short disyllables were now allowed to occur in the contexts they had been excluded from by traditional convention. In this connection, we may be reminded that unless directly preceded by a prominent position (the one filled by a foot) resolution came into play irrespective of the phonological properties of resolvable disyllables even in the traditional metre. Such a limited-scale opposition would certainly have facilitated the extended use of the short disyllables as well. In conclusion, the dysfunctioning of Kaluza’s Law in the Heliand was primarily a metrical event, largely attributed to the loss of type A2a as an independent metrical type. Added to this prime factor was vowel epenthesis, a phonological change

MOSP-03.qxd

5/11/04

216

2:41 PM

Page 216

3. Resolution and alliteration

characteristic of Old Saxon. These two phenomena in conjunction deprived the short disyllables of the basis for their identification, and with it the identity of the long disyllables as well. Considering further that the loss of type A2a was phonologically motivated (section 2.1.7), however, we should say that the disruption of Kaluza’s Law was phonologically conditioned in the final analysis. The unavailability of Kaluza’s Law in the Heliand would in turn have motivated the innovation of reconfiguring the opposition between operation and suspension of resolution for the second position by introducing subrule (19c) as its alternative. As may be recalled, operation and suspension of resolution were highly competitive in Beowulf as far as the second position was concerned (Table 3.3): both processes took place with more or less the same incidence. And Kaluza’s Law as specified in subrule (24c) fully regulated which of the alternatives occurred on particular instances according to their phonological structure. By contrast, for the penultimate position (the second lift of type C), the vast majority were subject to suspension of resolution by virtue of subrule (24d). Since this subrule was inherited as subrule (19b) in the Heliand, type C continued to be suspended from resolution in the Heliand with rare exceptions. It was accordingly the second position that was to be affected primarily by the inaccessibility to Kaluza’s Law in the Heliand. The disruption of the traditional generalisation in Old Saxon, as discussed in the foregoing paragraphs, would have obscured the originally regular distribution of resolved and unresolved configurations, which would have struck the poet as nearly out of control: the opposition between short and long disyllables was no longer available as a basis of generalisation. Such a lack of regularity would have invited the Heliand poet, an innovator with a keen sense of balance and harmony, to reorganise the metre for improvement. In reintegrating the distribution of the resolved and unresolved second position, the poet would naturally have resorted to the part of generalisation that he inherited from the traditional metre, namely subrules (19a) and (19b). As discussed in section 3.1.1.6 above, both were founded on the same principle: suspension of resolution concerned entities of lesser prominence. Drawing on this same principle, the poet then explored a viable solution to bring a new order to the disintegrated distribution that he had to work with. Since the variation between the primary and secondarystressed syllables as occupants of the second position in question would have appeared most visible among other variations, the poet presumably seized on this distinction and incorporated it as an integral part of the rule for suspension of resolution. Thus, subrule (19c) came to figure, alongside the inherited two others, subrules (19a) and (19b), in the reconstituted rule for suspension of resolution in the Heliand. In a less conspicuous way, the absence of Kaluza’s Law would have also borne on the implementation of resolution on the second lift of type C. By virtue of the continued function of rule (19b), this position seemed to be maintained as most amenable to suspension of resolution in the Heliand metre as in the Beowulf counterpart, as remarked above. On closer inspection, however, the Heliand metre proves to be more demanding in this respect. Specifically, the Heliand metre exceptionally allows resolution to occur on the second lift of type C with relatively lesser frequency than Beowulf: we find twenty-five such occurrences in the Heliand (as listed in section 3.1.1.2 above) and sixteen in Beowulf (Beo 164a, 1260a, 1703a, 1946a, 28b, 73b, 350b, 1154b, 1158b, 1603b, 1833b, 2096b, 2309b, 2796b, 2917b, 3128b).

MOSP-03.qxd

5/11/04

2:41 PM

Page 217

3.1. Resolution

217

Since the Heliand poet found himself incapable of distinguishing the two disyllabic variants in a principled way such as was done by recourse to Kaluza’s Law, he would have followed the relevant rule (19b) with greater stringency so as to minimise what stroke him as an unordered variation. 3.1.3. Apparent conflict between implementation and suspension of resolution: on the scansion of the configurations Px . . . Psx and Px . . . P#px The configurations Px . . . Psx and Px . . . P#px are apparently ambiguous as they seem open to two competing scansions, as type A2 on the one hand and type D* (subtype D*2a) on the other. The ambiguity resides in the divergent ways the penultimate short stressed syllable (s or p) may be treated in regard to resolution. If resolution is invoked, type A2 will result; if resolution is suspended, type D* will emerge. We find numerous such verses in the corpus, accounting for 225 in number, as exemplified below: (29) 473a hêlagna hebancuning 1180a sâlig barn godes 3895b hêlag barn godes The distribution of the configurations Px . . . Psx and Px . . . P#px, and the structurally related ones containing a long syllable S or P instead of s, p, sx, or px, is shown in Table 3.4. While the configuration Px . . . Psx is virtually limited to the a-verse with double alliteration, the primary-stressed counterpart Px . . . P#px is attested in the a-verse with single alliteration and the b-verse as well. In fact, with a few exceptions (3639a flêsk endi lîchamon; 3270b endi luggi geuuitskepi; 4753b the gêst endi the lîchamo), all a-verses with single alliteration and all b-verses, both with a short stressed syllable in penultimate position, are of the configuration Px . . . P#px, rather than Px . . . Psx. Disregarding for the moment this distributional property that is unique to the configuration Px . . . P#px and unknown to type A2 (Px . . . PS; Px . . . P#P) and subtype D*2a (Px . . . PSx; Px . . . P#Px) alike, we may address what seems to be the most urgent question, namely, scansion of the configurations Px . . . Psx and Px . . . P#px. As observed above, these configurations are subject to scansion as type A2 and subtype D*2a with apparently equal plausibility. As shown in Table 3.4 below, both type A2 and subtype D*2a, as instantiated by the configurations Px . . . PS and Px . . . PSx, respectively, are all but limited to the a-verse with double alliteration, precisely as is the ambiguous configuration Px . . . Psx. This distributional property is accordingly incapable of determining a proper scansion. No more distinguishing should be syntactic constituency: in both metrical types, the major verse-internal break usually comes between the first drop and the second lift, and an appeal to syntactic constituent structure is less likely to determine categorisation than is the case with the ambiguity between subtype D2b and type E (see section 3.2.2 below). We may then turn to the composition of the first drop for disambiguation. Table 3.5 shows the distribution of the varying number of syllables that occupy the first drop of the configurations Px . . . Psx (ambiguous), Px . . . PS (type A2), Px . . . PSx

MOSP-03.qxd

5/11/04

2:41 PM

218

Page 218

3. Resolution and alliteration

Table 3.4. Distribution of Px . . . Psx, Px . . . PS, Px . . . PSx, Px . . . P#px, Px . . . P#P, and Px . . . P#Px a-verse with double alliteration Px . . . Psx Px . . . PS Px . . . PSx Px . . . P#px Px . . . P#P Px . . . P#Px

139 65 146

a-verse with single alliteration

b-verse

98% 98% 97%

1 1 1

1% 2% 1%

2 0 4

1% 0% 3%

54 64% 11 100% 19 100%

12 0 0

14% 0% 0%

19 0 0

22% 0% 0%

Px . . . Psx a-verse with double alliteration: 229a, 301a, 341a, 346a, 360a, 378a, 386a, 411a, 424a, 454a, 473a, 480a, 535a, 627a, 651a, 668a, 672a, 679a, 691a, 727a, 742a, 765a, 980a, 1062a, 1115a, 1129a, 1176a, 1191a, 1228a, 1274a, 1322a, 1349a, 1359a, 1441a, 1530a, 1544a, 1646a, 1652a, 1661a, 1669a, 1722a, 1751a, 1783a, 1859a, 1860a, 1886a, 1896a, 1961a, 1981a, 2004a, 2009a, 2063a, 2174a, 2181a, 2229a, 2300a, 2317a, 2400a, 2515a, 2653a, 2730a, 2741a, 2766a, 2776a, 2780a, 2796a, 2855a, 2876a, 2886a, 2888a, 2889a, 2896a, 2899a, 2924a, 2926a, 2938a, 3053a, 3116a, 3126a, 3335a, 3476a, 3536a, 3686a, 3698a, 3800a, 3856a, 3943a, 3948a, 4033a, 4127a, 4190a, 4215a, 4242a, 4347a, 4443a, 4553a, 4634a, 4642a, 4662a, 4665a, 4743a, 4745a, 4759a, 4762a, 4766a, 4783a, 4812a, 4813a, 4897a, 4913a, 4944a, 4971a, 4981a, 5068a, 5125a, 5190a, 5195a, 5213a, 5242a, 5276a, 5303a, 5325a, 5338a, 5339a, 5367a, 5368a, 5497a, 5517a, 5550a, 5637a, 5643a, 5646a, 5658a, 5708a, 5734a, 5747a, 5774a, 5798a, 5803a; a-verse with single alliteration: 3639a; b-verse: 3270b, 4753b Px . . . PS (see under Table 2.38, section 2.2.1 above) Px . . . PSx (see under Table 2.82, section 2.8 above) Px . . . P#px a-verse with double alliteration: 49a, 193a, 276a, 399a, 598a, 648a, 895a, 1084a, 1282a, 1384a, 1400a, 1412a, 1485a, 1581a, 1642a, 1996a, 2122a, 2176a, 2251a, 2298a, 2355a, 2371a, 2403a, 2666a, 2847a, 2908a, 2937a, 2942a, 2943a, 3024a, 3070a, 3138a, 3248a, 3911a, 4011a, 4089a, 4115a, 4259a, 4270a, 4374a, 4722a, 4914a, 5084a, 5089a, 5133a, 5171a, 5332a, 5341a, 5395a, 5869a, 5882a, 5894a, 5946a, 5962a; a-verse with single alliteration: 7a, 518a, 798a, 1180a, 1440a, 2024a, 2038a, 2133a, 2325a, 2520a, 2975a, 5509a; b-verse: 400b, 812b, 832b, 847b, 1121b, 1165b, 1916b, 2172b, 2533b, 2975b, 3111b, 3293b, 3895b, 4796b, 4798b, 5093b, 5159b, 5974b, 5976b Px . . . P#P (see under Table 2.38, section 2.2.1 above) Px . . . P#Px (see under Table 2.82, section 2.8 above)

Table 3.5. Distribution of Px . . . Psx, Px . . . PS (type A2), Px . . . PSx (subtype D*2a), Px . . . P#px, Px . . . P#P (type A2), and Px . . . P#Px (subtype D*2a) according to the number of syllables in the first drop number of syllables Px . . . Psx 1 2 3 4 5 6

number of verses 69 44 14 10 2 1

percentage 49% 31% 10% 7% 1% 1%

MOSP-03.qxd

5/11/04

2:41 PM

Page 219

3.1. Resolution number of syllables 7 8 9

number of verses

219 percentage

1 0 1

1% 0% 1%

142

100%

Px . . . PS (type A2) 1 2 3 4 5

18 30 11 6 1

27% 45% 17% 9% 2%

total

66

100%

Px . . . PSx (subtype D*2a) 1 2 3 4 5 6

87 30 20 9 4 1

58% 20% 13% 6% 3% 1%

151

100%

Px . . . P#px 1 2 3 4 5

46 20 10 8 1

54% 24% 12% 9% 1%

total

85

100%

4 4 1 2

36% 36% 9% 18%

total

11

100%

Px . . . P#Px (subtype D*2a) 1 2 3 4

11 4 3 1

58% 21% 16% 5%

total

19

100%

total

total

Px . . . P#P (type A2) 1 2 3 4

MOSP-03.qxd

5/11/04

220

2:41 PM

Page 220

3. Resolution and alliteration

(subtype D*2a), Px . . . P#px (ambiguous), Px . . . P#P (type A2), and Px . . . P#Px (subtype D*2a). The configurations Px . . . Psx and Px . . . P#px prove to be closer to Px . . . PSx and Px . . . P#Px than to Px . . . PS and Px . . . P#P, respectively, in regard to the composition of the first drop. Specifically, the configuration Px . . . PS differs from Px . . . Psx and Px . . . PSx in terms of the commonest number of syllables involved (two as against one), and the relatively narrow range of the attested numbers of syllables (five as against eight and six). A similar grouping obtains for the set of the three configurations with a primary-stressed syllable: while the configurations Px . . . P#px and Px . . . P#Px have only a single syllable in the first drop in more than a half of all relevant instances, the corresponding figure for the configuration Px . . . P#P drops to c. 36 per cent, a proportion equal to that for the examples with two syllables in the same position. What is at stake here may indeed be a matter of degree rather than of sharp categorisation, yet we may reasonably suspect that the configurations Px . . . Psx and Px . . . P#px, more akin to the configurations Px . . . PSx and Px . . . P#Px than to Px . . . PS and Px . . . P#P, may plausibly be scanned as type D*, subtype D*2a. This supposition has its credibility strengthened when we take into account the differing proportion of the configuration Px . . . Psx to the configuration Px . . . PS on the one hand, and to the configuration Px . . . PSx on the other. As detectable from Table 3.4 above, the configuration Px . . . Psx outnumbers Px . . . PS more than twice in occurrence, whereas it is nearly as frequent as the configuration Px . . . PSx. On the interpretation that the sequence Px . . . Psx constitutes type A2 with resolution on the final position, we would be confronted with such an extremely uneven distribution in favour of the resolved variant. The abundance of resolution will become even more outstanding, when we bring into the picture the related configurations with a primary-stressed syllable, Px . . . P#px and Px . . . P#P. There seem no particular reasons, however, for favouring resolution in the context at issue; we would rather expect the preceding second lift to be resolved, particularly when the penultimate syllable is a primary-stressed one (p), given the generalisation that will be adduced to account for the prevalence of x . . . px#Px over x . . . P#Px (section 3.1.4), for example. In fact, we have eleven examples of Px . . . px#P (100 per cent) but none of Px . . . P#P (type A2); and eighteen examples of Px . . . px#Px (c. 95 per cent) but only one of Px . . . P#Px (subtype D*2a). On the view that the configurations Px . . . Psx and Px . . . P#px scan as type D*, subtype D*2a, however, the predominance of Px . . . P#px over Px . . . P#Px in frequency, and also the lack of such a difference in distribution for the pair of Px . . . Psx and Px . . . PSx, may yield to a principled explanation: since on this view the penultimate position constitutes a normal drop, the greater frequency with which the primary-stressed short syllable occurs than the long counterpart is seen as a natural consequence of the fact that the latter is counted as maximally marked, being farthest removed from the short unstressed syllable, the prototypical realisation of the normal drop (for a similar discussion in regard to the corresponding position of type D, see section 2.7 above). On the other hand, since the secondary stressed syllable is closer to the unstressed counterpart and characterised as correspondingly less marked, it is understandable that it is not affected by a comparable diversification based on syllable length.

MOSP-03.qxd

5/11/04

2:41 PM

Page 221

3.1. Resolution

221

The significant increase of type D* (section 2.8) in the Heliand including one of its major variants, subtype D*2a, may add further plausibility to the proposed scansion. Type A2, by contrast, does not seem firmly placed in the system, given its relatively low frequency in occurrence (section 2.2.1) as well as the generally marginalised status of the heavy drop in the metrical organisation. Taken altogether, then, scanning the configurations Px . . . Psx and Px . . . P#px as subtype D*2a recommends itself as a more plausible interpretation. We are now concerned with the special distributional property of the variant Px . . . P#px, observed above: this variant does not share with the other variant Px . . . Psx the near-exclusive occurrence of the a-verse with double alliteration, as it is also attested in no small numbers in the a-verse with single alliteration and the b-verse (Table 3.4). Moreover, the corresponding configurations with a long stressed syllable P instead of p in penultimate or final position, that is, Px . . . P#P (type A2; all being Px . . . px#P, similar to x . . . px#Px, type C, as noted above) and Px . . . P#Px (subtype D*2a; all being Px . . . px#Px, except for 2644a gildid god selbo, again similar to x . . . px#Px, type C, as noted above), are both confined to the a-verse with double alliteration. The questions arising then are thus twofold. First, why is the variant with a primary-stressed syllable (Px . . . P#px) rather than a secondarystressed one (Px . . . Psx) attended with single alliteration with such high incidence? Second, why is the configuration with a short primary-stressed syllable (Px . . . P#px) rather than a long one (Px . . . P#Px) accompanied with single alliteration so frequently? The first question has to do with alliterative patterning of heavy verses in general, the category of verses in which a drop is associated with a primary-stressed syllable. As will be discussed fully in section 3.2.2 below, the Heliand has at its disposal an enriched range of alliterative patterns for such heavy verses: it has access to the new patterns axa and axx, in addition to the pattern aax, which constituted nearly the only pattern available in the traditional metre. And one of these new patterns, axx, is precisely the one that is realised in the configuration Px . . . P#px when it is attended with single alliteration. Since this novel alliterative pattern is an exclusive property of heavy verses, it falls out as a matter of course that the variant Px . . . Psx, with the third lexical item (-sx) subordinated to the preceding one as the second element of a compound, remains unqualified for realising the pattern axx; it has to manifest double alliteration as otherwise required of the expanded metrical types. As will be shown in section 3.2.2 below, the emergence of the new alliterative patterns is made possible by the weakening of the originally canonical status of the pattern aax. And the selection of the particular pattern axx is conditioned by syntactic constituency: primarily responsible here is the absence of definite syntactic prominence on the second lexical word as against the third, the prominence that is coextensive with the syntactic organisation [A[BC]] (A, B, C ⫽ three independent words constitutive of a heavy verse; for details, see section 3.2.2 below). The majority of the heavy verses with the alliterative pattern axx, however, is characterised by the ambivalent structure [A[B]C]. The lack of syntactic prominence on the second word that serves as a necessary condition for the pattern axx has a significant bearing on the metrical dimension as well. The primary-stressed syllable of the second word that is exempted from alliteration has to be metrically in harmony with such a process of suppressing

MOSP-03.qxd

5/11/04

222

2:41 PM

Page 222

3. Resolution and alliteration

alliteration: in corresponding fashion it must be less than fully prominent. Crucially involved here are syntagmatically based considerations on relative prominence. When the third primary-stressed syllable P is long, the immediately preceding second one P has to be supplied with additional prominence so as to assert itself readily perceivable as a lift (the second lift) in differentiation from the following P, which constitutes a drop. This contextual adjustment normally calls for resolution on the second P as a prominence reinforcement: as pointed out above, out of a total of nineteen instances of the configuration Px . . . P#Px, and of a total of eleven occurrences of the configuration Px . . . P#P (Table 3.4), all but one example have resolution operated on the second P. Such added prominence, and the underlying motivation for it, may in turn be found largely at variance with suppressing alliteration on it. Accordingly, the configurations Px . . . px#P and Px . . . px#Px are hardly compatible with the alliterative pattern axx. By contrast, when the verse-final independent word is of the form px with a short stressed syllable, the preceding second primary-stressed syllable P need not be reinforced with extra prominence, because the following stressed syllable carries a weaker degree of prominence by virtue of its inherently lesser quantity. This divergence between long and short stressed syllables has its closest parallel in the sharp differentiation of the sequences x . . . P#Px and x . . . P#px in regard to implementation of resolution on the first lift of type C, discussed at length in section 3.1.1.6 above. The absence of additional prominence that owes to the following short stressed syllable p then allows the medial primary-stressed syllable P to do without alliteration, thereby giving rise to the alliterative pattern axx. In Beowulf, too, the configurations Px . . . Psx and Px . . . P#px seem to suffer comparable structural ambiguity in overall configurational terms, with the consequence of two alternative scansions, types A2b and D*2b, competing with apparently equal force (Suzuki 1996a: 80–1, 233–5). On an individual basis, however, these configurations could be largely disambiguated by resorting to Kaluza’s Law. As formulated in (24), section 3.1.1.6 above, the Rule for Suspension of Resolution in Beowulf is sensitive to the distinction between short and long disyllables that is encoded in Kaluza’s Law. Suspension of resolution may accordingly take place only when the resolvable disyllable is not short, that is, when the disyllable in question does not end in a reflex of pre-OE *i or *u. In this light, the seemingly ambiguous configurations may be scanned properly as type D*2a whenever the verse-final syllable is not a reflex of pre-OE *i or *u; otherwise we would obtain type A2b with the resolved second drop. Thus, the ambiguity at issue may be viewed as more apparent than real in Beowulf. Yet we should bear in mind that the ambiguity persists as an inherent structural feature of types A2b and D*2a as far as the metrical configurations are concerned: both implementation and suspension of resolution constitute viable metrical processes in verse-final position after the sequence Px . . . P. Disambiguation comes to the fore only at the level of individual verse instances with decisive reference to the segmental properties of verse-final syllables. Loss of Kaluza’s Law and the consequent remaking of the Rule for Suspension of Resolution in the Heliand (section 3.1.2) subsequently led to a wholesale recasting of the traditional ambiguity between types A2b and D*2a: the ambiguity in question, one of the structural loosenesses in the traditional metre, was largely removed at the level of metrical types by choosing type D*, subtype D*2a, as a more viable scansion than the alternative as type A2b in the innovated metre of the Heliand.

MOSP-03.qxd

5/11/04

2:41 PM

Page 223

3.1. Resolution

223

We may naturally wonder at this point why subtype D*2a was chosen over type A2b as a more plausible scansion. This is probably because suspension of resolution constituted a marked phenomenon as against implementation of resolution, which applied by default. Since the configurations at issue satisfied the rule for the specific case (suspension of resolution), regardless of whether they also happened to fall under the general case (implementation of resolution), they would have been subsumed primarily under the marked case by virtue of the specificity involved. 3.1.4. Resolution and variants of metrical types There are two significant dimensions along which operation of resolution varies in frequency: (i) metrical configurations, that is, metrical types and their realisation variants; and (ii) distinction between the first and second lifts. Table 3.6 provides varying incidence of resolution according to metrical types and their variants, as well as the status of the lifts involved in resolution. The variation on resolution in the Heliand relative to metrical configurations on the one hand and to the distinction between the first and second lifts on the other may be described in general terms as follows. As far as metrical configurations are concerned, the lift that is followed by a sequence of two positions with decreasing prominence (i.e., lift ⫹ drop) shows a strong tendency to be resolved, as in type C (x . . . P#Px), type D (P#Pxx; P#PS/sx; P#Px . . . S). With respect to the distinction between the first and second lifts, the former is resolved more frequently than the latter (Kauffmann 1887: 299 n. 2; Sievers 1893: §111.1), unless one of the following two conditions is met. First, the second lift located in verse-final position receives more extensive operation of resolution than the first lift in the same verse, as in types B1 (x . . . Px#P; x . . . PS#P) and E (PXx(x)#(x)P; PSx(x)#(x . . .)P). Second, the second lift that is immediately preceded or followed by a stressed syllable associated with a drop is resolved more extensively than the first lift, as in type A2 (Px . . . PS) and heavy variants of types A1 (PS#Px), B1 (x . . . PS#P), D (subtype D2a; P#PS/sx), and D* (subtype D*2a; Px . . . PS/sx). The second condition, however, may be regarded as part of a generalisation of wider scope, since it seems applicable to the first lift to a certain extent, as in variants of type A1 (Psx#Px; PS#Px; PSx#Px) and of type E (PSx(x)#(x)P).10 Accordingly, the occurrence in verse-final position seems the only condition that uniquely increases the frequency of resolution on the second lift. A comparison with Beowulf (Table 3.7; cf. Suzuki 1996a: 196–200) may reveal that the variation on resolution in the Heliand is partly inherited from tradition and partly innovative unknown to tradition. In Beowulf, too, the two positions of decreasing metrical strength (i.e., lift ⫹ drop; heavy drop ⫹ normal drop) seem to be responsible for high incidence of resolution on the preceding lift (Suzuki 1996a: 197–9), as, on the one hand, on the first lift of types D1 (P#Pxx), D2a (P#PSx), D2b (P#Px . . . S), and to a lesser extent types C (x . . . P#P/px) and E (PS/XxP) as well, and, on the other hand, on the second lift of types D*1(Px . . . Pxx) and D*2a (Px . . . PSx). And as in the Heliand, the first lift is generally resolved more frequently than the second one, 10

Since the first lift constitutes the first occurrence of a stressed syllable in a verse, the clause ‘preceded by a stressed syllable’ is immaterial for this position.

MOSP-03.qxd

5/11/04

2:41 PM

224

Page 224

3. Resolution and alliteration

Table 3.6. Varying frequency of resolution on the first and second lifts relative to metrical types and their major realisation variants

metrical type and realisation variant

total number of verses

number and percentage of resolution on first lift

number and percentage of resolution on second lift

A1 (/ ⫻ / ⫻) Px#Px Psx#Px Pxx#Px P#xPx PX#Px PS#Px PXx#Px Px#xPx P#xxPx P#xxxPx PSx#Px PS#xPx

4107 1212 11 113 283 25 16 24 841 253 147 19 16

722 (18%) 155 (13%) 3 (27%) 1 (1%) 109 (39%) 3 (12%) 4 (25%) 3 (13%) 114 (14%) 105 (42%) 68 (46%) 5 (26%) 3 (19%)

465 (11%) 75 (6%) 0 (0%) 8 (7%) 38 (13%) 2 (8%) 4 (25%) 1 (4%) 100 (12%) 20 (8%) 21 (14%) 1 (5%) 0 (0%)

A2 (/ ⫻ / \) Px . . . PS

77 66

14 (18%) 12 (18%)

36 (47%) 25 (38%)

B1 (⫻ / ⫻ /) x . . . Px#P x . . . PS#P

3142 1841 71

385 (12%) 189 (10%) 26 (37%)

877 (28%) 475 (26%) 35 (49%)

C (⫻ / / ⫻) x . . . P#P/px x . . . PS/sx x . . . PX/xx

1788 1187 437 166

706 (39%) 541 (46%) 160 (37%) 7 (4%)

25 (1%) 11 (1%) 10 (2%) 0 (0%)

D (/ / ⫻ ⫻) P#Pxx P#PS/sx P#Px . . . S

344 116 86 14

211 (61%) 89 (77%) 38 (44%) 6 (43%)

59 (17%) 5 (4%) 41 (48%) 0 (0%)

D* (/ ⫻ / ⫻ ⫻) Px . . . Pxx Px . . . PS/sx Px . . . Px . . . S

677 106 301 31

110 (16%) 28 (26%) 47 (16%) 1 (3%)

157 (23%) 4 (4%) 94 (31%) 2 (3%)

E (/ \ ⫻ /) PXx(x)#(x)P PSx(x)#(x . . .)P

428 123 231

121 (28%) 5 (4%) 89 (39%)

173 (40%) 29 (24%) 100 (43%)

MOSP-03.qxd

5/11/04

2:41 PM

Page 225

3.1. Resolution

225

Table 3.7. Varying frequency of resolution on the first and second lifts relative to metrical types and their major realisation variants in Beowulf

metrical type and realisation variant

total number of verses involved

number and percentage of resolution on first lift

A1 (/ ⫻ / ⫻) Px#Px Pxx#Px P#xPx Px#xPx P#xxPx P#xxxPx

2198 871 31 320 535 120 9

297 (14%) 130 (15%) 0 (0%) 66 (21%) 57 (11%) 29 (24%) 1 (11%)

162 (7%) 58 (7%) 0 (0%) 33 (10%) 51 (10%) 7 (6%) 2 (22%)

A2a (/ \ / ⫻) PS#Px

147 144

25 (17%) 24 (17%)

28 (19%) 28 (19%)

A2b (/ ⫻ / \) Px . . . PS

112 108

17 (15%) 17 (16%)

24 (21%) 24 (22%)

B1 (⫻ / ⫻ /) x . . . Px#P

1047 556

91 (9%) 52 (9%)

78 (7%) 48 (9%)

C (⫻ / / ⫻) x . . . P#P/px x . . . PS/sx x . . . PX/xx

1121 529 422 170

241 (21%) 148 (28%) 70 (17%) 23 (14%)

16 (1%) 13 (2%) 3 (1%) 0 (0%)

D1 (/ / ⫻ ⫻) P#Pxx

342 244

145 (42%) 116 (48%)

17 (5%) 12 (5%)

D2a (/ / \ ⫻) P#PS/sx

90 87

41 (46%) 40 (46%)

11 (12%) 10 (11%)

D2b (/ / ⫻ \) P#Px . . . S

75 42

34 (45%) 17 (40%)

1 (1%) 0 (0%)

subtype D*1 Px . . . Pxx

52 50

2 (4%) 2 (4%)

27 (52%) 27 (54%)

subtype D*2a Px . . . PSx

70 68

6 (9%) 5 (7%)

34 (49%) 33 (49%)

subtype D*2b Px . . . PxS

39 21

3 (8%) 2 (10%)

2 (5%) 0 (0%)

405 74 273

106 (26%) 19 (26%) 77 (28%)

33 (8%) 5 (7%) 23 (8%)

E (/ \ ⫻ /) PXx#(x)P PSx#(x)P

number and percentage of resolution on second lift

MOSP-03.qxd

5/11/04

226

2:41 PM

Page 226

3. Resolution and alliteration

unless the second lift on its own meets the condition just mentioned for increased implementation of resolution. Furthermore, as shown in types A2a and A2b, the second lift that is immediately preceded or followed by a stressed syllable associated with a drop (heavy drop) is slightly more susceptible to resolution than the first lift unless the latter is involved in the contexts particularly favourable to resolution. The immediate adjacency of a heavy drop thus induces a moderate increase in implementation of resolution on the second lift in Beowulf (Suzuki 1996a: 199).11 Since only types A2a and A2b are involved and to a limited extent, this generalisation has to be qualified as a minor one. In certain respects, however, the two works diverge in the variation pattern of resolution. Of paramount importance are the following two contexts: (i) the first lift of type C; (ii) the second lift of subtype D2a (P#PS/sx) in the Heliand (⫽ type D2a in Beowulf) and subtype D*1 of type D* (Px . . . Pxx) in the Heliand (⫽ subtype D*1 of type D1 in Beowulf). Starting with the seemingly less outstanding case: while high incidence of resolution on the first lift of type C may be generally favoured by the following two positions of decreasing prominence as noted above, it is further subject to a more specific condition in the Heliand, namely syllable length of the second lift: the first lift of type C is more likely to receive resolution than not when the second one is realised by a long stressed syllable; conversely, it is less likely to do so when it is followed by a short stressed syllable. In other words, the configuration x . . . px#Px occurs with much greater frequency than the configuration x . . . P#Px, as observed by Hirt (1891: 149). According to my count, there are 473 examples of x . . . px#Px (c. 72 per cent)12 as against 187 examples of x . . . P#Px 11

12

As will be shown below, the presence of a heavy drop rather than a stressed syllable as in the Heliand figures as a significant condition in the metre of Beowulf. The motivation for and consequences of this reconfiguration of the generalisation in the Heliand metre will be discussed in due course below. a-verse with double alliteration: 104a, 189a, 228a, 293a, 331a, 549a, 577a, 669a, 807a, 945a, 1017a, 1042a, 1407a, 1521a, 1797a, 1898a, 1913a, 1998a, 2019a, 2041a, 2154a, 2184a, 2256a, 2402a, 2423a, 2500a, 2595a, 2850a, 2858a, 2925a, 2956a, 3094a, 3234a, 3245a, 3295a, 4030a, 4041a, 4069a, 4098a, 4184a, 4210a, 4555a, 4964a, 5162a, 5178a, 5285a, 5293a, 5315a, 5319a, 5486a, 5628a, 5757a, 5763a, 5783a, 5790a, 5796a, 5814a, 5832a, 5871a, 5954a, 5960a; a-verse with single alliteration: 132a, 272a, 283a, 429a, 444a, 457a, 528a, 706a, 806a, 977a, 1168a, 1241a, 1394a, 1564a, 1793a, 1800a, 1969a, 2237a, 2310a, 2314a, 2479a, 2498a, 2509a, 2787a, 2905a, 3199a, 3342a, 3350a, 3475a, 3514a, 3590a, 3742a, 3748a, 3778a, 3901a, 3941a, 3979a, 4149a, 4678a, 4753a, 4839a, 4862a, 5188a, 5222a, 5736a, 5846a, 5895a, 5980a; b-verse: 19b, 25b, 41b, 105b, 120b, 132b, 165b, 208b, 210b, 215b, 218b, 225b, 235b, 245b, 251b, 253b, 257b, 258b, 265b, 283b, 287b, 296b, 302b, 323b, 328b, 375b, 385b, 412b, 423b, 426b, 442b, 452b, 457b, 472b, 473b, 489b, 494b, 499b, 541b, 576b, 592b, 596b, 607b, 623b, 636b, 642b, 644b, 654b, 673b, 674b, 679b, 731b, 737b, 752b, 758b, 773b, 777b, 793b, 809b, 817b, 819b, 835b, 841b, 948b, 949b, 957b, 968b, 1015b, 1018b, 1033b, 1049b, 1081b, 1101b, 1109b, 1158b, 1166b, 1174b, 1185b, 1234b, 1258b, 1297b, 1336b, 1342b, 1373b, 1394b, 1396b, 1420b, 1425b, 1447b, 1450b, 1460b, 1465b, 1503b, 1565b, 1577b, 1579b, 1587b, 1590b, 1604b, 1613b, 1633b, 1637b, 1649b, 1652b, 1662b, 1669b, 1671b, 1707b, 1711b, 1716b, 1722b, 1730b, 1742b, 1754b, 1757b, 1771b, 1777b, 1784b, 1787b, 1793b, 1795b, 1826b, 1839b, 1847b, 1854b, 1860b, 1875b, 1896b, 1913b, 1930b, 1937b, 1943b, 1948b, 1950b, 1954b, 1983b, 2015b, 2033b, 2140b, 2159b, 2177b, 2217b, 2233b, 2235b, 2305b, 2321b, 2326b, 2337b, 2370b, 2375b, 2406b, 2411b, 2434b, 2436b, 2521b, 2523b, 2557b, 2569b, 2594b, 2640b, 2642b, 2654b, 2663b, 2674b, 2682b, 2692b, 2711b, 2726b, 2727b, 2766b, 2769b, 2778b, 2786b, 2792b, 2832b, 2833b, 2840b, 2848b, 2866b, 2935b, 2939b, 2963b,

MOSP-03.qxd

5/11/04

2:41 PM

Page 227

3.1. Resolution

227

(c. 28 per cent)13 in the Heliand. Such a strong preference for resolution on the first lift stands in marked contrast to the low frequency with which resolution is implemented on the same position when the second lift is filled by an unresolved short syllable: the sequence x . . . px#px occurs fifty-seven times (c. 11 per cent),14 while the sequence x . . . P#px is counted 455 (c. 89 per cent).15 From another perspective,

13

14

15

2967b, 2977b, 3019b, 3059b, 3069b, 3082b, 3103b, 3109b, 3117b, 3119b, 3148b, 3158b, 3189b, 3190b, 3194b, 3238b, 3259b, 3289b, 3297b, 3300b, 3311b, 3317b, 3331b, 3343b, 3381b, 3385b, 3432b, 3460b, 3470b, 3489b, 3515b, 3520b, 3536b, 3577b, 3584b, 3605b, 3621b, 3645b, 3652b, 3655b, 3658b, 3662b, 3664b, 3670b, 3674b, 3679b, 3690b, 3695b, 3701b, 3719b, 3736b, 3739b, 3752b, 3769b, 3776b, 3782b, 3812b, 3817b, 3826b, 3839b, 3848b, 3857b, 3875b, 3893b, 3907b, 3931b, 3932b, 3944b, 3959b, 4020b, 4046b, 4067b, 4071b, 4078b, 4095b, 4099b, 4110b, 4129b, 4157b, 4178b, 4192b, 4195b, 4218b, 4242b, 4248b, 4295b, 4306b, 4307b, 4313b, 4318b, 4319b, 4320b, 4325b, 4340b, 4343b, 4397b, 4401b, 4410b, 4434b, 4438b, 4457b, 4459b, 4465b, 4495b, 4496b, 4506b, 4539b, 4570b, 4572b, 4599b, 4641b, 4649b, 4658b, 4661b, 4703b, 4706b, 4709b, 4710b, 4733b, 4739b, 4748b, 4781b, 4787b, 4791b, 4792b, 4829b, 4831b, 4871b, 4884b, 4937b, 4939b, 4940b, 4975b, 4995b, 5057b, 5067b, 5085b, 5112b, 5157b, 5160b, 5254b, 5255b, 5267b, 5272b, 5278b, 5287b, 5300b, 5311b, 5346b, 5347b, 5359b, 5374b, 5383b, 5400b, 5405b, 5410b, 5421b, 5432b, 5451b, 5585b, 5612b, 5616b, 5622b, 5744b, 5763b, 5765b, 5770b, 5790b, 5791b, 5795b, 5801b, 5804b, 5814b, 5823b, 5826b, 5856b, 5874b, 5900b, 5903b, 5950b, 5960b, 5964b. a-verse with double alliteration: 982a, 1477a, 1488a, 1871a, 1977a, 2062a, 2779a, 2894a, 2897a, 2906a, 2922a, 3135a, 4017a, 4740a, 5092a, 5555a, 5857a, 5900a; a-verse with single alliteration: 14a, 103a, 475a, 653a, 1249a, 1858a, 2383a, 2629a, 3358a, 3427a, 3632a, 3907a, 4022a, 4146a, 4235a, 4372a, 4458a, 4534a, 4787a, 4809a, 5181a, 5442a, 5780a; b-verse: 70b, 94b, 163b, 193b, 214b, 232b, 276b, 299b, 327b, 357b, 428b, 451b, 461b, 503b, 578b, 668b, 695b, 713b, 790b, 794b, 810b, 850b, 896b, 916b, 923b, 939b, 994b, 1039b, 1057b, 1062b, 1070b, 1270b, 1343b, 1348b, 1487b, 1538b, 1543b, 1597b, 1661b, 1664b, 1676b, 1789b, 1791b, 1832b, 1856b, 1862b, 1936b, 1941b, 1945b, 1955b, 1956b, 2053b, 2086b, 2101b, 2119b, 2122b, 2162b, 2239b, 2258b, 2323b, 2327b, 2333b, 2340b, 2452b, 2588b, 2607b, 2627b, 2657b, 2698b, 2709b, 2713b, 2715b, 2725b, 2729b, 2921b, 2946b, 2970b, 3034b, 3044b, 3060b, 3081b, 3198b, 3201b, 3207b, 3210b, 3243b, 3368b, 3369b, 3400b, 3442b, 3619b, 3667b, 3733b, 3802b, 3924b, 3952b, 3983b, 4001b, 4042b, 4083b, 4107b, 4138b, 4173b, 4205b, 4231b, 4328b, 4360b, 4377b, 4405b, 4416b, 4420b, 4429b, 4450b, 4460b, 4548b, 4642b, 4678b, 4686b, 4814b, 4817b, 4820b, 4841b, 4851b, 4949b, 5069b, 5239b, 5303b, 5339b, 5387b, 5388b, 5397b, 5419b, 5425b, 5438b, 5445b, 5502b, 5535b, 5561b, 5573b, 5698b, 5726b, 5792b, 5794b, 5821b, 5837b, 5906b. a-verse with single alliteration: 896a, 1456a, 2079a, 2671a, 3678a, 5599a, 5623a; b-verse: 628b, 913b, 1064b, 1084b, 1086b, 1407b, 1986b, 2066b, 2082b, 2192b, 2197b, 2295b, 2369b, 2385b, 2407b, 2435b, 2538b, 2565b, 2669b, 2677b, 2777b, 3132b, 3214b, 3336b, 3547b, 3552b, 3760b, 3906b, 4008b, 4080b, 4133b, 4181b, 4191b, 4203b, 4309b, 4432b, 4463b, 4503b, 4549b, 4738b, 4808b, 4845b, 4890b, 4920b, 4946b, 5042b, 5124b, 5238b, 5283b, 5338b. a-verse with double alliteration: 200a, 313a, 949a, 970a, 3401a, 3821a, 3893a, 4145a, 4454a, 4467a, 5071a, 5132a, 5570a, 5716a; a-verse with single alliteration: 17a, 73a, 94a, 115a, 177a, 227a, 784a, 1015a, 1079a, 1123a, 1131a, 1278a, 1363a, 1387a, 1391a, 1484a, 2143a, 2216a, 2303a, 2326a, 2469a, 2537a, 2539a, 2731a, 2816a, 2932a, 3266a, 3582a, 3634a, 4050a, 4198a, 4324a, 4362a, 4600a, 4628a, 4786a, 5234a, 5631a, 5770a; b-verse: 2b, 10b, 14b, 23b, 81b, 87b, 95b, 118b, 122b, 128b, 129b, 164b, 168b, 169b, 172b, 180b, 182b, 184b, 192b, 205b, 209b, 216b, 229b, 260b, 273b, 280b, 324b, 331b, 352b, 368b, 382b, 391b, 425b, 431b, 437b, 460b, 466b, 474b, 479b, 487b, 501b, 522b, 531b, 544b, 545b, 565b, 572b, 575b, 584b, 626b, 637b, 650b, 651b, 689b, 693b, 702b, 707b, 712b, 714b, 721b, 740b, 767b, 770b, 787b, 825b, 827b, 831b, 840b, 852b, 856b, 866b, 886b, 911b, 915b, 919b, 929b, 932b, 946b, 955b, 999b, 1003b, 1004b, 1065b, 1115b, 1154b, 1164b, 1196b, 1203b, 1260b, 1289b, 1337b, 1354b, 1355b,

MOSP-03.qxd

5/11/04

228

2:41 PM

Page 228

3. Resolution and alliteration

resolution on the first lift is implemented with far higher frequency when the second lift is realised by a long stressed syllable (x . . . px#Px) than by a short one (x . . . px#px): the configuration x . . . px#Px accounts for nearly 90 per cent of all such occurrences of the resolved first lift. Thus, we observe here that the diametrically opposed treatment of resolution is conditioned by the syllable length of the second lift. On the one hand, the privileged treatment of the configuration x . . . px#Px at the expense of x . . . P#Px may be attributed to the same underlying mechanism that motivates suspension of resolution from the second of two consecutive prominent positions in general and the preference for x . . . P#px over x . . . P#Px in particular, discussed earlier. Yet the actual adjustment implemented works in the opposite direction: instead of suppressing prominence on the second position as with the sequence x . . . P#px, additional prominence is given to the first position through association with a disyllable by means of resolution. On the other hand, the infrequent use of the configuration x . . . px#px in favour of x . . . px#Px may be explained as a consequence of a relatively low prominence on the second lift that is realised by a short stressed syllable: to the extent that the short syllable is less prominent than the long counterpart, the preceding first lift is subject to a weaker demand for realising additional prominence by resolution. There is reason for believing that such a near-symmetrical variation on resolution on the first lift that was keyed to syllable length of the second lift constituted another instance of innovation brought about by the Heliand poet by reorganisation. The Beowulf metre shows a distribution pattern which is somewhat similar in broad terms but different in detail: resolution on the first lift is only slightly preferred when the second one is occupied by a long stressed syllable, falling far short of the 1357b, 1359b, 1374b, 1376b, 1381b, 1392b, 1393b, 1432b, 1480b, 1563b, 1574b, 1583b, 1623b, 1627b, 1654b, 1675b, 1679b, 1724b, 1797b, 1806b, 1825b, 1845b, 1853b, 1864b, 1869b, 1888b, 1901b, 1968b, 1978b, 1980b, 1982b, 1985b, 1996b, 2003b, 2018b, 2023b, 2085b, 2105b, 2107b, 2158b, 2180b, 2195b, 2203b, 2225b, 2227b, 2288b, 2289b, 2307b, 2309b, 2312b, 2324b, 2334b, 2342b, 2346b, 2438b, 2446b, 2469b, 2472b, 2474b, 2476b, 2478b, 2481b, 2485b, 2559b, 2564b, 2571b, 2574b, 2591b, 2606b, 2633b, 2638b, 2649b, 2710b, 2750b, 2796b, 2807b, 2831b, 2851b, 2875b, 2881b, 2895b, 2928b, 2929b, 2960b, 2962b, 2964b, 2998b, 3025b, 3050b, 3102b, 3173b, 3178b, 3226b, 3233b, 3237b, 3262b, 3280b, 3284b, 3291b, 3340b, 3351b, 3374b, 3396b, 3402b, 3440b, 3481b, 3488b, 3560b, 3686b, 3707b, 3727b, 3732b, 3740b, 3755b, 3778b, 3785b, 3799b, 3819b, 3830b, 3862b, 3878b, 3910b, 3928b, 3939b, 3950b, 3951b, 4010b, 4032b, 4043b, 4047b, 4063b, 4077b, 4081b, 4098b, 4122b, 4164b, 4171b, 4193b, 4194b, 4230b, 4252b, 4257b, 4265b, 4274b, 4289b, 4317b, 4336b, 4348b, 4359b, 4366b, 4376b, 4379b, 4404b, 4470b, 4482b, 4485b, 4491b, 4492b, 4527b, 4534b, 4535b, 4550b, 4608b, 4609b, 4637b, 4665b, 4669b, 4670b, 4671b, 4687b, 4698b, 4730b, 4751b, 4770b, 4771b, 4779b, 4801b, 4822b, 4838b, 4844b, 4883b, 4903b, 4907b, 4929b, 4933b, 4943b, 4947b, 4957b, 4964b, 5052b, 5054b, 5055b, 5061b, 5074b, 5095b, 5101b, 5117b, 5122b, 5190b, 5203b, 5205b, 5206b, 5261b, 5270b, 5286b, 5296b, 5313b, 5365b, 5378b, 5390b, 5434b, 5464b, 5482b, 5520b, 5523b, 5534b, 5590b, 5608b, 5646b, 5653b, 5669b, 5686b, 5691b, 5729b, 5825b, 5865b, 5868b, 5870b, 5896b, 5908b, 5910b, 5940b, 5971b, 5980b.

1495b, 1753b, 1919b, 2046b, 2253b, 2376b, 2488b, 2650b, 2913b, 3125b, 3302b, 3583b, 3804b, 3987b, 4123b, 4280b, 4419b, 4578b, 4723b, 4853b, 5015b, 5130b, 5336b, 5539b, 5831b,

1550b, 1775b, 1927b, 2062b, 2264b, 2415b, 2497b, 2656b, 2916b, 3152b, 3304b, 3642b, 3808b, 3994b, 4160b, 4282b, 4441b, 4585b, 4726b, 4867b, 5031b, 5138b, 5342b, 5540b, 5840b,

1557b, 1788b, 1951b, 2070b, 2267b, 2419b, 2500b, 2675b, 2918b, 3161b, 3322b, 3653b, 3810b, 4007b, 4161b, 4287b, 4448b, 4588b, 4727b, 4882b, 5032b, 5178b, 5361b, 5562b, 5863b,

MOSP-03.qxd

5/11/04

2:41 PM

Page 229

3.1. Resolution

229

pronounced preference for resolution in the Heliand: x . . . px#Px (135 examples; c. 54 per cent) versus x . . . P#Px (116 examples; c. 46 per cent); by contrast, the avoidance of resolution on the first lift when followed by the short stressed syllable is nearly mandatory (Hirt 1891: 153; Heusler 1956: §232a): x . . . px#px (three examples; c. 1 per cent) versus x . . . P#px (262 examples; c. 99 per cent). Confronted with the lack of symmetry between the two pairs x . . . px#Px/x . . . P#Px and x . . . px#px/x . . . P#px in their distribution pattern, the Heliand poet would have remade the inherited uneven distribution into a more balanced pattern in accord with his high respect for order and harmony, even as he made more extensive use of resolution in the contexts followed by two positions of decreasing prominence, as we saw earlier in this section: as a consequence, these two pairs of configurations were made more like each other in their overall distribution, so that the configurations x . . . px#Px and x . . . P#px were preferred in comparable degrees to x . . . P#Px and x . . . px#px, respectively. It thus seems unwarranted to attribute the increased use of x . . . px#px to the Heliand poet’s ‘weniger feine metrische Gefühl’ as did Hirt (1891: 153). An analogous generalisation partly applies to the set of variants in which the second lift is occupied by a secondary-stressed syllable falling on the second element of a compound. At stake here are the following four configurations, each attested in the corpus as indicated in parentheses: x . . . pxSx (124 examples),16 x . . . PSx (120 examples),17 x . . . Psx (156 examples),18 x . . . pxsx 16

17

18

a-verse with single alliteration: 80a, 162a, 388a, 434a, 436a, 513a, 609a, 618a, 632a, 869a, 986a, 1116a, 1226a, 1290a, 1575a, 1651a, 1839a, 1849a, 1979a, 2029a, 2131a, 2201a, 2287a, 2437a, 2487a, 2508a, 2605a, 2613a, 2624a, 2736a, 2757a, 2791a, 2829a, 3150a, 3175a, 3259a, 3744a, 3843a, 3852a, 4180a, 4269a, 4363a, 4515a, 4559a, 4668a, 4694a, 4708a, 4878a, 4927a, 4999a, 5174a, 5220a, 5235a, 5386a, 5396a, 5563a, 5896a; b-verse: 45b, 275b, 321b, 325b, 411b, 414b, 488b, 551b, 672b, 859b, 1002b, 1022b, 1043b, 1051b, 1096b, 1098b, 1153b, 1161b, 1303b, 1375b, 1510b, 1635b, 1867b, 1870b, 1914b, 2133b, 2275b, 2421b, 2486b, 2495b, 2623b, 2626b, 2637b, 2645b, 2842b, 2885b, 3080b, 3199b, 3217b, 3222b, 3330b, 3597b, 3617b, 3643b, 3659b, 4141b, 4255b, 4277b, 4301b, 4357b, 4422b, 4461b, 4473b, 4519b, 4887b, 4918b, 4922b, 5014b, 5094b, 5224b, 5391b, 5488b, 5604b, 5714b, 5842b, 5962b, 5969b. a-verse with double alliteration: 1027a, 2144a, 3294a, 3447a; a-verse with single alliteration: 4a, 84a, 89a, 99a, 155a, 183a, 210a, 238a, 244a, 353a, 376a, 398a, 417a, 616a, 640a, 655a, 824a, 839a, 857a, 868a, 987a, 1034a, 1038a, 1111a, 1121a, 1155a, 1193a, 1212a, 1254a, 1351a, 1410a, 1520a, 1534a, 1636a, 1817a, 1834a, 1874a, 1953a, 2117a, 2419a, 2468a, 2641a, 2651a, 2660a, 2679a, 2811a, 2880a, 3041a, 3049a, 3120a, 3158a, 3167a, 3424a, 3492a, 3511a, 3592a, 3727a, 3826a, 4169a, 4294a, 4387a, 4412a, 4551a, 4594a, 4803a, 5026a, 5051a, 5131a, 5172a, 5221a, 5236a, 5275a, 5290a, 5301a, 5328a, 5382a, 5403a, 5404a, 5521a, 5645a, 5739a, 5807a, 5877a; b-verse: 248b, 998b, 1001b, 1036b, 1200b, 1274b, 1469b, 1490b, 1615b, 1878b, 1929b, 2084b, 2492b, 2600b, 2756b, 3461b, 3598b, 3722b, 3762b, 4356b, 4573b, 4971b, 4979b, 4983b, 5043b, 5063b, 5152b, 5193b, 5204b, 5237b, 5444b, 5498b, 5920b. a-verse with single alliteration: 44a, 108a, 138a, 230a, 285a, 309a, 344a, 362a, 709a, 730a, 789a, 870a, 875a, 1050a, 1413a, 1445a, 1553a, 1868a, 1899a, 1905a, 1909a, 2128a, 2441a, 2461a, 2610a, 2727a, 2728a, 2762a, 2783a, 2785a, 2979a, 3010a, 3260a, 3283a, 3285a, 3338a, 3366a, 3397a, 3555a, 3731a, 3736a, 3801a, 3873a, 3917a, 4045a, 4085a, 4099a, 4148a, 4156a, 4157a, 4197a, 4201a, 4219a, 4225a, 4228a, 4466a, 4477a, 4530a, 4557a, 4623a, 4734a, 4773a, 5005a, 5143a, 5305a, 5340a, 5408a, 5415a, 5422a, 5462a, 5662a, 5672a, 5704a, 5772a, 5787a, 5852a, 5853a; b-verse: 241b, 249b, 539b, 582b, 622b, 786b, 879b, 920b, 921b, 1032b, 1046b, 1080b, 1095b, 1113b, 1119b, 1239b, 1401b, 1414b, 1438b, 1488b, 1540b, 1728b, 1750b, 1833b,

MOSP-03.qxd

5/11/04

230

2:41 PM

Page 230

3. Resolution and alliteration

(27 examples).19 Resolution on the first lift is strongly favoured when the second lift is realised by a long syllable (x . . . pxSx versus x . . . pxsx), in parallel to the opposition between x . . . px#Px and x . . . px#px. The parallelism does not go any further, however. The difference between x . . . pxSx and x . . . PSx is hardly similar to the one involving the pair of x . . . px#Px and x . . . P#Px: the association of the second lift with a long syllable fails to correlate with a significant increase in implementation of resolution on the first lift. This divergence between the primary and secondary stresses associated with the second lift, however, would seem readily explainable as a natural consequence of the difference in prominence between the two stresses: the secondary stress is not nearly as prominent as the primary one, and thus hardly endangers the underlying prominence contour, so that it is scarcely needed to reinforce the preceding lift by added prominence through resolution. Moving onto the second major difference between the Heliand and Beowulf, the second lift of the configuration P#PS/sx is resolved with great incidence in the Heliand, but not so frequently in Beowulf, while that of the configuration Px . . . Pxx is resolved extensively in Beowulf, yet it is minimally affected by resolution in the Heliand. The diametrically opposed treatment in the two works of the second lift in P#PS/sx on the one hand and in Px . . . Pxx on the other is systematically related, as will become clear below. As regards the configuration P#PS/sx, the presence of P in the immediately preceding position (i.e., the first lift) was of vital significance in Beowulf (Suzuki 1996a: 198–9): since the first lift had to be resolved most extensively as observed above, a similar extensive operation on the second lift would have been found at odds with the requirement to maximise difference in prominence between the two consecutive strong positions, the first and second lifts. Accordingly, while the second lift of the configuration Px . . . PS/sx was frequently resolved by virtue of the immediately preceding, intervening weak element (x), that of the minimally distinct configuration P#PS/sx was usually precluded from resolution in Beowulf. The consideration of this kind, however, would not have been appreciated by the Heliand poet, who attached greater importance to the common property of two positions of decreasing prominence being present in the following. Such a re-evaluation may plausibly be viewed as a manifestation of the overall metrical organisation of the Heliand, in which consideration of maximal balance and harmony overrode that of maximal asymmetry and contrast underlying the metrical organisation of Beowulf (Chapter 5; see also Suzuki 1996a: 389–92). In this connection, it may be reminded that the traditional distinction between the first and second lifts was considerably weakened in the Heliand, as witnessed, for example, in the partial dysfunctioning of the principle of left dominance largely responsible for alliterative patterning (see section 3.2.2 below). Thus, the original motivation for keeping the two lifts distinct

19

1871b, 1910b, 1932b, 2110b, 2271b, 2301b, 2351b, 2575b, 2604b, 2733b, 2738b, 3221b, 3327b, 3348b, 3375b, 3405b, 3453b, 3458b, 3468b, 3523b, 3528b, 3595b, 3816b, 3834b, 3898b, 4014b, 4221b, 4352b, 4362b, 4437b, 4544b, 4553b, 4614b, 4620b, 4638b, 4754b, 4799b, 5002b, 5058b, 5075b, 5128b, 5199b, 5200b, 5221b, 5250b, 5256b, 5266b, 5277b, 5309b, 5389b, 5489b, 5553b, 5583b, 5724b, 5902b. a-verse with single alliteration: 931a, 3790a, 3885a, 4574a, 4930a, 5334a, 5409a, 5441a, 5461a, 5465a; b-verse: 58b, 100b, 343b, 533b, 2294b, 2503b, 2718b, 2874b, 3792b, 3905b, 4227b, 4742b, 5246b, 5284b, 5285b, 5420b, 5586b.

MOSP-03.qxd

5/11/04

2:41 PM

Page 231

3.1. Resolution

231

as sharply as possible on the surface would not have appealed to the poet’s sense of balance and harmony in organisation. Once freed from the traditional practice of respecting the preceding strong position, the Heliand poet set out to explore an alternative way of organisation that he appreciated to be more balanced and systematised in overall structural terms. In doing so, the poet could hardly have missed the differing behaviour in the traditional metre of three major variants of type C, x . . . P#P/px, x . . . PS/sx, and x . . . PX/xx, which varied in the incidence of resolution on the first lift, as shown in Table 3.7. This insight would have been readily available to the poet as he came to recognise the variation of resolution on the first lift to be based on syllable length of the second lift, as discussed in this section above. The variation at issue on resolution in type C was clearly determined by the stress properties of the immediately following syllable, primary-stressed (the roots of independent words), secondary-stressed (the second elements of compounds), or non-lexical-stressed (the derivational elements of non-compound words). Worth emphasising here is that these variants were indistinguishable in metrical structure in terms of the configuration of metrical positions involved: they all reduced to the underlying form ⫻ / / ⫻ (i.e., drop ⫹ lift ⫹ lift ⫹ drop), with the third position identified as a lift. Distinctive rather were the linguistic (as opposed to metrical) properties that were associated with the third position in question.20 The patterned variation on resolution as observed above in the variants of type C was paralleled in part by that of types D1 (P#Pxx) and D2a (P#PSx) in the traditional metre, this time concerning the second lift: it was subject to resolution with higher frequency in type D2a than in type D1, and this difference may be attributed to the nature of the following (third) position. Yet these two configurations constituted distinct metrical types in the traditional metre, the third position being distinguished primarily in terms of the status of the metrical positions involved, a normal drop versus a heavy drop. In traditional practice, the variation on resolution that we find relatable to distinction in stress was thus less than self-evident and therefore remained unexploited for structural organisation. Acutely recognising potential significance of stress in effecting variation on resolution in the traditional metre, the Heliand poet built on this insight to reorganise variation on resolution including the treatment of the second lift of the configurations P#PS/sx and P#Pxx. In so doing, the poet differentiated the inherited variation sharply and made it demonstrably patterned. Since these two configurations differed in terms of stress much in the same way as did the two variants of type C, x . . . PS/sx and x . . . PX/xx, they became analogously graded in sharper form than in the traditional metre, so much so that the second lift of the configuration P#PS/sx, as well as the first lift of x . . . PS/sx and also of x . . . P#P/px (type C), were made to undergo resolution with far greater frequency. Conversely, in opposition to these, resolution was rendered hardly applicable to the first lift of the configuration x . . . PX/xx (type C; cf. Tables 3.6 and 3.7), as well as to the second lift of the configuration Px . . . Pxx (subtype D*1), as more fully discussed in the following paragraph. 20

In more general terms, the Heliand metre attaches greater importance to stress properties of individual language materials than to underlying metrical positions, as observed in the poet’s reorganisation of hypermetric verses (see section 4.2 below).

MOSP-03.qxd

5/11/04

232

2:41 PM

Page 232

3. Resolution and alliteration

In regard to the configuration Px . . . Pxx, the Heliand would appear to fall short of making full of the generalisation that applies to Beowulf: the second lift of the configuration Px . . . Pxx is followed by two drops that decrease in prominence according to the linearity-based scaling of prominence, whereby the like positions become less prominent toward the end of the verse (Suzuki 1996a: 165–7, 199; see also section 1.2 above). On closer consideration, however, we may become convinced that the Heliand poet consistently followed consequences of his own generalisation that we have identified above. Unlike the Beowulf poet, the Heliand poet did not feel himself bound by the need to distinguish categorically between the first and second lifts; accordingly he disregarded the additional condition of the absence of an immediately preceding stronger position. With such a logic in mind, he would hardly have failed to notice a structural disparity between the configurations P#Pxx and Px . . . Pxx with respect to the treatment of the second lift: one was resolved extensively and the other scarcely so despite the indistinguishable (according to his own thinking) contexts involved. Most consistent would be to have the second lift of both configurations resolved with comparable frequency, whether enhanced or diminished. Such an inconsistency inherent in traditional practice would have prompted the poet to explore a consistent and more harmonious treatment of the two configurations at issue. In the face of the two logically possible, mutually exclusive solutions, the poet decided on retaining the inherited low incidence of resolution for P#Pxx at the expense of Px . . . Pxx, thereby suppressing frequency of resolution on the second lift of Px . . . Pxx. His decision proved to be reasonable against the background of the reorganisation that he implemented elsewhere by systematising the extant but unexploited resources of variation: because the second lift of the configuration Px . . . Pxx was immediately followed by an unstressed syllable, it was made largely immune to resolution, as were the first lift of the light variants of type C (x . . . PX/xx) and the second lift of subtype D1 (P#Pxx). The Heliand poet made further innovative use of the traditional variation by generalising it to cover the two variants of type E, PXx . . . P and PSx . . . P. As regards the frequency of resolution on the first lift, the two configurations came to be subsumed under the same patterned variation that we have seen characteristic of the Heliand. Indeed, they may be distinguished in the traditional metre to a limited extent, but not much in pronounced form (Table 3.7). Consistently following the overall scheme of reorganisation, the Heliand poet thus articulated the distinction more sharply by polarising the inherited pattern (Table 3.6). A further consequence of the reorganisation of the variation on resolution concerns the second lift of type A2 (Px . . . PS) in the Heliand, which is characterised by the enhanced incidence of resolution in comparison with its counterpart in Beowulf (compare Tables 3.6 and 3.7). As may be recalled, resolution on the second lift of type A2b was favoured in Beowulf only to a moderate extent by the immediate adjacency of a stressed syllable or a heavy drop, a generalisation of obviously limited applicability. In the presence of a stressed syllable after the lift in question, the Heliand poet would have subsumed this case as well under the new generalisation that he came to establish between varied preference for resolution and the stress properties of the following syllable. As a result of being integrated into the rule of extensive applicability, the second lift of type A2 was made more

MOSP-03.qxd

5/11/04

2:41 PM

Page 233

3.1. Resolution

233

amenable to resolution in the Heliand than in Beowulf, and concomitantly this type became differentiated in sharper shape from type A1. The Heliand poet’s appropriation of the new generalisation did not stop here, however. It went into finer detail and brought about more profound effect: he not only applied resolution to the second lift of type A2 with greater frequency, but also made the implementation of resolution graded as keyed to the following three distinct groups of the language materials that occupied the second lift and drop of the configuration at issue: (i) independent lexical words; (ii) true compounds; (iii) quasi compounds. The three variants of type A2 thus differentiated are exemplified in (30) below: (30) 1557a gerno thurh godes thanc (independent word) 2354a helið an helsîð (true compound) 2892a hebbian ênigan hêrdôm (quasi compound) These three variants of type A2 vary significantly in the extent to which the second lift is affected by resolution, as represented in Table 3.8. The neat gradation pattern of resolution that we recognise in the Heliand stands in contrast to the situation in Beowulf, as shown in Table 3.9. The difference in morpholexical status is in turn most likely to have had as its phonological correlate a comparable difference in degrees of prominence: a greater degree of independence in morpholexical terms would have been matched with a higher degree of perceptibility in phonological terms. Given such a plausible correlation, we may be justified in claiming that the graded implementation of resolution on the second lift of type A2 would have been an innovation that the Heliand poet introduced by carrying his reorganisation of the variation on resolution to its logical conclusion. We witness here once again the Heliand poet’s high sensitivity to and firm control of the metrical organisation of the poem in its totality. Table 3.8. Distribution of resolution on the second lift of type A2 in relation to the morpholexical status of the words appearing in the second lift and drop

independent lexical word true compound quasi compound

resolved

not resolved

11 (100%) 23 (43%) 2 (15%)

0 (0%) 30 (57%) 11 (85%)

Note: the underlined verses have their second lift resolved independent word: 1557a, 1865a, 2204a, 3082a, 3450a, 3478a, 3734a, 4275a, 4622a, 4909a, 5970a true compound: 156a, 264a, 554a, 570a, 674a, 753a, 948a, 1032a, 1345a, 1508a, 1916a, 1955a, 2070a, 2173a, 2233a, 2240a, 2268a, 2289a, 2343a, 2354a, 2360a, 2585a, 2655a, 2682a, 2702a, 2712a, 2881a, 2931a, 2947a, 3106a, 3303a, 3323a, 3747a, 3842a, 3860a, 4134a, 4330a, 4469a, 4658a, 4688a, 4692a, 4695a, 4731a, 4954a, 5060a, 5116a, 5124a, 5137a, 5167a, 5210a, 5299a, 5321a, 5400a quasi compound: 848a, 1024a, 1393a, 1638a, 1846a, 1891a, 2005a, 2892a, 3897a, 4343a, 4492a, 5209a, 5252a

MOSP-03.qxd

5/11/04

2:41 PM

234

Page 234

3. Resolution and alliteration

Table 3.9. Distribution of resolution on the second lift of type A2b in Beowulf in relation to the morpholexical status of the words appearing in the second lift and drop

independent lexical word true compound quasi compound

resolved

not resolved

1 (20%) 23 (23%) 5 (16%)

4 (80%) 76 (77%) 27 (84%)

Note: the underlined verses have their second lift resolved independent word: 330a, 736a, 2509a, 2638a, 3105a true compound: 33a, 49a, 58a, 65a, 82a, 127a, 130a, 131a, 193a, 216a, 217a, 277a, 305a, 308a, 320a, 380a, 408a, 485a, 515a, 558a, 594a, 608a, 619a, 641a, 688a, 690a, 693a, 755a, 773a, 780a, 823a, 826a, 912a, 922a, 989a, 1060a, 1087a, 1127a, 1161a, 1200a, 1277a, 1329a, 1424a, 1430a, 1444a, 1489a, 1506a, 1522a, 1527a, 1529a, 1533a, 1545a, 1546a, 1559a, 1563a, 1564a, 1575a, 1698a, 1719a, 1758a, 1785a, 1794a, 1799a, 1800a, 1881a, 1921a, 2037a, 2085a, 2132a, 2172a, 2175a, 2196a, 2252a, 2289a, 2292a, 2296a, 2342a, 2370a, 2420a, 2586a, 2593a, 2625a, 2660a, 2661a, 2682a, 2691a, 2704a, 2892a, 2893a, 2953a, 2957a, 2980a, 2991a, 3021a, 3041a, 3090a, 3118a, 3132a, 3155a quasi compound: 134a, 192a, 215a, 232a, 251a, 287a, 336a, 350a, 357a, 413a, 761a, 1000a, 1246a, 1267a, 1562a, 1650a, 1925a, 2086a, 2109a, 2119a, 2154a, 2188a, 2214a, 2339a, 2449a, 2564a, 2618a, 2624a, 2739a, 2929a, 3135a, 3168a

The central significance that the Heliand poet thus attached to the immediately following stressed syllable for an increased use of resolution would have brought about a further reorganisation in implementation of resolution: it would have made more visible the role of the immediately adjacent stressed syllable for resolution on the second lift in traditional practice. In Beowulf, as observed above, the second lift of types A2a and A2b was resolved slightly more frequently than the first one and also than the second lift of type A1 variants (Table 3.7) by virtue of the immediate proximity of a heavy drop. Figuring now centrally in metrical organisation, the presence of the immediately following stressed syllable would have provided a firmer basis of generalisation for the minor rule at issue, which was constructed in terms of the immediately adjacent heavy drop in the traditional metre.21 Subject to a greater degree of integration in the metrical system through its recasting in terms of the stressed syllable, the traditional rule in question would then have been capable of extending its domain of operation. Specifically, the configuration x . . . PS#P, a variant of type B1, came to fall under the reconfigured rule in the Heliand metre, because the second lift was immediately preceded by a stressed syllable rather than a heavy drop. As noted in section 2.4.2.2 and shown in Table 3.6 above, the second lift of this configuration received resolution extensively in the Heliand. By contrast, resolution occurred on none of the eight instances of the configuration in Beowulf (202a, 615a, 821a, 232b, 537b, 1940b, 1941b, 2109b; cf. Suzuki 1996a: 92), precisely because in it the rule was predicated on the presence of a heavy drop rather than a stressed syllable. Such 21

This may be regarded as a further instance of the Heliand poet’s privileged use of the syllable over the metrical position.

MOSP-03.qxd

5/11/04

2:41 PM

Page 235

3.2. Alliteration

235

reorganisation of the rule was thus made possible by the overall reconfiguration of the immediately following stressed syllable as a central inducing factor of resolution in the Heliand.

3.2. Alliteration Reduced to its essentials, the alliteration of the Heliand is no different from that of Beowulf (cf. Sievers 1893: §119; Lehmann 1953: 23–4): it consistently conforms to the set of rules for alliteration that was laid out for Beowulf in Suzuki (1996a: 277), as summarised at the beginning of the three sections below that are devoted to the three complementary bases of alliteration, metrical (3.2.1), lexical and syntactic (3.2.2), and phonological (3.2.3). Yet the Heliand poet reconstituted some of the traditional rules and constraints in order to adjust whatever discrepancies arose between linguistic and metrical systems and thereby to achieve greater balance and harmony in verse structure. 3.2.1. The metrical basis of alliteration 3.2.1.1. Introductory remarks Alliteration is an obligatory metrical device that involves repetition of initial consonants of lifts in the line (for details on the phonological basis of alliteration, see section 3.2.3 below). Since the minimal and maximal numbers of the lift per verse are one and two for the normal verse, respectively, alliteration may occur once (single alliteration) or twice (double alliteration) in a verse. In the a-verse and the b-verse alike, the first lift must always materialise alliteration (the principle of left dominance). By contrast, the second lift may but need not alliterate in the a-verse, while it may not alliterate in the b-verse. Thus, the a-verse allows both for single and double alliteration, whereas the b-verse is compatible only with single alliteration. In accordance with the variable and constant numbers of the lift, one or two for the a-verse, and two for the b-verse, respectively, we have three distinct patterns of alliteration (a ⫽ alliterative; x ⫽ non-alliterative): aa (double alliteration; the a-verse, with two lifts); ax (single alliteration; the a-verse and the b-verse, both with two lifts); xa (single alliteration; the a-verse with a single lift, the non-alliterative element x here being a drop, that is, a replacement of the first lift; see section 2.12.1 above). Of the two or three alliterating lifts occurring in a line, the first lift of the b-verse may be regarded as pivotal for alliteration because of its invariable involvement in the metrical device. Apart from the defective verses presumably due to manuscript corruptions (verses 2a, 2516b, 4264b, 4704a; Lehmann 1953: 24), we should note verse 3962b That scolda sinnon uuel, which shows the alliterative pattern xa, rather than ax, in the b-verse. Other examples of b-verses with the xa pattern in the manuscripts, as provided in Rieger (1876: 7–8), were subsequently emended in one way or another, and exhibit the normal pattern ax in our current text. Moreover, we find several b-verses with double alliteration in our text as follows (cf. Rieger 1876: 9–10, 12): 314b huô he thea thiornun thô; 3020b huelpos huerbad; 3236b Ef he than ôk uuendien ne uuili; 3691b thes thu te uuârun ni uuêst; 5892b Than uuas eft gihêlid hugi.

MOSP-03.qxd

5/11/04

2:41 PM

236

Page 236

3. Resolution and alliteration

Table 3.10. Overall distribution of single and double alliteration in the Heliand and Beowulf

Heliand Beowulf

single alliteration

double alliteration

total

8220 (69%) 4785 (75%)

3740 (31%) 1571 (25%)

11960 (100%) 6356 (100%)

Table 3.11. Distribution of alliterative patterns in the a-verse in the Heliand and Beowulf

Heliand Beowulf

aa

ax

xa

total

3740 (63%) 1571 (49%)

1817 (30%) 1291 (41%)

426 (7%) 315 (10%)

5983 (100%) 3177 (100%)

Table 3.10 shows the overall distribution of single and double alliteration in the Heliand (disregarding the apparently aberrant verses mentioned in the last paragraph) and Beowulf, while Table 3.11 illustrates the distribution of the three available patterns in the a-verse (cf. von See 1967: 65). Through comparison with Beowulf (Tables 3.10 and 3.11), we find that double alliteration is sharply increased in the Heliand. Given that well over the half of the a-verses manifest double alliteration, it may be warranted to conclude that double alliteration is no longer a marked pattern within the a-verse in the Heliand, although it is still marked in the whole context of verses. The enhanced incidence of double alliteration in the Heliand may be related to the obliteration of increased metrical types which are distinguished by marked preference for the a-verse with double alliteration. We shall explore implications of this relocated status of double alliteration in the Heliand in due course. A further rule concerns the placement of a single alliterating element in the a-verse: avoid single alliteration in the fourth position. The verse-final position is thus prohibited from bearing alliteration on its own. Accordingly, the alliterative pattern xa nearly always involves the third position as an alliterative element. This generalisation, however, admits of the following exceptions: (31) 235a 573a 1533a 2367a 2756a 3693a 3962a 4038a 5745a

Thô nam he thia bôk an hand huuand im habde forliuuan that gi sô ni uurecan huô sie scoldin gehalon sô hues sô thu mi bidis huô thu noh uuirðis behabd an is hêlagun uuord that he it thi sân fargibid huô sia eft te them grabe

It would appear far from accidental that the majority of the above exceptions have their alliterative lifts resolved. Such a high proportion of occurrence of resolution on the fourth position would make sense if we assumed that implementation of

MOSP-03.qxd

5/11/04

2:41 PM

Page 237

3.2. Alliteration

237

resolution pre-empted alliteration from occurring at the very end of a verse in terms of syllable concatenation, thereby rendering the deviant verses involved less exceptional. Resolution thus appears to bring about a lesser violation to the generalisation that alliteration may not exclusively affect the fourth position, although the fact remains that with or without resolution the verse-final position carries alliteration. Since the maximal number of alliterating elements allowed in a verse is two, apparent triple alliteration as in the following verses should be regarded as only accidental: (32) 1315a 3062a 3251a 5170a 5452a

thie hebbiad iro herta gihrênod sâlig bist thu Simon, quað he, sunu Ionases sibun sîðun sibuntig diap dôdes dalu an heliðhelme bihelid

3.2.1.2. Metrical types and alliteration As fully discussed in Chapter 2 above, variants of a metrical type show differing patterns of alliteration. Specifically, we have examined in detail the metrical properties of major variants of each type, including their varying distribution in the a-verse and the b-verse, and thereby identified an array of conditioning factors leading to an increased preference for the a-verse with double alliteration within each metrical type. The pattern of single and double alliteration is thus in no small measure determined by structural properties of a given metrical configuration. In this section, we focus on the minimal (least prominent) variant of each metrical type that is characterised as being the least favourable to the a-verse with double alliteration in default of particular conditions for inducing double alliteration, and consider whether there is any notable difference among the major metrical types in regard to the distinction between single and double alliteration. Since the secondarystressed syllable is far less likely to manifest alliteration, we should disregard those variants in which the second lift is realised by a non-primary-stressed syllable. As shown in Table 3.12, the five-position metrical types, i.e., types D* and E*, are clearly distinguished from the other types by the overwhelming frequency of double alliteration. The preponderance of double alliteration in these five-position Table 3.12. Proportion of single and double alliteration in the minimal variants of major metrical types in the a-verse metrical type

double alliteration

single alliteration

type A1 (Px#Px) type B1 (x . . . Px#P) type C (x . . . P#P/px) type D (P#PX/xx) type D* (Px#PX/xx) type E (PXx#P) type E* (Px#Px#P)

221 (51%) 141 (33%) 95 (45%) 20 (43%) 29 (91%) 11 (29%) 23 (85%)

210 (49%) 281 (67%) 117 (55%) 26 (57%) 3 (9%) 27 (71%) 4 (15%)

MOSP-03.qxd

5/11/04

2:41 PM

Page 238

238

3. Resolution and alliteration

types may be viewed as a consequence of the presence of an extra drop, which adds to their overall prominence. Among the four-position metrical types, types A1, C, and D stand out from the rest by virtue of the nearly even distribution of single and double alliteration. At the opposite extreme are found types B1 and E, which show marked preference for single alliteration: nearly two-thirds or more of the a-verses realise single alliteration. Since these two metrical types exclusively share the common property of ending in a lift, it seems warranted to isolate this feature as significantly correlative to increased incidence of single alliteration. We may then generalise that the four-position types are not keyed to particular preference for single or double alliteration, unless they end in a lift, in which case single alliteration is markedly preferred. Of interest in this connection will be a comparison with the situation in Beowulf, shown in Table 3.13. As it turns out, Beowulf shows a rather different pattern on two counts. First, single alliteration prevails over double alliteration except for type E, which is unique for its preference for double alliteration. Thus, the markedly stronger preference for single alliteration in Beowulf than in the Heliand, the pattern pointed out in section 3.2.1.1 above, is common to all but one metrical type in the poem. It should be noted further that even the expanded verse form, subtype D*1, strongly favours single alliteration in sharp contrast to its comparable type in the Heliand. This seems hardly surprising, however, when we are reminded that this expanded configuration (subtype D*1) is a variant of type D1 and accordingly that the apparent extra position between the lifts does not count as an independent position contrary to the corresponding element in the Heliand; therefore, subtype D*1, a variant of type D1, is hardly differentiated from the prototype of type D1 in terms of alliterative patterning in Beowulf. Second, type E, distinguished by its unique preference for double alliteration as noted above, is found diametrically opposed to its comparable in the Heliand, which realises double alliteration with the lowest frequency. We may recall here that the presence of a lift at the end of a verse constitutes a favourable condition for single alliteration in the Heliand metre. The relatively high incidence of double alliteration for type E in Beowulf, on the other hand, may be attributed to the status of the second position being a heavy drop, as this position is incompatible with a medial short unstressed syllable (x), a syllable with minimal prominence.

Table 3.13. Proportion of single and double alliteration of the minimal variants of major metrical types in the a-verse in Beowulf metrical type

double alliteration

single alliteration

type A1 (Px#Px) type B1 (x . . . Px#P) type C (x . . . P#P/px) type D1 (P#PX/xx) subtype D*1 (Px# PX/xx) type E (PXx#P)

118 (31%) 54 (34%) 38 (47%) 16 (29%) 5 (14%) 11 (65%)

266 (69%) 104 (66%) 43 (53%) 39 (78%) 31 (86%) 6 (35%)

MOSP-03.qxd

5/11/04

2:41 PM

Page 239

3.2. Alliteration

239

The divergent treatment of type E in Beowulf and the Heliand in regard to alliterative patterning may lend itself to a credible account: as discussed in detail in section 2.9 above, in the Heliand metre the first drop of type E came closest to, if not identified with, a normal drop through reorganisation by analogy with the corresponding position of type A1: the first drop of type E became qualified for association with a wide spectrum of language materials in ways highly comparable to that of type A1. As a consequence of the reorganisation, the position in question would have been deprived of the original privilege that it had received in the traditional metre concerning double alliteration. 3.2.1.3. Anacrusis and alliteration In section 2.11 above, we discovered that anacrusis was reconstituted as an optional underlying metrical position in the Heliand; as a consequence, the strict restrictions on its use in the traditional versecraft were obliterated, so much so that anacrusis may now attend on the b-verse as well as on the a-verse; that it may co-occur with single alliteration as well as double alliteration; and that it may attach to type E as well as to types A1 and D. By far the most significant change in use of anacrusis in relation to alliteration is accordingly that anacrusis no longer serves as an embellishment privileged to the a-verse with double alliteration. A question arising then is whether there is no more significant correlation between anacrusis and the a-verse with double alliteration; or posed another way, whether anacrusis is implemented in a manner scarcely sensitive to the distinction in alliterative patterning. A preliminary answer to this question may be provided by a cursory look at the overall distribution of anacrusis in the Heliand in comparison with that in Beowulf, shown in Tables 3.14 and 3.15. As indicated in Table 3.14, anacrusis in the Heliand cannot be characterised as a privilege of the a-verse with double alliteration: only slightly over a half of anacrustic verses manifest double alliteration. Further, a comparison with the distribution within the a-verse (Table 3.15) will show that the dissociation of anacrusis from double alliteration is largely a consequence of the radical increase of anacrustic b-verses, rather than of a-verses with single alliteration. Anacrustic a-verses with single alliteration remain largely underrepresented in the Heliand as well. Table 3.14. Distribution pattern of anacrustic verses in the Heliand and Beowulf

Heliand Beowulf

a-verse (aa)

a-verse (ax)

b-verse

total

911 (58%) 75 (84%)

126 (8%) 5 (6%)

533 (34%) 9 (10%)

1570 (100%) 89 (100%)

Table 3.15. Distribution of anacrustic verses in the a-verse in the Heliand and Beowulf

Heliand Beowulf

a-verse (aa)

a-verse (ax)

total

911 (88%) 75 (94%)

126 (12%) 5 (6%)

1037 (100%) 80 (100%)

MOSP-03.qxd

5/11/04

2:41 PM

240

Page 240

3. Resolution and alliteration

In points of detail, however, the above overall picture obscures what correlations there are between anacrusis and alliterative patterning. To this end, we need a more microscopic perspective: we have to consider alliterative patterns at the level of variants of metrical types, rather than in broad terms across types or at the level of metrical types. Tables 3.16 through 3.19 show alliterative patterning of the major variants of types A1, D, D*, and E, with and without anacrusis. Table 3.16. Alliterative patterning of the major variants of type A1 with and without anacrusis a-verse (aa)

a-verse (ax)

b-verse

total

174 (19%) 47 (16%) 221 (18%)

163 (18%) 47 (16%) 210 (17%)

588 (64%) 193 (67%) 781 (64%)

925 (100%) 287 (100%) 1212 (100%)

Psx#Px x . . . Psx#Px (x . . .)Psx#Px

1 (10%) 0 (0%) 1 (9%)

3 (30%) 0 (0%) 3 (27%)

6 (60%) 1 (100%) 7 (64%)

10 (100%) 1 (100%) 11 (100%)

Pxx#Px x . . . Pxx#Px (x . . .)Pxx#Px

14 (19%) 8 (20%) 22 (19%)

17 (23%) 5 (12.5%) 22 (19%)

42 (58%) 27 (67.5%) 69 (61%)

73 (100%) 40 (100%) 113 (100%)

P#xPxa x . . . P#xPxa (x . . .)P#xPxa

28 (16%) 10 (28%) 38 (26%)

10 (9%) 14 (39%) 24 (16%)

72 (65%) 12 (33%) 84 (56%)

110 (100%) 36 (100%) 146 (100%)

P#xPxb x . . . P#xPxb (x . . .)P#xPxb

74 (73%) 29 (81%) 103 (75%)

2 (2%) 1 (3%) 3 (2%)

25 (25%) 6 (17%) 31 (23%)

101 (100%) 36 (100%) 137 (100%)

P#xPxc x . . . P#xPxc (x . . .)P#xPxc

102 (48%) 39 (54%) 141 (50%)

12 (6%) 15 (21%) 27 (10%)

97 (46%) 18 (25%) 115 (41%)

211 (100%) 72 (100%) 283 (100%)

PS#Px x . . . PS#Px (x . . .)PS#Px

6 (50%) 1 (25%) 7 (44%)

1 (8%) 0 (0%) 1 (6%)

5 (42%) 3 (75%) 8 (50%)

12 (100%) 4 (100%) 16 (100%)

10 (50%) 4 (100%) 14 (58%)

4 (20%) 0 (0%) 4 (17%)

6 (30%) 0 (0%) 6 (25%)

20 (100%) 4 (100%) 24 (100%)

62 (10%) 6 (3%) 68 (8%)

209 (34%) 94 (42%) 303 (36%)

619 (100%) 222 (100%) 841 (100%)

Px#Px x . . . Px#Px (x . . .)PxPx

PXx#Px x . . . PXx#Px (x . . .)PXx#Px Px#xPx x . . . Px#xPx (x . . .)Px#xPx

348 (56%) 122 (55%) 470 (56%)

MOSP-03.qxd

5/11/04

2:41 PM

Page 241

3.2. Alliteration

241

a-verse (aa)

a-verse (ax)

b-verse

total

P#xxPx x . . . P#xxPx (x . . .)P#xxPx

101 (55%) 42 (60%) 143 (57%)

27 (15%) 1 (1%) 28 (11%)

55 (30%) 27 (39%) 82 (32%)

183 (100%) 70 (100%) 253 (100%)

P#xxxPx x . . . P#xxxPx (x . . .)P#xxxPx

62 (83%) 66 (92%) 128 (87%)

1 (1%) 2 (3%) 3 (2%)

12 (16%) 4 (6%) 16 (11%)

75 (100%) 72 (100%) 147 (100%)

P#xxxxPx x . . . P#xxxxPx (x . . .)P#xxxxPx

44 (81%) 29 (91%) 73 (85%)

1 (2%) 0 (0%) 1 (1%)

9 (17%) 3 (9%) 12 (14%)

54 (100%) 32 (100%) 86 (100%)

PSx#Px x . . . PSx#Px (x . . .)PSx#Px

9 (64%) 4 (80%) 13 (68%)

2 (14%) 0 (0%) 2 (11%)

3 (21%) 1 (20%) 4 (21%)

14 (100%) 5 (100%) 19 (100%)

PS#xPx x . . . PS#xPx (x . . .)PS#xPx

15 (94%) 0 (0%) 15 (94%)

1 (6%) 0 (0%) 1 (6%)

0 (0%) 0 (0%) 0 (0%)

16 (100%) 0 (0%) 16 (100%)

#x ⫽ prefix #x ⫽ independent word c #x ⫽ prefix or independent word a

b

Table 3.17. Alliterative patterning of the major variants of type D with and without anacrusis a-verse (aa)

a-verse (ax)

b-verse

total

P#PX/xx x . . . P#PX/xx (x . . .)P#PX/xx

19 (19%) 1 (6%) 20 (17%)

25 (25%) 1 (6%) 26 (22%)

56 (56%) 14 (88%) 70 (60%)

100 (100%) 16 (100%) 116 (100%)

P#PS/sx x . . . P#PS/sx (x . . .)P#PS/sx

50 (67%) 9 (82%) 59 (69%)

10 (13%) 1 (9%) 11 (13%)

15 (20%) 1 (9%) 16 (19%)

75 (100%) 11 (100%) 86 (100%)

0 (0%) 0 (0%) 0 (0%)

2 (17%) 0 (0%) 2 (14%)

12 (100%) 2 (100%) 14 (100%)

P#Px . . . S 10 (83%) x . . . P#Px . . . S 2 (100%) (x . . .)P#Px . . . S 12 (86%)

At first glance, the above distribution would appear far from patterned: sometimes anacrusis enhances the incidence of double alliteration, other times it reduces it; similarly, in some cases anacrustic verses occur more frequently in the b-verse than do their non-anacrustic counterparts, in other cases the converse takes place. Yet a closer look will bring to light some patterning involved: the configurations characterised by their minimal preference for the a-verse with double alliteration

MOSP-03.qxd

5/11/04

2:41 PM

242

Page 242

3. Resolution and alliteration

Table 3.18. Alliterative patterning of the major variants of type D* with and without anacrusis a-verse (aa) Px . . . PX/xx x . . . Px . . . PX/xx (x . . .)Px . . . PX/xx

59 (82%) 28 (82%) 87 (82%)

Px . . . PS/sx x . . . Px . . . PS/sx (x . . .)Px . . . PS/sx

215 (98%) 78 (95%) 293 (97%)

Px . . . Px . . . S x . . . Px . . . Px . . . S (x . . .)Px . . . Px . . . S

23 (96%) 7 (100%) 30 (97%)

a-verse (ax)

b-verse

total

4 (6%) 1 (3%) 5 (5%)

9 (13%) 5 (15%) 14 (13%)

72 (100%) 34 (100%) 106 (100%)

1 (less than 1%) 1 (1%) 2 (1%)

3 (1%) 3 (4%) 6 (2%)

219 (100%) 82 (100%) 301 (100%)

0 (0%) 0 (0%) 0 (0%)

1 (4%) 0 (0%) 1 (3%)

24 (100%) 7 (100%) 31 (100%)

Table 3.19. Alliterative patterning of the major variants of type E with and without anacrusis a-verse (aa) PXx(x)#(x)P x . . . PXx#(x)P (x . . .)PXx(x)#(x)P PSx(x)#(x . . .)P x . . . PSx(x)#(x . . .)P (x . . .)PSx(x)#(x . . .)P

a-verse (ax)

b-verse

total

25 (20%) 7 (12%) 32 (18%)

39 (31%) 8 (14%) 47 (26%)

59 (48%) 43 (74%) 102 (56%)

123 (100%) 58 (100%) 181 (100%)

89 (48%) 20 (70%) 109 (47%)

44 (24%) 9 (20%) 53 (23%)

54 (29%) 15 (34%) 69 (30%)

187 (100%) 44 (100%) 231 (100%)

have the frequency of the a-verse with double alliteration reduced and/or the frequency of the b-verse increased, when provided with anacrusis at their head: these are Px#Px, Psx#Px, Pxx#Px (type A1), P#PX/xx (type D), and PXx(x)#(x)P (type E).22 The only exception to this generalisation is the configuration P#xPx, where #x ⫽ prefix. The configuration might possibly have assimilated to the pattern embodied by the closely related configuration in which #x constitutes an independent word. On the contrary, the configurations in which double alliteration prevails over single alliteration (the a-verse and the b-verse counted together) tend to occur with still greater frequency in the a-verse with double alliteration when accompanied with anacrusis, as do the configurations P#xPx (#x ⫽ independent word), P#xxxPx, P#xxxxPx, PSx#Px (type A1), P#PS/sx, P#Px . . . S (type D), and Px . . . Px . . . S (type D*). Finally, the configurations that do not have such marked preference either for the a-verse with double alliteration or the b-verse do not seem to show consistent patterning among themselves when expanded with anacrusis: at issue are the configurations P#xPx (#x ⫽ prefix/independent word), 22

As may be recalled, by virtue of inherent structural ambiguity the configuration x . . . PXx(x)#(x)P is scannable as type E with anacrusis with no less plausibility than as type B1.

MOSP-03.qxd

5/11/04

2:41 PM

Page 243

3.2. Alliteration

243

PS#Px, PXx#Px, Px#xPx, P#xxPx (type A1), and PSx(x)#(x . . .)P (type E). We may accordingly generalise that anacrustic verses tend to reinforce the verse distribution of their non-anacrustic counterparts if the latter show a polarity in patterning on their own.23 The tendency observed above may be ascribed to a generalisation that the Heliand poet brought into being by elaborating on traditional practice. Apart from the increased metrical types (i.e., types A2 and D2), anacrusis was limited to relatively heavy variants of type A1 in the traditional metre. This would mean that anacrusis was integral to the a-verse with double alliteration, since those heavy type A1 variants in themselves showed marked preference for the a-verse with double alliteration. This traditional restriction continued to be respected in the Heliand as well, if in reorganised form: anacrusis remained closely associated with the a-verse with double alliteration to the extent that the base configurations were endowed with such a property. With extension beyond its original domain of use, however, anacrusis could hardly have been retained as a privilege for the a-verse with double alliteration: anacrusis had to co-occur with single alliteration, as well as with the b-verse. Prompted by his outstanding sensitivity to balance and harmony, the Heliand poet then would have reshaped the patterning by harmonising anacrusis with the newly admitted base configurations that showed lesser preference for the a-verse with double alliteration and conversely greater preference for the b-verse. Moreover, confronted with the metrical configurations lacking in marked preference pattern, he abstained from imposing such a reconstitution arbitrarily. In this way, the Heliand was subject to essentially the same distributional constraint on anacrusis as applied to Beowulf. The major difference is that the asymmetrical preference pattern of anacrusis characteristic of Beowulf was harmonised as a symmetrical preference patterning in the Heliand, so that anacrusis reinforced the existing uneven verse distribution pattern of base configurations, not only in favour of the a-verse with double alliteration as in traditional practice, but also in favour of the b-verse. 3.2.1.4. Resolution and alliteration This section addresses the question whether there is a correlation between resolution and alliteration. More specifically, we may ask: do the resolved configurations exhibit a greater (or lesser) preference for the a-verse with double alliteration than their unresolved counterparts? There are two distinct metrical levels to be examined: the level of metrical types and the level of realisation variants. Tables 3.20 through 3.25 compare the alliterative patterning in the Heliand of resolved (differentiated according to the first or the second lift involved) and unresolved configurations of the major metrical types and their representative realisation variants, with special reference to the difference in prominence realised in the drops. Since the secondary-stressed syllable is by nature far less likely to be resolved than the primary-stressed counterpart, it does not figure as an occupant of a lift in the specific 23

Based on a limited sample consisting of type A1 verses, Russom (1998: 147–8) correctly observes that the configuration Px#Px maintains a low frequency of double alliteration when it is accompanied with anacrusis, while more complex type A1 variants show an increased preference for double alliteration when provided with anacrusis. The real picture of this correlation, however, comes to light only when placed in the whole context of distribution of anacrustic and non-anacrusitc verses across distinct metrical types, as shown in the text.

MOSP-03.qxd

5/11/04

2:41 PM

244

Page 244

3. Resolution and alliteration

Table 3.20. Alliterative patterning of the unresolved and resolved variants of type A1 ranking

a-verse (aa)

a-verse (ax)

b-verse

total

/⫻/⫻ px ⫻ / ⫻

132 132

1645 (49%) 414 (57%)

388 (11%) 64 (9%)

1351 (40%) 244 (34%)

3384 (100%) 722 (100%)

/⫻/⫻ / ⫻ px ⫻

132 132

1766 (48%) 293 (63%)

400 (11%) 52 (11%)

1476 (41%) 119 (26%)

3642 (100%) 464 (100%)

Px#Px pxx#Px

331 231

194 (18%) 27 (17%)

188 (18%) 22 (14%)

675 (64%) 106 (68%)

1057 (100%) 155 (100%)

Px#Px Px#pxx

231 231

201 (18%) 20 (27%)

198 (17%) 12 (16%)

738 (65%) 43 (57%)

1137 (100%) 75 (100%)

Pxx#Px pxxx#Px

331 331

22 (20%) 0 (0%)

22 (20%) 0 (0%)

68 (61%) 1 (100%)

111 (100%) 1 (100%)

Pxx#Px Pxx#pxx

231 321

21 (20%) 1 (12.5%)

20 (19%) 2 (25%)

64 (62%) 5 (62.5%)

104 (100%) 8 (100%)

P#xPxa px#xPxa

231 321

29 (26%) 9 (26%)

13 (12%) 11 (31%)

69 (62%) 15 (43%)

111 (100%) 35 (100%)

P#xPxa P#xpxxa

231 321

37 (29%) 1 (6%)

15 (12%) 9 (53%)

77 (60%) 7 (41%)

129 (100%) 17 (100%)

P#xPxb px#xPxb

132 132

44 (70%) 59 (80%)

2 (3%) 1 (1%)

17 (27%) 14 (19%)

63 (100%) 74 (100%)

P#xPxb P#xpxxb

132 132

88 (76%) 15 (71%)

3 (3%) 0 (0%)

25 (22%) 6 (29%)

116 (100%) 21 (100%)

P#xPxc px#xPxc

231 132

73 (42%) 68 (62%)

15 (9%) 12 (11%)

86 (49%) 29 (27%)

174 (100%) 109 (100%)

P#xPxc P#xpxxc

132 132

125 (51%) 16 (42%)

18 (7%) 9 (24%)

102 (41%) 13 (34%)

245 (100%) 38 (100%)

PXx#Px pxXx#Px

132 132

12 (57%) 2 (67%)

4 (19%) 0 (0%)

5 (24%) 1 (23%)

21 (100%) 3 (100%)

PXx#Px PXx#pxx

132 331

14 (61%) 0 (0%)

4 (17%) 0 (0%)

5 (22%) 1 (100%)

23 (100%) 1 (100%)

Px#xPx pxx#xPx

132 132

404 (56%) 66 (58%)

62 (9%) 6 (5%)

261 (36%) 42 (37%)

727 (100%) 114 (100%)

Px#xPx Px#xpxx

132 132

424 (57%) 46 (46%)

49 (7%) 19 (19%)

268 (36%) 35 (35%)

741 (100%) 100 (100%)

P#xxPx px#xxPx

132 132

72 (49%) 71 (68%)

16 (11%) 12 (11%)

60 (41%) 22 (21%)

148 (100%) 105 (100%)

MOSP-03.qxd

5/11/04

2:41 PM

Page 245

3.2. Alliteration ranking

a-verse (aa)

a-verse (ax)

245 b-verse

total

P#xxPx P#xxpxx

132 132

129 (55%) 14 (70%)

26 (11%) 2 (10%)

78 (33%) 4 (20%)

233 (100%) 20 (100%)

Px#xxPx pxx#xxPx

132 133

219 (83%) 22 (69%)

11 (4%) 5 (16%)

34 (13%) 5 (16%)

264 (100%) 32 (100%)

Px#xxPx Px#xxpxx

132 133

210 (81%) 31 (84%)

13 (5%) 3 (8%)

36 (14%) 3 (8%)

259 (100%) 37 (100%)

P#xxxPx px#xxxPx

132 132

70 (89%) 58 (85%)

0 (0%) 3 (4%)

9 (11%) 7 (10%)

79 (100%) 68 (100%)

P#xxxPx P#xxxpxx

132 132

111 (88%) 17 (81%)

3 (2%) 0 (0%)

12 (10%) 4 (19%)

126 (100%) 21 (100%)

P#xxxxPx px#xxxxPx

132 132

38 (83%) 35 (88%)

0 (0%) 1 (3%)

8 (17%) 4 (10%)

46 (100%) 40 (100%)

P#xxxxPx P#xxxxpxx

132 132

45 (82%) 28 (90%)

0 (0%) 1 (3%)

10 (18%) 2 (6%)

55 (100%) 31 (100%)

PSx#Px pxSx#Px

132 132

9 (64%) 4 (80%)

2 (14%) 0 (0%)

3 (21%) 1 (20%)

14 (100%) 5 (100%)

PSx#Px PSx#pxx

132 133

12 (67%) 1 (100%)

2 (11%) 0 (0%)

4 (22%) 0 (0%)

18 (100%) 1 (100%)

PS#xPx pxS#xPx

123 133

12 (92%) 3 (100%)

1 (8%) 0 (0%)

0 (0%) 0 (0%)

13 (100%) 3 (100%)

PS#xPx PS#xpxx

123 –

15 (94%) 0 (0%)

1 (6%) 0 (0%)

0 (0%) 0 (0%)

16 (100%) 0 (0%)

#x ⫽ prefix #x ⫽ independent word c #x ⫽ prefix or independent word a

b

Table 3.21. Alliterative patterning of the unresolved and resolved variants of type B1 ranking

a-verse (aa)

a-verse (ax)

b-verse

total

⫻/⫻/ ⫻ px ⫻ /

321 321

220 (8%) 27 (7%)

455 (17%) 50 (13%)

2082 (76%) 308 (80%)

2757 (100%) 385 (100%)

⫻/⫻/ ⫻ / ⫻ px

321 321

140 (6%) 107 (12%)

372 (16%) 133 (15%)

1753 (77%) 637 (73%)

2265 (100%) 877 (100%)

x . . . Px . . . P 321 x . . . pxx . . . P 321

201 (8%) 25 (7%)

386 (15%) 45 (13%)

2013 (77%) 283 (80%)

2600 (100%) 353 (100%)

x . . . Px . . . P 321 x . . . Px . . . px 321

124 (6%) 102 (13%)

315 (15%) 116 (14%)

1701 (79%) 595 (73%)

2140 (100%) 813 (100%)

MOSP-03.qxd

5/11/04

2:41 PM

Page 246

246

3. Resolution and alliteration

Table 3.22. Alliterative patterning of the unresolved and resolved variants of type C ranking

a-verse (aa)

a-verse (ax)

b-verse

total

⫻//⫻ ⫻ px / ⫻

321 321

41 (4%) 62 (9%)

337 (31%) 128 (18%)

704 (65%) 518 (73%)

1082 (100%) 708 (100%)

⫻//⫻ ⫻ / px ⫻

321 231

101 (6%) 2 (8%)

464 (26%) 1 (4%)

1200 (68%) 22 (88%)

1765 (100%) 25 (100%)

x . . . P#P/px 321 x . . . px#P/px 231

33 (5%) 62 (11%)

62 (10%) 55 (10%)

551 (85%) 424 (78%)

646 (100%) 541 (100%)

x . . . P#P/px x . . . P#pxx

93 (8%) 2 (13%)

117 (10%) 0 (0%)

962 (82%) 13 (87%)

1172 (100%) 15 (100%)

321 231

Table 3.23. Alliterative patterning of the unresolved and resolved variants of type D ranking

a-verse (aa)

a-verse (ax)

b-verse

total

//⫻⫻ px / ⫻ ⫻

132 231

70 (53%) 73 (35%)

20 (15%) 38 (18%)

43 (32%) 100 (47%)

133 (100%) 211 (100%)

//⫻⫻ / px ⫻ ⫻

231 132

109 (38%) 34 (58%)

48 (17%) 10 (17%)

128 (45%) 15 (25%)

285 (100%) 59 (100%)

P#PX/xx px#PX/xx

231 321

10 (37%) 10 (11%)

3 (11%) 23 (26%)

14 (52%) 56 (63%)

27 (100%) 89 (100%)

P#PX/xx P#pxX/xx

321 231

19 (17%) 1 (20%)

26 (23%) 0 (0%)

66 (59%) 4 (80%)

111 (100%) 5 (100%)

P#PS/sx px#PS/sx

132 132

32 (67%) 27 (71%)

7 (15%) 4 (11%)

9 (19%) 7 (18%)

48 (100%) 38 (100%)

P#PS/sx P#pxS/sx

132 132

35 (78%) 24 (59%)

3 (7%) 8 (20%)

7 (16%) 9 (22%)

45 (100%) 41 (100%)

P#Px . . . S px#Px . . . S

132 132

0 (0%) 0 (0%)

1 (12.5%) 1 (17%)

P#Px . . . S P#pxx . . . S

132 –

0 (0%) 0 (0%)

2 (14%) 0 (0%)

7 (87.5%) 5 (83%) 12 (86%) 0 (0%)

8 (100%) 6 (100%) 14 (100%) 0 (0%)

realisation variants in the following discussion, although it is included in the overall counting of each metrical type. Also disregarded from specific discussion are those realisation variants that are attested in small numbers. Beginning with the more concrete level, i.e., the level of realisation variants, we notice that operation of resolution brings about hardly appreciable difference in overall alliterative patterning, with rare exceptions: with or without resolution, we find unchanged the proportion of the a-verse with double alliteration to the one with

MOSP-03.qxd

5/11/04

2:41 PM

Page 247

3.2. Alliteration

247

Table 3.24. Alliterative patterning of the unresolved and resolved variants of type D* ranking

a-verse (aa)

a-verse (ax)

b-verse

total

/⫻/⫻⫻ px ⫻ / ⫻ ⫻

132 132

497 (88%) 103 (94%)

22 (4%) 0 (0%)

48 (8%) 7 (6%)

567 (100%) 110 (100%)

/⫻/⫻⫻ / ⫻ px ⫻ ⫻

132 132

447 (86%) 153 (97%)

22 (4%) 0 (0%)

51 (10%) 4 (3%)

520 (100%) 157 (100%)

Px . . . PX/xx pxx . . . PX/xx

132 132

65 (79%) 26 (93%)

5 (6%) 0 (0%)

12 (15%) 2 (7%)

82 (100%) 28 (100%)

Px . . . PX/xx Px . . . pxX/xx

132 133

87 (82%) 4 (100%)

5 (5%) 0 (0%)

14 (13%) 0 (0%)

106 (100%) 4 (100%)

Px . . . PS/sx pxx . . . PS/sx

132 132

247 (97%) 46 (98%)

2 (1%) 0 (0%)

5 (2%) 1 (2%)

254 (100%) 47 (100%)

Px . . . PS/sx Px . . . pxS/sx

133 132

202 (98%) 91 (97%)

2 (1%) 0 (0%)

3 (1%) 3 (3%)

207 (100%) 94 (100%)

Px . . . Px . . . S 132 pxx . . . Px . . . S 133

29 (97%) 1 (100%)

0 (0%) 0 (0%)

1 (3%) 0 (0%)

30 (100%) 1 (100%)

Px . . . Px . . . S 132 Px . . . pxx . . . S 133

28 (97%) 2 (100%)

0 (0%) 0 (0%)

1 (3%) 0 (0%)

29 (100%) 2 (100%)

Table 3.25. Alliterative patterning of the unresolved and resolved variants of type E ranking

a-verse (aa)

a-verse (ax)

b-verse

total

/\⫻/ px \ ⫻ /

131 132

115 (37%) 58 (48%)

78 (25%) 28 (23%)

114 (37%) 35 (29%)

307 (100%) 121 (100%)

/\⫻/ / \ ⫻ px

231 123

74 (29%) 99 (57%)

67 (26%) 39 (23%)

114 (45%) 35 (20%)

255 (100%) 173 (100%)

PXx . . . P pxXx . . . P

321 313

24 (19%) 2 (29%)

40 (32%) 3 (43%)

61 (49%) 2 (29%)

125 (100%) 7 (100%)

PXx . . . P PXx . . . px

321 312

17 (13%) 9 (19%)

30 (29%) 13 (48%)

51 (58%) 12 (33%)

98 (100%) 34 (100%)

PS/sx . . . P pxS/sx . . . P

132 132

75 (48%) 48 (51%)

35 (22%) 21 (22%)

47 (30%) 25 (27%)

157 (100%) 94 (100%)

PS/sx . . . P PS/sx . . . px

132 133

52 (38%) 71 (63%)

35 (25%) 21 (19%)

51 (37%) 21 (19%)

138 (100%) 113 (100%)

MOSP-03.qxd

5/11/04

248

2:41 PM

Page 248

3. Resolution and alliteration

single alliteration to the b-verse for a given metrical configuration in the Heliand. The proportion in question may be defined here as the relative ranking of the three kinds of verse (the a-verse with double alliteration, the a-verse with single alliteration, and the b-verse) in terms of the percentages that they represent for a given pair of resolved and unresolved configurations, as indicated in the second column of Tables 3.20 through 3.25. However, there are several configurations of substantial occurrence which are subject to reranking in verse distribution: px#xPx (#x ⫽ prefix; type A1), px#xPx (#x ⫽ prefix/independent word; type A1), P#xpxx (#x ⫽ prefix; type A1), x . . . px#P/px and x . . . P#pxx (type C), px#PX/xx (type D), and PXx . . . px (type E). Of these seven deviant configurations, all but two – px#xPx (#x ⫽ prefix) and px#PX/xx – involve a significant increase in the frequency of the a-verse with double alliteration, a point we will return to below. Against the background of the overall similarity between resolved and unresolved configurations in terms of the unchanged proportion of the three kinds of verse represented, we can hardly avoid recognising another tendency in force: that the proportion of the a-verse with double alliteration tends to rise when resolution operates, although there are a number of notable exceptions with varying degrees of deviation as follows: P#xpxx (regardless of the morpholexical status of #x; type A1), Px#xpxx (type A1), pxx#xxPx (type A1), px#PX/xx (type D), P#pxS/sx (type D). The question arising then is whether the above generalisation is also true of traditional practice or whether this constitutes a distinctive feature of the Heliand metre. Tables 3.26 through 3.31 show how resolution is treated in relation to alliterative patterning in Beowulf. Although not perfectly patterned in every detail, the overall situation in Beowulf may lend itself to a generalisation that slightly differs from the one obtained for the Heliand: resolution on either the first or second lift has a stronger tendency to increase frequency of the a-verse with double alliteration. Of particular importance, there are numerous resolved configurations in Beowulf in which resolution (either on the first or the second lift) is observed to be correlated to enhanced incidence of double alliteration at the expense of preserving the relative verse ranking of their unresolved counterparts, such as px#xPx (#x ⫽ prefix/independent word), pxx#xPx, Px#xpxx (type A1), x . . . pxx . . . P, x . . . Px . . . px (type B1), x . . . px#P/px, x . . . P#pxx (type C), and PS/sx . . . px (type E). There are a few exceptions to this generalisation, however: disregarding the configurations attested only in small numbers, the configurations P#xpxx (#x ⫽ independent word, or prefix/ independent word; type A1), Px . . . pxX/xx (subtype D*), pxXx . . . P (type E) exhibit a markedly decreased incidence of the a-verse with double alliteration when resolved. Apart from the configurations px#xPx and pxXx . . . P, the exceptions involve the second lift. In this light, we may refine the generalisation in more precise terms as follows: resolution associates with increased frequency of double alliteration, particularly insofar as the first lift is resolved. Through comparison of the Heliand with Beowulf, we may accordingly be led to conclude that the resolved verses tend to retain the alliterative patterning of their unresolved counterparts to a markedly greater extent in the Heliand metre than in the Beowulf metre, and also that, to a certain extent, resolution is correlated to increase of double alliteration. By contrast, the second generalisation is of central significance in Beowulf, so much so that resolution often incurs increase of double alliteration by bringing about change in verse distribution patterning.

MOSP-03.qxd

5/11/04

2:41 PM

Page 249

3.2. Alliteration

249

Table 3.26. Alliterative patterning of the unresolved and resolved variants of type A1 in Beowulf ranking

a-verse (aa)

a-verse (ax)

b-verse

total

/⫻/⫻ px ⫻ / ⫻

231 231

563 (30%) 113 (38%)

395 (21%) 51 (17%)

942 (50%) 133 (45%)

1900 (100%) 297 (100%)

/⫻/⫻ / ⫻ px ⫻

231 231

616 (30%) 60 (37%)

420 (21%) 26 (16%)

999 (49%) 76 (47%)

2035 (100%) 162 (100%)

Px#Px pxx#Px

321 321

94 (13%) 21 (16%)

225 (30%) 43 (33%)

422 (57%) 66 (51%)

741 (100%) 130 (100%)

Px#Px Px#pxx

321 321

104 (13%) 11 (19%)

249 (31%) 19 (33%)

460 (57%) 28 (48%)

813 (100%) 58 (100%)

Pxx#Px Pxxx#Px

231 –

9 (29%) 0 (0%)

4 (13%) 0 (0%)

18 (58%) 0 (0%)

31 (100%) 0 (0%)

Pxx#Px Pxx#pxx

231 –

9 (29%) 0 (0%)

4 (13%) 0 (0%)

18 (58%) 0 (0%)

31 (100%) 0 (0%)

P#xPxa px#xPxa

231 231

13 (10%) 5 (20%)

10 (8%) 1 (4%)

107 (82%) 19 (76%)

130 (100%) 25 (100%)

P#xPxa P#xpxxa

231 231

14 (11%) 4 (17%)

11 (8%) 0 (0%)

107 (81%) 19 (83%)

132 (100%) 23 (100%)

P#xPxb px#xPxb

132 132

93 (75%) 30 (67%)

2 (2%) 1 (2%)

29 (23%) 14 (31%)

124 (100%) 45 (100%)

P#xPxb P#xpxxb

132 132

116 (73%) 7 (70%)

3 (2%) 0 (0%)

40 (25%) 3 (30%)

159 (100%) 10 (100%)

P#xPxc px#xPxc

231 132

106 (42%) 35 (52%)

12 (5%) 2 (3%)

136 (54%) 33 (45%)

254 (100%) 70 (100%)

P#xPxc P#xpxxc

231 231

130 (45%) 11 (33%)

14 (5%) 0 (0%)

147 (50%) 22 (67%)

291 (100%) 33 (100%)

Px#xPx pxx#xPx

231 132

226 (47%) 28 (49%)

22 (5%) 4 (7%)

230 (48%) 25 (44%)

478 (100%) 57 (100%)

Px#xPx Px#xpxx

231 132

229 (47%) 25 (49%)

23 (5%) 3 (6%)

232 (48%) 23 (45%)

484 (100%) 51 (100%)

P#xxPx px#xxPx

132 132

63 (69%) 22 (76%)

2 (2%) 0 (0%)

26 (29%) 7 (24%)

91 (100%) 29 (100%)

P#xxPx P#xxpxx

132 133

78 (69%) 7 (100%)

2 (2%) 0 (0%)

33 (29%) 0 (0%)

113 (100%) 7 (100%)

MOSP-03.qxd

5/11/04

2:41 PM

Page 250

250

3. Resolution and alliteration ranking

a-verse (aa)

a-verse (ax)

b-verse

total

Px#xxPx pxx#xxPx

132 133

38 (83%) 2 (100%)

1 (2%) 0 (0%)

7 (15%) 0 (0%)

46 (100%) 2 (100%)

Px#xxPx Px#xxpxx

132 133

36 (82%) 4 (100%)

1 (2%) 0 (0%)

7 (16%) 0 (0%)

44 (100%) 4 (100%)

P#xxxPx px#xxxPx

132 133

6 (75%) 1 (100%)

0 (2%) 0 (0%)

2 (25%) 0 (0%)

8 (100%) 1 (100%)

P#xxxPx P#xxxpxx

132 131

6 (86%) 1 (50%)

0 (2%) 0 (0%)

1 (14%) 1 (50%)

7 (100%) 2 (100%)

#x ⫽ prefix #x ⫽ independent word c #x ⫽ prefix or independent word a

b

Table 3.27. Alliterative patterning of the unresolved and resolved variants of type B1 in Beowulf ranking

a-verse (aa)

a-verse (ax)

b-verse

total

⫻/⫻/ ⫻ px ⫻ /

321 231

75 (8%) 15 (16%)

208 (22%) 10 (11%)

673 (70%) 66 (73%)

956 (100%) 91 (100%)

⫻/⫻/ ⫻ / ⫻ px

321 321

72 (7%) 18 (23%)

194 (20%) 24 (31%)

703 (73%) 36 (46%)

969 (100%) 78 (100%)

x . . . Px . . . P 321 x . . . pxx . . . P 231

73 (8%) 15 (19%)

156 (18%) 10 (11%)

639 (74%) 66 (73%)

868 (100%) 91 (100%)

x . . . Px . . . P 321 x . . . Px . . . px 231

70 (8%) 18 (28%)

152 (17%) 14 (22%)

673 (75%) 32 (50%)

895 (100%) 64 (100%)

Table 3.28. Alliterative patterning of the unresolved and resolved variants of type C in Beowulf ranking

a-verse (aa)

a-verse (ax)

b-verse

total

⫻//⫻ ⫻ px / ⫻

321 321

37 (4%) 20 (8%)

396 (45%) 70 (29%)

447 (51%) 151 (63%)

880 (100%) 241 (100%)

⫻//⫻ ⫻ / px ⫻

321 331

55 (5%) 2 (12.5%)

464 (42%) 2 (12.5%)

586 (53%) 12 (75%)

1105 (100%) 16 (100%)

x . . . P#P/px x . . . px#P/px

321 231

21 (6%) 17 (11%)

38 (10%) 5 (3%)

322 (85%) 126 (85%)

381 (100%) 148 (100%)

x . . . P#P/px x . . . P#pxx

321 231

36 (7%) 2 (15%)

43 (8%) 0 (0%)

437 (85%) 11 (85%)

516 (100%) 13 (100%)

MOSP-03.qxd

5/11/04

2:41 PM

Page 251

3.2. Alliteration

251

Table 3.29. Alliterative patterning of the unresolved and resolved variants of types D1, D2a, and D2b in Beowulf ranking

a-verse (aa)

a-verse (ax)

b-verse

total

//⫻⫻ px / ⫻ ⫻

321 321

10 (5%) 12 (8%)

49 (25%) 31 (21%)

138 (70%) 102 (70%)

197 (100%) 145 (100%)

//⫻⫻ / px ⫻ ⫻

321 321

19 (6%) 3 (18%)

75 (23%) 5 (29%)

231 (71%) 9 (53%)

325 (100%) 17 (100%)

P#PX/xx px#PX/xx

321 321

6 (5%) 10 (9%)

20 (16%) 19 (16%)

102 (80%) 87 (75%)

128 (100%) 116 (100%)

P#PX/xx P#pxX/xx

321 331

13 (6%) 3 (25%)

36 (16%) 3 (25%)

183 (79%) 6 (50%)

232 (100%) 12 (100%)

//\x px / \ x

132 132

41 (84%) 32 (78%)

0 (0%) 0 (0%)

8 (16%) 9 (22%)

49 (100%) 41 (100%)

//\x / px \ x

132 132

63 (80%) 10 (91%)

0 (0%) 0 (0%)

16 (20%) 1 (9%)

79 (100%) 11 (100%)

P#PS/sx px#PS/sx

132 132

39 (83%) 31 (78%)

0 (0%) 0 (0%)

8 (17%) 9 (22%)

47 (100%) 40 (100%)

P#PS/sx P#pxS/sx

132 132

61 (79%) 9 (90%)

0 (0%) 0 (0%)

16 (21%) 1 (10%)

77 (100%) 10 (100%)

//⫻\ px / ⫻ \

132 133

39 (95%) 32 (94%)

0 (0%) 1 (3%)

2 (5%) 1 (3%)

41 (100%) 34 (100%)

//⫻\ / px ⫻ \

132 133

70 (95%) 1 (100%)

1 (1%) 0 (0%)

3 (4%) 0 (0%)

74 (100%) 1 (100%)

P#Px . . . S px#Px . . . S

132 132

22 (92%) 15 (94%)

0 (0%) 0 (0%)

2 (8%) 1 (6%)

24 (100%) 16 (100%)

P#Px . . . S P#pxx . . . S

132 –

37 (93%) 0 (0%)

0 (0%) 0 (0%)

3 (7%) 0 (0%)

40 (100%) 0 (0%)

Entering into the more abstract dimension of analysis involving the level of metrical types, we encounter the recurring difference between the Heliand and Beowulf. As at the level of realisation variants, the alliterative patterning in the Heliand at the level of metrical types is also governed primarily by preservation of the distribution patterning that obtains for unresolved configurations, although a few deviant cases as follows are attested: type C (second lift), type D (both lifts), and type E (second lift) are subject to reranking in preference pattern. Furthermore, all but one case of reranking (type D, first lift) manifest an increase in incidence of the a-verse with double alliteration. Noteworthy here are the following two observations. First, as in Beowulf, the second lift shows a stronger tendency for deviation from the generalisation

MOSP-03.qxd

5/11/04

2:41 PM

252

Page 252

3. Resolution and alliteration

Table 3.30. Alliterative patterning of the unresolved and resolved variants of subtypes D*1, D*2a, and D*2b in Beowulf ranking

a-verse (aa)

a-verse (ax)

b-verse

total

/⫻/⫻⫻ px ⫻ / ⫻ ⫻

312 133

6 (12%) 2 (100%)

33 (66%) 0 (0%)

11 (22%) 0 (0%)

50 (100%) 2 (100%)

/⫻/⫻⫻ / ⫻ px ⫻ ⫻

321 313

7 (28%) 1 (4%)

8 (32%) 25 (93%)

10 (40%) 1 (4%)

25 (100%) 27 (100%)

Px . . . PX/xx pxx . . . PX/xx

312 133

6 (13%) 2 (100%)

31 (65%) 0 (0%)

11 (23%) 0 (0%)

48 (100%) 2 (100%)

Px . . . PX/xx Px . . . pxX/xx

231 313

7 (30%) 1 (4%)

6 (26%) 25 (93%)

10 (43%) 1 (4%)

23 (100%) 27 (100%)

/⫻/\⫻ px ⫻ / \ ⫻

132 132

62 (97%) 5 (83%)

0 (0%) 0 (0%)

2 (3%) 1 (17%)

64 (100%) 6 (100%)

/⫻/\⫻ / ⫻ px ⫻ ⫻

132 132

34 (94%) 33 (97%)

0 (0%) 0 (0%)

2 (6%) 1 (3%)

36 (100%) 34 (100%)

Px . . . PS/sx pxx . . . PS/sx

132 133

61 (98%) 5 (100%)

0 (0%) 0 (0%)

1 (4%) 0 (0%)

62 (100%) 5 (100%)

Px . . . PS/sx Px . . . pxS/sx

132 133

34 (97%) 32 (100%)

0 (0%) 0 (0%)

1 (3%) 0 (0%)

35 (100%) 32 (100%)

/⫻/⫻\ px ⫻ / ⫻ \

123 133

35 (97%) 3 (100%)

1 (3%) 0 (0%)

0 (0%) 0 (0%)

36 (100%) 3 (100%)

/⫻/⫻\ / ⫻ px ⫻ \

123 133

36 (97%) 2 (100%)

1 (3%) 0 (0%)

0 (0%) 0 (0%)

37 (100%) 2 (100%)

Px . . . Px . . . S 133 pxx . . . Px . . . S 133

18 (100%) 2 (100%)

0 (0%) 0 (0%)

0 (0%) 0 (0%)

18 (100%) 2 (100%)

Px . . . Px . . . S 133 Px . . . pxx . . . S –

20 (100%) 0 (0%)

0 (0%) 0 (0%)

0 (0%) 0 (0%)

20 (100%) 0 (0%)

than the first. Second, the exceptionality at issue largely stems from the increase of the a-verse with double alliteration, a competing generalisation that is partly at work in the Heliand and in fuller force in Beowulf, as pointed out above and shown further below. Beowulf is in this way distinguished from the Heliand by its more consistent correlation between resolution and double alliteration, which occasionally realises through reranking of the proportion of the three verses that obtains in the absence of resolution, as in type B1 (first lift). The Heliand thus partly reconstituted resolution by giving priority to attaining a high degree of parallelism between resolved and unresolved verses at the cost of the original close association between resolution and double alliteration. This loss of the

MOSP-03.qxd

5/11/04

2:41 PM

Page 253

3.2. Alliteration

253

Table 3.31. Alliterative patterning of the unresolved and resolved variants of type E in Beowulf ranking

a-verse (aa)

a-verse (ax)

b-verse

total

/\⫻/ px \ ⫻ /

231 231

77 (25%) 28 (26%)

21 (7%) 8 (7%)

209 (68%) 71 (66%)

307 (100%) 107 (100%)

/\⫻/ / \ ⫻ px

231 131

89 (23%) 16 (50%)

27 (7%) 2 (0%)

264 (69%) 16 (50%)

380 (100%) 34 (100%)

PXx . . . P pxXx . . . P

231 321

12 (22%) 1 (5%)

3 (5%) 3 (16%)

40 (73%) 15 (79%)

55 (100%) 19 (100%)

PXx . . . P PXx . . . px

231 231

12 (17%) 1 (20%)

6 (9%) 0 (0%)

51 (74%) 4 (80%)

69 (100%) 5 (100%)

PS/sx . . . P pxS/sx . . . P

231 231

52 (27%) 23 (30%)

16 (8%) 5 (6%)

127 (65%) 49 (64%)

195 (100%) 77 (100%)

PS/sx . . . P PS/sx . . . px

231 132

63 (25%) 12 (52%)

21 (8%) 0 (0%)

165 (66%) 11 (48%)

249 (100%) 23 (100%)

traditional clustering of these two metrical features may suggest that resolution was in no small measure deprived of its earlier function of marking increased metrical strength of a given configuration. On the other hand, what there remained as minor exceptions to the similarity of the verse patterning in resolved and unresolved configurations in the Heliand may be regarded as an inherited feature from traditional versecraft, in which this correlation between resolution and double alliteration was in fuller effect. Of particular interest in this connection is the highly deviant configuration P#xpxx (regardless of the morpholexical status of #x; type A1), which is exceptional in the Heliand: it fails to exhibit a higher percentage of double alliteration with resolution than without (Table 3.20). The configuration in question is also exceptional in Beowulf, in which it fails to correlate with an increase in double alliteration in violation of the dominant patterning (Table 3.26). We might be tempted to conjecture that the deviation at issue in the Heliand would have been part of inherited practice. A further outrageous exception found in the Heliand is the configuration px#PX/xx (type D with the first lift resolved; Table 3.23): as with the configuration P#xPx, it undergoes reranking in verse distribution (231 —⬎ 321), and also receives resolution concomitantly with a decrease in double alliteration. An appeal to traditional practice as with the configuration P#xpxx would not be likely to provide a credible account, however: in Beowulf, the same configuration fully conforms to the general pattern, with an increased use of double alliteration as well as with the overall verse distribution preserved (Table 3.29). As far as resolution is concerned, however, the configuration in question in the Heliand reinforces the extensive use of it on the first lift characteristic of traditional practice, as discussed in section 3.1.4 above (Tables 3.6 and 3.7). The radical break from tradition that we witness as regards alliterative patterning in the configuration px#PX/xx in

MOSP-03.qxd

5/11/04

254

2:41 PM

Page 254

3. Resolution and alliteration

the Heliand would accordingly appear least compatible with the reinforcement of the inherited preference for resolution in the same work. Sharply distinguishing the configuration P#PX/xx in the Heliand from its counterpart in Beowulf is a greater frequency with which it manifests double alliteration than in Beowulf, as indicated in Tables 3.23 and 3.29. So high is the frequency that the a-verse with double alliteration comes next to the most frequently occurring b-verse, resulting in the ranking 231 distinct from the one obtained in Beowulf (321), in which the a-verse with double alliteration is minimally represented. By contrast, the resolved variant px#PX/xx exhibits the normal pattern identical to the one in Beowulf. Of greater importance still seems another distinct property of the configuration P#PX/xx in the Heliand: it is the only configuration that occurs with resolution conspicuously more frequently than without: nearly four instances out of every five realise resolution on the first lift (Table 3.6, section 3.1.4). Accordingly, the unmarked variant of the configuration at issue has to be identified as px#PX/xx, rather than P#PX/xx. Since for other configurations the resolved variants are less common and thus demonstrably marked in status, we might be prompted to speculate that the reversal in markedness status would have had to do with the highly exceptional behaviour of the configuration P#PX/xx: more specifically, much as a resolved, hence marked variant is normally attended by a higher frequency of double alliteration, the marked variant P#PX/xx manifests a greater percentage of double alliteration in the exceptional case at hand. Inasmuch as the reason for the alleged reversal in status remains unclarified as yet, we are far removed from anything approaching principled explanation, however. At this point, we need to explore the motivation for the partial reconstitution of resolution in its relation to alliterative patterning in the Heliand, the reorganisation whereby resolution does not coeval with double alliteration as fully as in Beowulf: as may be recalled, in the Beowulf metre, both resolution and double alliteration prototypically served as markers of heightened metrical strength of a lift (Suzuki 1996a). There are two changes that seem to be of vital importance in motivating the reconstitution in question. Both changes affected the alliteration of the Heliand and thereby contributed to a disruption of the close association between double alliteration and resolution that obtained in the traditional metre. First, as pointed out in section 3.2.1.1 above, double alliteration came to figure centrally in the a-verse: about 62 per cent of the a-verses realised double alliteration in the Heliand (Table 3.11). By contrast, less than the half of the a-verses manifested double alliteration in Beowulf. Such an increase in incidence of double alliteration in the Heliand would have nearly undermined its traditional status as a marked opposite of single alliteration: the more frequent occurrence of double alliteration in the a-verse came to stand at odds with such a simplistic categorisation. Second, double alliteration came to be largely dissociated from the notion of enhanced prominence falling on the lift, through obliteration of increased metrical types, i.e., derived metrical types that were nearly coextensive with the a-verse with double alliteration on categorical grounds. Double alliteration accordingly ceased to function as an exponent of added metrical strength that attended on the lift. In its stead, double alliteration was reorganised as a fully graded device that operated with varying degrees of preference depending on an array of syntagmatic properties of verses, even as metrical strength came to reconfigure as a variable weighted primarily

MOSP-03.qxd

5/11/04

2:41 PM

Page 255

3.2. Alliteration

255

along the horizontal dimension of verse length (section 2.1.7). In this way, double alliteration was redefined as a property of an entire verse in the Heliand, rather than a privilege to the lift as a marker of its enhanced prominence. By contrast, resolution continued to be a property of a strong metrical position in general and a lift in particular. Prosodically equated with a long stressed syllable by virtue of foot construction, a resolved disyllable continued to constitute a marked counterpart of it as a realisation variant of the foot. Since the lift is mapped to the foot in verse-making, it may follow that resolution should be characterised as an attribute of the lift. The conceptual link between resolution and lift accordingly was maintained in the Heliand much as in the traditional metre. Thus, while resolution remained closely associated with the lift, its former partner, double alliteration, was dissociated from it in the Heliand. The loss of the traditional common conceptual basis would then have led the poet to see resolution in its own right, rather than regard it as unified with double alliteration. As a consequence, removed from double alliteration, resolution stood alone on its own traditional terms, that is, as an attribute of a lift, while double alliteration came to be placed in a wider context of a whole verse, as a property of a relatively long verse. As it turns out, then, it is double alliteration that underwent reconstitution in itself; as a result of this reconstitution, resolution had to be recharacterised as independent from double alliteration. In this way emerged the reorganisation of the relation between resolved and unresolved configurations in the Heliand. Furthermore, since a resolved disyllable remained equivalent to a long stressed syllable in prosodic terms, there would have been no special motivation for encoding with some metrical distinction the substitution of a long stressed syllable with a resolved disyllable by resolution. Resolution thus came to exert least demonstrable change in the alliterative patterning of the unresolved configurations. 3.2.2. The lexical and syntactic basis of alliteration In addition to the obligatory alliteration on the first lift of a verse (the principle of left dominance or alliteration by position; section 3.2.1.1), as well as the configurational properties of the verse involved (sections 3.2.1.2), the lexical categories of words contained in it play a vital role in determining its alliterative pattern. More specifically, lexical categories largely determine which word or words may serve as a carrier or carriers of alliteration. According to such varying qualification for realising alliteration, the major lexical categories fall into three classes (Rieger 1876: 18–32; Sievers 1893: 42–6; Lehmann 1953: 32–3; Heusler 1956: §§137–52). Class 1 consists of substantives (nouns, adjectives, substantive and adjectival forms of verbs); class 2, finite verbs and adverbs; and class 3, function words (pronouns, pronominal adjectives, prepositions, conjunctions). Class 1 words (substantives) are fully qualified to alliterate and usually participate in alliteration; class 3 words (function words) are least likely to realise alliteration, because they normally function as a drop, as shown in the following paragraph; between these two extremes are located class 2 words (finite verbs and some adverbs), which may manifest alliteration according to context. In this way, alliterative patterning is determined in part by membership of lexical categories (alliteration by nature). Such a lexical-based ranking in terms of the varying qualifications for alliteration conditions among other things the way the second lift of the a-verse is affected by alliteration: it has to alliterate if

MOSP-03.qxd

5/11/04

256

2:41 PM

Page 256

3. Resolution and alliteration

it is occupied by a class 1 word and if the first lift is realised by a word of a lower class in terms of the ranking in question. By contrast, when both lifts are realised by class 1 words, the second lift may alliterate only optionally (Suzuki 1996a: 282; cf. Rieger 1876: 19; Heusler 1956: §§138–9). The lexical categorisation underlying alliterative patterning is identical with the one responsible for determining the status of metrical positions – the lift and the drop – for constituent words in a verse. Given such a duplicate function performed by the lexical categorisation in question, and more importantly in view of the status of lift being a prerequisite for alliteration, it may be appropriate at this point to remark on the mechanism by which a given word is assigned a lift (and by implication the mechanism of drop formation as well). Two different modes of lift formation are at work, one complementing with the other: (i) lift formation by nature; and (ii) lift formation by position. Lift formation by nature stipulates that class 1 words are by nature most likely to be chosen as lifts wherever they appear in a verse, insofar as not more than two such words are contained in the verse. This inherent full qualification of class 1 words for forming a lift may be held responsible for their nearly consistent participation in alliteration even when they serve as the second lift, provided that the first one is realised by a word of another class, as observed in the preceding paragraph. Conversely, class 3 words are largely excluded from this process of lift formation by nature: they are usually identified as drops accordingly, unless they are exceptionally assigned emphatic stress on pragmatic grounds or perceived as (more or less coincidentally) alliterative primarily due to their phonological identity with the following lift (cf. Sievers 1893: §28). Because a single drop may be associated with a sequence of unstressed syllables, the number of class 3 words occurring in a verse is immaterial. Standing between these two extremes are class 2 words, whose qualification for lift formation by nature is partly context-sensitive. Indeed, they are by nature capable of serving as lifts, yet their actual status largely depends on a syntagmatic condition, namely a relative ranking of the lexical classes involved in the verse. More specifically, a class 2 word is likely to be chosen as a lift by nature when the remaining words contained in the verse are all class 3; on the other hand, in the presence of a class 1 or 2 word in the same verse, it may, but need not, be chosen as a lift. In this way, class 2 is inherently ambivalent, standing in the interstices between the two clearly defined classes (classes 1 and 3) and circumscribed by fuzzy boundaries in itself. This inherent ambiguity then gives rise to a number of ambivalent verses that are susceptible to alternative scansions, as shown below. Lift formation by position, as implied by the name, is conditioned by position in the verse. It assigns the status of lift to those words that are not chosen as lifts on the basis of their inherent properties (lift formation by nature), provided that they are located in the third or fourth metrical positions. Accordingly, mainly class 3 and, to a lesser extent, class 2 are susceptible to this second mode of lift formation, all members of class 1 falling under the scope of the first mode. Both lift formations, applying in conjunction by nature and by position as specified above, single out at most two words per verse as lifted, that is, strong in metrical terms. Subsequently, the most prominent feet contained in them are identified as the lifts of the verse, the remaining syllables delegated to the role of drops. The lift formations in question thus scan the following two-word concatenations as

MOSP-03.qxd

5/11/04

2:41 PM

Page 257

3.2. Alliteration

257

Table 3.32. Scansion of two-word concatenations as metrical configurations two-word concatenations

metrical configurations

class 1 ⫹ class 1 class 1 ⫹ class 2 class 1 ⫹ class 3 class 2 ⫹ class 1 class 2 ⫹ class 2 class 2 ⫹ class 3 class 3 ⫹ class 1 class 3 ⫹ class 2 class 3 ⫹ class 3

lift ⫹ lift lift ⫹ lift lift ⫹ lift lift ⫹ lift; drop ⫹ lift lift ⫹ lift; drop ⫹ lift lift ⫹ lift drop ⫹ lift drop ⫹ lift drop ⫹ lift

metrical configurations consisting of lifts and drops, as shown prototypically in Table 3.32. Thus, we find two concatenations which are ambiguous in their metrical value: (i) class 2 word ⫹ class 1 word; (ii) class 2 word ⫹ class 2 word. As noted above, a class 2 word is by nature subject to two interpretations when it may be scanned as occupying the first or second positions of a verse (that is, the two positions that are immune to lift formation by position) and when it is followed by a word of the equal or more prominent classes (i.e., classes 1 and 2). We accordingly encounter a number of ambiguous a-verses that are amenable to alternative scansions depending on how we characterise such class 2 words, as an alliterative lift or as a drop: if scanned as alliterative, we obtain verses with double alliteration; otherwise, we get ones with single alliteration. These ambivalent verses fall into three major groups and are open to the following alternative scansions: (i) types A1 or A3; (ii) types B1 or D* (subtype D*2b); (iii) types C or D* (subtype D*2a). These three groups of verses are exemplified and listed below, and the exact metrical status of the underlined words crucially determines scansion:24 (33) types A1 or A3; thirty-seven examples 847a Habda im sô bihalden 1080a lêt ina thô lêdean 1099a endi all sulic ôdes 1215a af sulicun suhtiun 1417a endi ôc sulicu suuîðor 1469a Êr scalt thu thi simbla gesônien 1481a that he beginna thero girnean 24

In addition, there are a few verses that may be scanned as types B1 or D (subtype D2b), as follows: 2813a fôr folcun tô 3117a stigun stên endi berg 3408a that sie eft ubil ettha gôd 5380a sô manag mislîc thing

MOSP-03.qxd

5/11/04

258

2:41 PM

Page 258

3. Resolution and alliteration 1516a 1525a 1610a 2012a 2049a 2087a 2101a 2561a 2645a 2879a 3031a 3140a 3264a 3321a 3429a 3435a 3895a 3929a 4158a 4220a 4426a 4861a 4936a 4967a 5156a 5292a 5528a 5548a 5684a 5809a

so uuirðid is simbla uuirsa sô cumid it al fan ubile Ne lât ûs farlêdean uuârun thar an uunneun sô ni mahte he bemîðan that he hîr bihalde quað that he thar quâmi ne uuelleo ik, that gi it uuiodon, quathie sô ina hêr gihaldid that he uuâri uuirðig habde iru giholpen that man thi hêr an thesaru hôhe biûtan the êno that he thurh mîna minnea that man them mannon allon them ando Habde iru thô giholpen habdun it im te hosca Ni uuas it thoh is uuillean ac uuârun im sô uurêða than ni habde ik thar ênige helpe uuâri it nu thîn uuillio, quâðun sie bethiu ni mahtun sie is bemîðan giuuêt ina thô uuarmien umbi sulica sundea Sia hietun im thuo te hoske that iu hier bihlîdan huilic iro scoldi hebbian that that uuâri te uuâren uuas im is giuuâdi

(34) types B1 or D* (subtype D*2b); twenty-four examples 340a obar alla thesa irminthiod 717a that uuârun thea uuîson man 801a eft an ôðrun daga 1114a suuîðo sêragmôd 1324a aftar te êuuandage 1536a Dôe alloro erlo gehuilic 1752a huuô alloro erlo gehuilic 2051a alloro erlo gehuilic 2243a suang gisuerc an gimang 2279a drêf thea diublas thanan 2284a Sô deda the drohtines sunu 2444a huô man thea mârien scal 2553a Thuo sprak eft thie aðales man 3144a sô tilêt thiu luft an tuê 3425a that man thero manno gihuem

MOSP-03.qxd

5/11/04

2:41 PM

Page 259

3.2. Alliteration 3432a 3482a 4165a 4319a 4325a 4836a 5458a 5499a 5973a

259

that man im mêra lôn farit im forð mid thiu alla irminthiod ac uuirðid uuîg sô maneg uuirðid uuôl sô mikil behuî lêdis thu mi sô these liudi tô allaro erlo gihuem hietun thuo hôbidband thar hôf he is hendi up

(35) types C or D* (subtype D*2a); sixteen examples 38a all so hie it fan them anginne 55a habda them heriscipie 282a aftar them ârundie 507a Sie môsta aftar ira magaðhêdi 564a ûse ârundi 594a huan êr sie gisâuuin ôstana 1105a alles thes ôduuelon 1240a aftar iro êndagon 1444a al sulikes urdêlies 2543a uuolda im thar sô uunsames 2768a endi allumu themu erlskepie 2994a ênig anduuordi 3363a allaro is armôdio 3883a Thô gifragn ik that sie frâgode 4382a alle thea engilos 4973a ni bist thu thesoro burgliudio, quâðun sie The particular scansions provided to the above verses depend decisively on the characterisation of the class 2, or sometimes class 3, words involved, such as pronominal adjectives (e.g., all), adverbs (e.g., eft), finite verbs (e.g., hebbian, wesan), man in pronominal use, and prepositions (e.g., aftar). In this respect, Lehmann (1953) and Hofmann (1991) largely diverge in their metrical analysis: Lehmann scans most of these crucial words as alliterative, whereas Hofmann treats them as drops and hence non-alliterative. Yet both metrists would appear less than consistent in their practice. For example, Lehmann treats uuirðid in verse 4325a uuirðid uuôl sô mikil as nonalliterative (given under (34) above), but scans as alliterative the same lexical item contained in verse 1516a so uuirðid is simbla uuirsa (given under (33) above). Hofmann, on the other hand, characterises alla in verse 4165a alla irminthiod as alliterative, yet reads the same word in verse 340a obar alla thesa irminthiod as nonalliterative (given under (34) above); further, Hofmann (1991: 88–90) postulates alliteration on finite verbs when they carry specific lexical meanings, as opposed to those with very general contents such as uuesen and hebbian as given above. Moreover, Lehmann often, and Hofmann sporadically, go so far as to scan class 3 words as alliterative: while treating aftar in verse 1240a aftar iro êndagon (given under (35) above) as non-alliterative, both Lehmann and Hofmann scan the same word in verse 1324a aftar te êuuandage (given under (34) above) as alliterative; and

MOSP-03.qxd

5/11/04

260

2:41 PM

Page 260

3. Resolution and alliteration

Lehmann regards ûse in verse 564a ûse ârundi (given under (35) above) as participating in alliteration. Highly arbitrary as the specific decisions would appear, we have to recognise that such apparent lack of consistency and resultant variation in scansion ultimately resides in the inherent ambivalence of the lexical class at issue (i.e., class 2) and of the prototype-based categorisation at large. Accordingly, rather than taken as a flaw of particular analytical frameworks, this feature should better be acknowledged as one of the structural indeterminacies of the underlying metrical system (cf. Suzuki 1996a: 7–10). In this light, I simply assume primarily for the sake of practical consistency alliteration on the above words and scan the relevant verses as types A1, subtypes D*2b, or D*2a, rather than types A3, B1, or C. The radical increase in double alliteration in general (section 3.2.1.1), and the extensive occurrence of type D* (section 2.8), in the Heliand, may lend some plausibility to scanning the words at issue as alliterative. Since these ambivalent verses are not many in number, however, the inevitably arbitrary scansions would be most unlikely to bias seriously our overall understanding of structural properties of the metrical types involved. Returning to alliterative patterning, we examine in what follows the applicability to the Heliand of the above rules for alliteration that are in full force in the traditional versecraft, as represented in Beowulf. Let us begin with the alliterative patterning of verses which are not heavy, that is, verses that do not contain more than two class 1 words. The tripartite division of word categories (alliteration by nature) and the principle of left dominance (alliteration by position) brought together should only allow for the following patterns of alliteration for the a-verse and the b-verse, as illustrated in Tables 3.33 and 3.34, respectively (cf. Heusler 1956: §§139–44, 150). The differential treatment of the a-verse and the b-verse stems from the categorical requirements imposed exclusively on the b-verse that it must invariably contain two lifts and that the first one always alliterates solely on its own (section 3.2.1.1). Accordingly, the b-verse constantly realises the pattern ax regardless of the exact lexical composition of individual verses. The principle of left dominance thus has a maximally transparent effect in the b-verse. Since alliteration is a privilege of the lift, class 3 words are generally excluded from realising it, unless they are chosen as lifts by position, as prescribed above, or occasionally perceived as alliterative presumably due to their phonological coincidence with the following lift. Accordingly, with rare exceptions the word concatenations beginning with a class 3 word, that is, class 3 ⫹ class 1 and class 3 ⫹ class 2, are most likely to manifest single alliteration on the only lift available in the verse, namely the alliterative pattern xa (the concatenation class 3 ⫹ class 3 is subject to a further constraint, as remarked below). Moreover, these three concatenations beginning with a class 3 word are disallowed as legitimate metrical configurations from the b-verse (as represented by the shaded background in Table 3.34) by virtue of the unique pattern ax accessible to it. And even when class 3 words are selected as lifts by position, they are usually non-alliterative. Thus, the concatenations of class 1 ⫹ class 3 and of class 2 ⫹ class 3 nearly always manifest the alliterative pattern ax, rather than aa. By the same token, the concatenation of class 3 ⫹ class 3 seems to be at variance with the pattern xa. These extra limitations may be ascribed to the inherent minimal prominence of class 3 words: in addition to making the words

MOSP-03.qxd

5/11/04

2:41 PM

Page 261

3.2. Alliteration

261

Table 3.33. Alliterative patterns of normal (non-heavy) a-verses lexical category

1

2

3

1 2 3

1 ⫹ 1: aa, ax, *xa 2 ⫹ 1: aa, *ax, xa 3 ⫹ 1: (*)aa, *ax, xa

1 ⫹ 2: aa, ax, *xa 2 ⫹ 2: aa, ax, xa 3 ⫹ 2: *aa, *ax, xa

1 ⫹ 3: (*)aa, ax, *xa 2 ⫹ 3: (*)aa, ax, *xa 3 ⫹ 3: *aa, (*)ax, *xa

Key: * ⫽ non-existent; (*) ⫽ rare Examples: 1 ⫹ 1 (aa): 2934a uuatares uuîti 1 ⫹ 1 (ax): 2639a hellie fiures 1 ⫹ 2 (aa): 2272a handun hêlde 1 ⫹ 2 (ax): 3346a hungar tholode 1 ⫹ 3 (aa; rare): 1239a uuurðun is thegnos te thiu 1 ⫹ 3 (ax): 2198a barnes thînes 2 ⫹ 1 (aa): 2292a fôr im te them friundun 2 ⫹ 1 (xa): 2997a bâdun iro hêrron 2 ⫹ 2 (aa): 4820a cussiu ine endi quaddiu 2 ⫹ 2 (ax): 3960a deda thar sô hie giuuonoda 2 ⫹ 2 (xa): 3974a bâdun that tharod quâmi 2 ⫹ 3 (aa; rare): 5498a dedun im eft ôðer an 2 ⫹ 3 (ax): 2502a Than farlêdead ina 3 ⫹ 1 (aa; rare): 1324a aftar te êuuandage 3 ⫹ 1 (xa): 1090a mid thînun fôtun 3 ⫹ 2 (xa): 894a that he sô gilêstea 3 ⫹ 3 (ax; rare): 3073a that thu môst aftar mi

Table 3.34. Alliterative patterns of normal (non-heavy) b-verses lexical category

1

2

3

1 2 3

1 ⫹ 1: *aa, ax, *xa 2 ⫹ 1: *aa, ax, *xa 3 ⫹ 1: *ax, *aa, *xa

1 ⫹ 2: *aa, ax, *xa 2 ⫹ 2: *aa, ax, *xa 3 ⫹ 2: *ax, *aa, *xa

1 ⫹ 3: *aa, ax, *xa 2 ⫹ 3: *aa, ax, *xa 3 ⫹ 3: *ax, *aa, *xa

Key: * ⫽ non-existent Examples: 1 ⫹ 1 (ax): 1601b himila rîkea 1 ⫹ 2 (ax): 2270b helpa sôhte 1 ⫹ 3 (ax): 1935b than gi an themu hûse mid im 2 ⫹ 1 (ax): 2019b sagda im mid uuordun 2 ⫹ 2 (ax): 3278b sô thu mi lêris nu 2 ⫹ 3 (ax): 3027b sô thu bâdi te mi

concerned scarcely capable to serve as lifts on their own (i.e., lift formation by nature), it largely disqualifies them from realising alliteration, in sharp contrast to the intrinsic maximal prominence of class 1 words, which requires them to alliterate even when used as the second lift under the specific conditions noted above. There are a small number of apparent exceptions to the above constraints on lift formation and alliterative patterning in the Heliand. The following verses are seemingly contradictory to the rule of lift formation by nature, and also to the rule of alliteration by nature, by implication (cf. Rieger 1876: 29–32).

MOSP-03.qxd

5/11/04

262

2:41 PM

Page 262

3. Resolution and alliteration

(36) 4600a an them dagun (class 3 ⫹ class 1) 2407b an themu dage (class 3 ⫹ class 1) Although class 3 words are, on their inherent grounds, scarcely capable of serving as lifts and hence of carrying alliteration, them and themu constitute the first lifts in the above verses. It may be reminded here that class 3 words are chosen as lifts primarily when they are found in the last two positions of the verse (lift formation by position), the contexts that are obviously out of a question here. Furthermore, with verse 4600a the second word them participates in alliteration on its own at the expense of the following more prominent word of class 1, dagun. On the other hand, the alliterative pattern ax of verse 2407b may be understood as a natural consequence of the unique canonical pattern available to the b-verse (cf. Rieger 1876: 30). On closer consideration, however, the apparently aberrant alliteration of verses 4600a and 2407b proves to be well-motivated: the class 3 words involved bear emphatic stress, a pragmatically based prominence. Put another way, these class 3 words are treated as indistinguishable from class 1 words by virtue of their heightened importance in the sentences concerned. In this light, the alliteration on them at the cost of dagun in verse 4600a may be explained as a consequence of the pragmatically motivated promotion of them as equivalent to a class 1 word: the superficial concatenation involved, class 3 ⫹ class 1, should rather be rescanned as consisting of class 1 ⫹ class 1. Given such a pragmatically motivated rescansion, them may freely participate on its own in disregard of the lexical-based alliterative patterns xa and aa that would otherwise obtain (see Table 3.33). The concatenation of class 2 and class 1 words in the a-verse exceptionally shows the alliterative pattern ax, rather than aa or xa, in defiance of the lexical hierarchy involved. We count four such exceptions to the generalisation (cf. Stübiger 1953: 6; Heusler 1956: §142): (37) 952a 1085a 4705a 4867a

quâmun thar te Iohannese scrîd thi te erðu hinan ne druobie iuuua herta uuell imu innan hugi

We would have expected the class 1 words underlined above to be alliterative either on its own or in conjunction with the preceding class 2 words. The exceptionality of the above four verses may be born out by the more frequent occurrence of comparable b-verses, thirty-eight in all, according to Stübiger’s (1953: 1–4) counting. The higher incidence in the b-verse of the configuration with a finite verb and a noun serving as lifts in this order is attributed to the canonical alliterative pattern ax in the b-verse, which overrides the alliteration rule based on lexical categories (alliteration by nature), as remarked above. The Heliand metre, however, is not unique in allowing for such exceptions. Comparable examples are found in Beowulf as well (cf. Lehmann 1953: 33): (38) Beo 758a Gemunde þa se goda Beo 1537a Gefeng þa be eaxle

MOSP-03.qxd

5/11/04

2:41 PM

Page 263

3.2. Alliteration

263

The principle of left dominance (alliteration by position) seems to admit of the following exceptions (cf. Heusler 1956: §139): (39) 19a Lucas endi Iohannes 235a Thô nam he thia bôk an hand 5552a Iesus fan Nazarethburh In the above examples, Lucas, bôk ‘book’, and Iesus, the first class 1 words contained in the verses concerned, are skipped over, and alliteration falls on the following class 1 words, Iohannes, hand ‘hand’, and Nazarethburh, respectively. In this connection, we may be reminded of a similar exception in Beowulf (Suzuki 1996a: 287–8): (40) Beo 316a Mæ¯l is me to feran Yet the exceptional behaviour of mæ¯l ‘time’ seems explainable to some extent on syntactic and pragmatic grounds: mæ¯l, standing in sentence-initial position, serves as predicate; used as a predicate nominal followed by a sequence of function words, the word may have failed to assert its full inherent noun categoriality; as a consequence, it would have been treated on a par with the other class 3 words, therewith found incapable of bearing alliteration on its own (Suzuki 1996a: 288). The Heliand examples, however, seem hardly amenable to a similar account. Furthermore, Lucas, bôk, and Iesus are all highly unlikely to have been recategorized on lexical grounds as class 3 words as was man ‘man’ discussed below. We may therefore have to admit that these three verses are true exceptions to the principle of left dominance.25 Having confirmed the strict conformity of the Heliand to traditional practice in determining alliterative pattern for non-heavy verses, we now turn to examine a subset of heavy verses, that is, the group of verses that contain three class 1 words. In traditional versecraft, heavy verses are expected to show double alliteration in manners conforming to the principle of left dominance: given three words of class 1 in a verse, the first two instances normally participate in alliteration to the exclusion of the last one (Suzuki 1996a: 49).26 The implementation of double rather than single alliteration on heavy verses is derived from the increased metrical strength, in comparison with the ordinary two-word verses, due to the presence of a third class 1 word (Suzuki 1996a: 383). Thus, the heavy verses should manifest the alliterative pattern aax for the a-verse as illustrated in (41) below, and axx for the b-verse, although the b-verse tends to avoid heavy verses by virtue of their inherently marked metrical status that is largely at variance with the unmarked status of the b-verse. 25

26

In view of such intractability, we may naturally be prompted to endorse Martin’s (1896: 126) emendation Thô nam he an hand hia bôk. Alternatively, one might follow Hofmann’s (1991: 166–7) account. Much the same may apply to hypermetric verses, particularly hypermetric a-verses. On this special category of overlong verses, see Chapter 4.

MOSP-03.qxd

5/11/04

264

2:41 PM

Page 264

3. Resolution and alliteration

(41) List of heavy verses with expected alliterative patterns aax in the a-verse, and their syntactic constituency27 9a frummian, firiho barn [A[BC]] (also 2594a) 15a hêlag himilisc uuord [A[BC]] 261a idis enstio fol [A[BC]] 372a mahtig an manno lioht [A[BC]] 412a fagar folc godes (cf. 1440a sâlig folc godes (axx); 2870a grôt craft godes (axa)) [A[B]C] 440a hêlag himilisc barn [A[BC]] 557a erlos fon ôðrun thiodun [A[BC]] 558a cunnies fon cnôsle gôdun [A[BC]] 559a êri fon ôðrun thiodun [A[BC]] 598a cuman thurh craft godes [A[BC]] 600a hêdro fon himiles tunglun [A[BC]] 626a liof landes uuard [A[BC]] (also 3759a) 661a berht bôcan godes [A[B]C] 671a thea uueros uualdand Krist [A[BC]] 763a môdag manno drôm [A[BC]] 776a Iosep godes têcan [A[BC]] 855a uuirkean uuilleon godes [A[BC]] 961a diurlîc drohtines sunu [A[BC]] (also 1005a) 984a liof liudio uuard [A[BC]] 990a hlûd fon them hôhon radura [A[BC]] 991a Crista, allaro cuningo bezton [A[BC]] 992a selbo fon sînun rîkea [A[BC]] 993a bezt allaro giboranaro manno [A[BC]] 1059a hêleg himiles uuard [A[BC]] 1107a hôrean the hêlago Crist [A[BC]] 1163a faran folc manag [A[BC]] 1271a faran fan them folke âðrum [A[BC]] 1282a gumon umbi thana godes sunu [A[BC]] (also 1384a, 1581a) 1348a gibidig grimmora thing [A[BC]] 1412a brêdean, that gibod godes [A[BC]] 1557a gerno thurh godes thanc [A[BC]] 1810a ne uuâg ne uuatares strôm [A[BC]] 1865a gêstos an godes uuang [A[BC]] 1921a gangan an that godes rîki [A[BC]] 2122a biddien, barn godes [A[BC]] 2145a hard helleo gethuing [A[BC]] (also 5169a) 2176a mid thiu brahtmu that barn godes [A[BC]] 2251a thie guodo godes suno [A[BC]] (also 4011a) 2298a brengean for that barn godes [A[BC]] 2348a hôrien is hêlag uuord [A[BC]] 2355a Crist thurh is craft mikil [A[BC]] 2371a be biliðiun that barn godes [A[BC]] 27

The issue of syntactic constituency will be discussed at length further below.

MOSP-03.qxd

5/11/04

2:41 PM

Page 265

3.2. Alliteration 2556a 2576a 2601a 2821a 2832a 2847a 2908a 2917a 2937a 2991a 3005a 3046a 3059a 3065a 3070a 3082a 3172a 3298a 3450a 3563a 3644a 3653a 3994a 4089a 4115a 4256a 4275a 4379a 4495a 4633a 4909a 4916a 4992a 5094a 5140a 5231a 5395a 5511a 5551a 5561a 5655a 5722a 5743a 5782a 5802a 5869a 5894a

fîond fêcni crûd [A[BC]] mâri mahtig Crist [A[BC]] (also 2581a, 3099a) hôh himiles lioht [A[BC]] (also 3669a) gumon te themu godes barne [A[BC]] Philippus frôd gumo [A[BC]] the gôdo godes sunu [A[BC]] (also 5089a) skêðan skîr uuater [A[BC]] sebo sorgono ful [A[BC]] (also 5966a) drokno obar diap uuater [A[BC]] selbo sunu Dauides [A[BC]] drîbad im dernean hugi [A[BC]] diulîc drohtines bodo [A[BC]] (also 5806a) Crist cuning êuuig [A[BC]] (also 5087a) fader allaro firiho barno [A[BC]] (also 3241a) hêlag hûs godes [A[B]C] grôni godes uuang [A[BC]] mahtig is mâgo hêm [A[BC]] unôði ôdagumu manne [A[BC]] gumon an godes uuang [A[BC]] drohtin Dauides sunu [A[BC]] Crist allaro cuningo best [A[BC]] open êuuig lioht [A[BC]] diurlîc drohtines thegan [A[BC]] the mikilon maht godes [A[B]C] thiu mikile maht godes [A[B]C] garu gôdo mêst [A[BC]] thie gumon umbi that godes hûs [A[BC]] mâri mannes sunu [A[BC]] uuâr uualdand Krist [A[B]C] uualdand uuîn endi brôd [A[BC]] diurlîc dages lioht [A[BC]] môdag manno folc [A[BC]] selbo te Simon Petruse [A[BC]] mârean mannes sunu [A[BC]] an themu dage derbies uuiht [A[BC]] findan fêknea uuord [A[BC]] mâri maht godes [A[B]C] sueltan sundiono lôs [A[BC]] Cristes an crûce scrîban [A[BC]] tuêna fartalda man [A[BC]] mînon gêst an godes uuillion [A[BC]] hêlag himilo rîkies [A[BC]] thes gumen grimman dôð [A[BC]] suigli sunnun lioht [A[BC]] lîf langerun huîl [A[BC]] cûðian thia craft godes [A[BC]] mârian thia maht godes [A[BC]]

265

MOSP-03.qxd

5/11/04

266

2:41 PM

Page 266

3. Resolution and alliteration 5907a sehan seldlîc thing [A[BC]] 5962a gangandi godes suno [A[BC]] 5970a mid them grôtun godes craft [A[BC]]

The Heliand allows for a greater range of alliterative patterns for the heavy a-verses than is possible in traditional practice, however: it permits not only the pattern aax as expected, but axx and axa as well. (42) Examples of heavy verses with apparently deviant alliterative patterns axx (see further a full list (43) below): 798a mahtig barn godes 1774a uuîd strâta endi brêd axa (see further a full list (44) below): 2171a godes uuilleon gumun 2870a grôt craft godes xaa: 945a uuesan euua hugi hrômag28 5419a huô thiu thiod habda duomos adêlid xax: 5623a thuo man thena godes suno29 While the patterns xaa and xax are rare in the extreme if not totally absent, axx and axa are frequently attested, as shown below. The frequent attestation of the alliterative patterns axx and axa in the Heliand may be understood as a weakening of the canonical alliterative pattern aax that fully obtains in Beowulf. Stated in more precise terms, the pattern axx stands in violation of the rule that requires double alliteration for heavy verses. This rule is exceptionless in Beowulf. Verse Beo 1422a Flod blode weol would appear to be the only example of a heavy a-verse with the alliterative pattern axx. The last word (weol), however, is a finite verb. Thus, Beowulf offers not a single instance of the verse that consists of three class 1 words and manifests the alliterative pattern axx.30 In view of this observation, then, we will be led to conclude that double alliteration for heavy verses is fully obligatory for Beowulf, whereas the binding force of the rule is considerably curtailed for the Heliand. (43) List of heavy verses with the alliterative pattern axx (in violation of the double alliteration rule for heavy verses), and their syntactic constituency 7a hêlag uuord godas [A[B]C] 518a that hêlage barn godes [A[B]C] 700a godes engil cumen (also 1117a) [[AB]C] 794a godes êgan barn (also 838a, 1010a, 1335a, 3085a) [A[BC]] 28

29

30

The authenticity of this verse is in doubt, as it is subjected to emendation. For example, Hofmann (1991: 52) reads verses 944b–945a as than eu lango scal uuesan / euua hugi hrômag. As will be discussed at length below, this verse should be a pseudo example of the alliterative pattern xax, since what appears to be a class 1 word, man, is pronominal and accordingly serves as part of a drop. For a comprehensive list of heavy verses in Beowulf and other Old English poems, see Hutcheson (1995: 257–69).

MOSP-03.qxd

5/11/04

2:41 PM

Page 267

3.2. Alliteration 798a 1180a 1440a 1774a 2133a 2218a 2520a 2975a 3147a 3981a 4062a 4799a

267

mahtig barn godes (also 2024a, 2038a, 2325a) [A[B]C] sâlig barn godes (also 5509a) [A[B]C] sâlig folc godes [A[B]C] uuîd strâta endi brêd (cf. 1707a hard trio endi hebig (axa)) [[AB]C] hêlag folc godes [A[B]C] dages lioht sehan [[AB]C] that hêlaga gibod godes [A[B]C] ênag barn godes [A[B]C] hêlag stemne godes [A[B]C] tuâ naht endi dagas [[AB]C] Krist godes sunu (also 5584a) [A[BC]] thriddeon sîðu te bedu [[AB]C]

The remaining pattern axa is no less common in the Heliand. This pattern, while conforming to the requirement of double alliteration for heavy a-verses, is deviant in another respect, however: alliteration occurs on the third rather than the second class 1 word contained in a verse, thereby disobeying the principle of left dominance. Given the second and third class 1 words, this principle should give priority to the former in realising alliteration, much as the first word is chosen over the second in two-word verses. (44) List of heavy verses with the alliterative pattern axa (in violation of the principle of left dominance), and their syntactic constituency 6a liudo barno lobon [[AB]C] 16a firiho barno frummian [[AB]C] 52a firio barnon ti frumon [[AB]C] 95a uuîsa man mid uuordun [[AB]C] 268a thes uuîdon rîkeas giuuand [[AB]C] 493a the aldo man an them alaha [[AB]C] 503a uuîsas mannas uuord [[AB]C] 593a threa man fon thero thiodu [[AB]C] 595a that godes bôcan gangan [[AB]C] 754a Krist selbon aquellian [[AB]C] 865a gôdlîc stemna godes [A[B]C] 1159a godes rîkeas forgeben [[AB]C] 1344a an godes rîkia garu [[AB]C] 1471a te them godes altere ageban [[AB]C] 1473a godes uuilleon fulgân [[AB]C] 1494a suâs man an saca [[AB]C] 1634a liudio barn ne lobon [[AB]C] 1707a hard trio end hebig (cf. 1774a uuîd strâta endi brêd (axx)) [[AB]C] 1710a suâses mannes gesiun [[AB]C] 1931a brêd strâta te burg [[AB]C] 2138a gôd lioht mid gode [[AB]C] 2170a liudeo barnum leof [[AB]C] 2171a godes uuilleon gumun [[AB]C] 2206a hêlandi Crist an hand [[AB]C]

MOSP-03.qxd

5/11/04

2:41 PM

268

Page 268

3. Resolution and alliteration 2234a 2285a 2805a 2870a 2882a 3132a 3323a 3483a 3665a 3710a 3783a 4002a 4231a 4826a 5148a 5804a

thie godes suno mid is iungron [[AB]C] gôd uuerk mid is iungeron [[AB]C] gôdes mannes forgang [[AB]C] grôt craft godes [A[B]C] sô grôte craft mid gode [[AB]C] sô gôd uuord undar gumun [[AB]C] tehan sîðun tehinfald [[AB]C] guod lôn at gode [[AB]C] themu landes hirdie te lobe [[AB]C] hlûd stemnie afhaben [[AB]C] grôt folc Iudeono [A[B]C] guod uuord for gumon [[AB]C] firiho barnun te frumu (also 5029a, 5712a, 5773a) [[AB]C] grim folc Iudeono [A[B]C] sundea lôsen gisald [[AB]C] thie grôto stên fan them grabe [[AB]C]

By way of comparison, the principle of left dominance for heavy verses applies in Beowulf almost without exception. As far as the configuration PPxP is concerned, only two examples exhibit the alliterative pattern axa in Beowulf as given below (Suzuki 1996a: 113), whereas there are as many as thirty-three examples with the alliterative pattern aax (Suzuki 1996a: 410 n. 54): (45) Beo 147a Beo 545a

twelf wintra tid [[AB]C] fif nihta fyrst [[AB]C]

In this way, Beowulf shows an overwhelmingly strong preference for PPxP as opposed to PPxP (P ⫽ alliterative; P ⫽ non-alliterative). Of further interest is the frequent occurrence of PPxP with the syntactic constituency [[PPx]P] in the b-verse, that is, verses of the composition characteristic of type E with single alliteration. The restriction of the verse form [[PPx]P] to the b-verse may therefore be ascribed to the conflict between the principle of left dominance on the one hand and the rule of double alliteration for heavy verses on the other: the verse form [[PPx]P], normally scanned as type E, would result in PPxP in the a-verse in accord with the rule of double alliteration for heavy verses; the resulting configuration PPxP, however, would violate the principle of left dominance. Confronted with the structural incompatibility inherent in the configuration [[PPx]P], the Beowulf metre came closest to excluding its occurrence from the a-verse. As for the configuration PPPx, both PPPx (type A2a) and PPPx (type D2a) are rarely encountered in Beowulf: we find three examples of PPPx and a single example of PPPx alongside an example of PPpx (type D2a; Suzuki 1996a: 81, 96): (46) Beo 517a Beo 2313a Beo 2987a Beo 1485a Beo 90a

seofon niht swuncon A2a [[AB]C] beorht hofu bærnan A2a [[AB]C] heard swyrd hilted A2a [A[B]C] geseon sunu Hræ¯dles D2a [A[BC]] swutol sang scopes D2a [A[B]C]

MOSP-03.qxd

5/11/04

2:41 PM

Page 269

3.2. Alliteration

269

At first glance, the lack of preference for PPPx at the expense of PPPx might be regarded as contradictory with the principle of left dominance. On closer consideration, however, an alternative, more plausible account may present itself. What we witness here is not a preference for PPPx over PPPx, a distribution pattern contrary to what is empirically implied by the principle at issue, but an avoidance of PPPx as well as PPPx. We may accordingly be justified in assuming that something other than the principle of left dominance is responsible for the rarity of both PPPx and PPPx. Specifically, building on Neuner’s (1920: 61) observation, we may generalise that the succession of three primary-stressed syllables (PPP) would have disfavoured the use not only of PPPx (type D2a) as Neuner originally suggested, but also of PPPx (type A2a). Thus, the general avoidance of PPPx and PPPx in Beowulf does not seem to be incompatible with the principle of left dominance that is otherwise found in full force. Through comparison with Beowulf, the situation in the Heliand proves to be outstanding not only on account of the large number of axa heavy verses attested, but by virtue of their greater variety in terms of the metrical types represented. While the five examples with the alliterative pattern axa in Beowulf as mentioned in (45) and (46) above are either type A2a or type E, the corresponding verse forms are embodied in the Heliand by subtype A1s, types D*, and E*, as well as by types A1 and E, as exemplified as follows: (47) 2138a 2170a 2234a 2285a 2870a

gôd lioht mid gode (type E) liudeo barnum leof (type E*) thie godes suno mid is iungron (type A1) gôd uuerk mid is iungeron (type D*) grôt craft godes (subtype A1s)

Both suno (2234a) and craft (2870a) occupy the second position (the first drop) of type A1; uuerk (2285a) is scanned as part of the first drop of type D*; and lioht (2138a) and barnum (2170a) are associated with the heavy drop of types E and E*, respectively. Of particular interest about these examples are verses 2234a, 2285a, and 2870a, which involve the association of a class 1 word with a normal drop, the metrical position of minimal prominence, which is generally realised by an unstressed syllable or its concatenation. We might be tempted to assume here a complete reduction of lexical-stressed syllables for metrical purposes. By contrast, Beowulf does not provide any example of verse in which a class 1 word occupies a normal (as opposed to a heavy) drop; the Heliand verses like 2234a would accordingly have been excluded as unmetrical in Beowulf. Furthermore, while the second position of subtype A1s is prototypically filled by the second element of a compound in Beowulf, there is no instance of the configuration P#P#px, where the same position is occupied by an independent lexical word (Suzuki 1996a: 90). The configuration P#P#Px, on the other hand, is attested in Beowulf, as given in (46) above. The crucial difference here resides in the status of the second position: a normal drop in subtype A1s and a heavy one in type A2a. Finally, no examples are attested in Beowulf in which the first drop of type D* is realised by an independent word (section 2.8; Suzuki 1996: 110). In this way, the Heliand enlarges the range of permissible alliterative patterns for heavy verses. A question arising then first is whether there is any rule-governed

MOSP-03.qxd

5/11/04

2:41 PM

270

Page 270

3. Resolution and alliteration

motivation for choosing a particular pattern for a given verse from the available options, particularly the patterns aax, axa, and axx. Of greater interest still will be to account for the reason for the enriched variety of alliterative patterns for heavy verses in the Heliand. More specifically, we need to explore the mechanism that led to a weakening of the originally canonical alliterative pattern aax for heavy verses as prescribed by the principle of left dominance. As the data presented above indicate, aax is by far the commonest alliterative pattern for heavy verses, followed by axa and then axx. This appears to show that the principle of left dominance is still largely at work in the Heliand, yet it is weakened so as to allow for a certain range of exceptions. It does not seem warranted to assume, however, that the principle of left dominance is simply curtailed and that exceptions are permitted to occur on a random basis, as claimed by Rieger (1876: 21–2). When we look into the constituent structure of verses rather than the overall frequency of alliterative patterns, we find that alliterative patterning is closely correlated with the syntactic relation of the constituent words involved. Specifically, given three constituent class 1 words of a verse (represented for the sake of reference as A, B, C, counted from the beginning of a verse), we may recognise three different configurations in terms of syntactic constituency, as follows: (48) a. [A[BC]] b. [[AB]C] c. [A[B]C] In (48a), in which B and C are closely related to the exclusion of A, the second and third words form a constituent. In (48b), the overall relationship is reversed: the first and second words count as being organised into a constituent, the third one being external to it. (48c) is a hybrid of (48a) and (48b) in that the first and second words on the one hand, and the second and third on the other, may be parsed as forming a constituent on their own. In other words, (48c) is amenable with equal plausibility to two different manners of parsing, characteristic of (48a) and (48b), respectively. The constituent structure (48c) is exemplified by the construction adjective ⫹ noun ⫹ noun (genitive), like verse 2133a hêlag folc godes (cf. Hofmann 1991: 78–80). In addition to the overall constituency of verses, the linear order figures as a conditioning factor. Given the two words that form a syntactic constituent, of vital importance is the linear order in which they appear in the constituent on the surface, rather than the syntactic functions that they play in that constituent (cf. Hofmann 1991: 77–8). The significance of linear ordering is evidenced by the following pairs of verse in which godes serves as a modifier of barn: godes alliterates when used as the second word in a verse (2821a); however, the word does not alliterate when located after the head noun barn (2371a). (49) 2821a gumon te themu godes barne [A[BC]] 2371a be biliðiun that barn godes [A[BC]] On the basis of the data provided in (41), (43), and (44), we may summarise the distribution of the three constituent types and the three alliterative patterns as in Table 3.35.

MOSP-03.qxd

5/11/04

2:41 PM

Page 271

3.2. Alliteration

271

Table 3.35. Distribution of alliterative patterns in relation to three constituent types

aax axa axx

[A[BC]]

[[AB]C]

[A[B]C]

100 93% 0 0% 7 27%

0 0% 39 91% 6 23%

7 7% 4 9% 13 50%

In the light of the data given in Table 3.35, we can hardly miss complementarity between the patterns aax and axa with respect to co-occurrence with the constituent structures [A[BC]] and [[AB]C]: aax and [A[BC]] on the one hand, and axa and [[AB]C] on the other, are matched so consistently that the alternative pairing is completely ruled out (cf. Heusler 1956: §146). No less significant seems that the hybrid constituent type [A[B]C] is found capable of realising all the three alliterative patterns, aax, axa, and most frequently axx. This apparent lack of regularity, first pointed out by Rieger (1876: 21–2), seems to be illustrated most tellingly by verses of the same composition X folc godes, as given below: (50) aax: 412a fagar folc godes axa: 3783a grôt folc Iudeono axx: 2133a hêlag folc godes Yet such a highly flexible alliterative patterning of [A[B]C] may be readily explainable by referring to the structure that it shares with [A[BC]] on the one hand and [[AB]C] on the other. We may therefore be led to restate the generalisation that the alliterative patterns aax and axa are incompatible with the constituent structures that are by no means parsable as [A[BC]] and [[AB]C], respectively; that axx is most favourable to [A[B]C]; and that inasmuch as [A[B]C] is partly equivalent to [A[BC]] and [[AB]C], the alliterative pattern axx may be manifested in these two constituent structures as well, though less frequently. A comparison with Beowulf will reveal that the generalisation proposed above on the correlation between alliterative patterning and constituent structure in the Heliand is a systematic extension of what was traditionally practised. Of the five examples of heavy verses with axa in Beowulf, four (Beo 147a, 517a, 545a, 2313a) embody the constituent structure [[AB]C], the remaining one (Beo 2987a) being [A[B]C], as indicated in (45) and (46) above. By contrast, of the two heavy verses with aax in Beowulf, one (Beo 1485a) instantiates [A[BC]], the other (Beo 90a) [A[B]C], as shown in (46) above. Thus, we find the same correlation recurring in Beowulf and the Heliand. Having identified the syntax-based regularity that controls the alliterative patterning of heavy verses in the Heliand, as well as its inheritance from traditional versecraft, we move onto the second question raised above: why and how did the Heliand poet come to enrich the range of alliterative patterns for heavy verses by generalising the traditional practice? One might conjecture that types E (PPxP) and E* (PxPxP) derived from subtypes D2b (PPxP) and D*2b (PxPxP), respectively, by demoting the second lift and

MOSP-03.qxd

5/11/04

272

2:41 PM

Page 272

3. Resolution and alliteration

conversely promoting the heavy drop; by the same token, subtype A1s (PPpx) would have been derived from subtype D2a (PPpx) by reversing the status of the second and third positions. Such an account may appear viable in purely operational terms, yet something deeper seems to be involved here. Moreover, the above account would have to leave the emergence of type A1 (PPxPx) unexplained, because the logical base configuration for it, PPxPx, was simply unavailable as a normal verse form in the traditional metre. As argued in section 2.1.7 above, the categorical distinction between normal and heavy drops was largely lost for class A, and therewith the traditionally distinct type A2a was integrated as a graded member of type A1 by metrical reorganisation. As a consequence, the first drop of type A1 was given access to a wider range of language materials for association than allowed in the traditional metre: among others, the strings -S#x and -Sx# came to be established as legitimate occupants of the position at issue in the Heliand. Thereupon it would have been only a short step to making available the string P#x as a further occupant by analogically substituting the secondary-stressed syllable (the second element of a nominal compound) with the primary counterpart (an independent nominal); a sequence containing a class 1 word was accordingly introduced to the Heliand metre as an acceptable realisation form of the first drop of type A1. Similarly, the sequence PPpx would have been re-evaluated as a metrical variant of subtype A1s in the a-verse, partly by analogy to the well-established subtype A1s variant PSpx, and partly by analogical extension of the traditionally accepted b-verse variant of subtype A1s, PPpx (Suzuki 1996a: 89–90). The first process would scarcely require explanation: much as with the string P#x discussed above, it would have been implemented by analogically substituting the secondarystressed syllable with the primary-stressed counterpart building on their common lexical basis. The second source of analogy was based on the prior existence of the string PPpx as a subtype A1s verse in the b-verse: e.g., Beo 908b snotor ceorl monig. By contrast, the apparently identical sequence was unknown as an a-verse in Beowulf, with a single exception: Beo 376a heard her cumen. This verse, however, cannot be scanned as subtype A1s on account of double alliteration: it should be type D2a (Suzuki 1996a: 90). Beowulf was thus implicated in a complete asymmetry with respect to the distribution of subtype A1s: the heavy variant PPpx is limited to the b-verse. The Heliand poet, confronted with such an asymmetrical distribution, would have been induced to redress it in a way highly characteristic of his keen sense of balance and harmony. Specifically, in keeping with the reorganisation of the first drop of type A1 in general as considered in section 2.1.7 above, the same position of the same metrical type was subjected to a comparable harmonisation, so that the heavy variant PPpx of subtype A1s came to be used in the a-verse as well. A similar account may extend to the use of a class 1 word in the first drop of type D* (verse 2285a, given in (47) above). As discussed in section 2.8 above, the position at issue allows for a greater range of language material to be associated with, so much so that it may be occupied even by a lexical-stressed syllable, that is, the second element of a compound (e.g., 5042a breosthugi blôðora). The appearance of a class 1 word in independent use may therefore be understood as a further case of analogical extension.

MOSP-03.qxd

5/11/04

2:41 PM

Page 273

3.2. Alliteration

273

As regards the occurrence of PPxP (type E) in defiance of the principle of left dominance, we may draw attention to the sharp increase of the type E variant PSx . . . P in the a-verse in the Heliand. As discussed in detail in section 2.9 above, the variant in question, which had been favoured as much as other variants of type E in the traditional metre, came to be used with greater frequency in the Heliand as a consequence of gradation and harmonisation. Once again, the lexicalbased analogical substitution of the secondary-stressed syllable with the primarystressed one would have been instrumental in making the string PPxP available as a further variant of type E in the a-verse. Furthermore, as with the PPpx variant of subtype A1s discussed above, the prior existence of the string PPxP as an acceptable variant of type E in the b-verse would have served as a source of analogy for making possible the extensive use of the same form as a metrical variant of type E in the a-verse, which may be interpreted as another manifestation of the harmonisation of the a-verse and the b-verse that the Heliand poet accomplished. As noted above, Beowulf offers only four examples of PPxP in the a-verse, whereas there are 132 instances of comparable type E b-verses in it. The Heliand poet’s strive for balance and harmony then subverted the traditional avoidance of using the variant at issue as an a-verse. Finally, the variant PxPxP as a heavy variant of type E* may be understood as a logical extension of this type E variant (PPxP). Underlying the introduction of the novel forms PPxPx (type A1), PPpx (subtype A1s), PPxP (type E), PxPxP (type E*) is thus recognised the Heliand poet’s profound sense of structural harmony as an ultimate source of innovation, and the exact process of innovation was implemented by analogical extension and levelling. The traditional restriction against use of these forms, namely the principle of left dominance, however, obviously failed to work as an impediment to the innovation. This principle would have ruled out the new forms in favour of PPxPx (unscannable), PPpx (subtype D2a), PPxP (subtype D2b), and PxPxP (subtype D*2b), respectively. The principle of left dominance is reducible to the principle of maximal contrast, an underlying grand principle that governs the wholesale metrical organisation of traditional metre in general and Beowulf in particular (Suzuki 1996a: 389–92). The principle of left dominance may therefore be regarded as a corollary of the principle of maximal contrast that is derived as a metrical counterpart of the predominant word stress pattern / x (Suzuki 1996a: 385–6). As will be shown in the concluding chapter below, the underlying principle of maximal contrast is diametrically opposed to the one of balance and harmony that motivated and implemented the metrical reorganisation of the Heliand as a whole. The keen sense of balance and harmony that has been characterised above as a significant trait of the Heliand poet may well be conceptualised as a mental equivalent of the principle at issue. In this light, it would be hardly surprising that the principle of left dominance, which constituted a derivative of the grand principle of maximal contrast and which was in full force in the traditional metre, gave way to the innovation that the Heliand poet designed building on the alternative underlying principle. In the Heliand, with the diminished canonicity of the alliterative pattern aax for heavy verses, and the resultant multiplicity of alliterative patterns made available to them, syntactic constituency ascended to figure as a major organising principle for

MOSP-03.qxd

5/11/04

274

2:41 PM

Page 274

3. Resolution and alliteration

determining alliterative patterning and consequently for scansion of heavy a-verses. In the light of such a vital role of syntactic constituency in metrical organisation, it may be justified to accord it a similar organising function also for comparable verses provided with single alliteration. Posited thus as an organising principle, syntactic constituency may help categorise otherwise ambivalent configurations. To be disambiguated by reference to syntactic constituency are the following three configurations: (i) P#P#px, (ii) P#Px . . . P, and (iii) Px . . . Px . . . P. Characteristic of the first configuration P#P#px is that the second position is realised as an independent lexical word, which, in the absence of double alliteration, may be identified either as the first drop (subtype A1s) or the second lift (subtype D2a) with apparently equal plausibility, as pointed out in sections 2.1.8 and 2.7 above. By virtue of double alliteration, on the other hand, verse 2870a grôt craft godes, also of the configuration P#P#px, constitutes a definite instance of subtype A1s, whereas verse 1163a faran folc manag, precisely of the same configuration, scans as subtype D2a. Given the distinct syntactic constituent structure [A[BC]] for subtype D2a, as substantiated by verse 1163a and the like, the two metrical types may accordingly be distinguished for instances with single alliteration: (51) [[AB]C]; subtype A1s 2218a dages lioht sehan 2836a tuê hund samad 2648b that godes barn sehan (with anacrusis) 3471b godes anst manot (52) [A[BC]]; type D, subtype D2a 4062a Krist godes sunu (also 5584a) 2908b Skrêd lioht dages 3364b hriop up thanen 5605b sehan lioht godes A comparison with the situation in Beowulf would seem particularly revealing. Altogether, there are twelve verses of the configuration P#P#px in Beowulf, all being b-verses, which I argued to be scanned as subtype A1s (Suzuki 1996a: 89–90). Of central interest here is the total absence of the same configuration in the a-verse, whether with single or double alliteration. Such an uneven distribution pattern would seem all the more significant when we are reminded that subtype A1s is used more extensively in Beowulf and that the variant P#P#px is correspondingly more frequent in it. We may naturally suspect that the avoidance of the configuration in the a-verse would have been motivated on structural grounds, rather than a matter of pure accident. As remarked in Suzuki (1996a: 49, 90), the complete absence would have been a consequence of the canonical alliterative pattern that was in full force in Beowulf, whereby given a heavy verse, that is, a verse containing three words qualified for serving as a lift by nature, the first two had to bear alliteration. Accordingly, as with heavy verses in general, the configuration P#P#px had to carry alliteration on the first two words, thereupon manifesting the alliterative pattern aax and resulting in type D, subtype D2a (type D2a in the system of the Beowulf metre). No structural

MOSP-03.qxd

5/11/04

2:41 PM

Page 275

3.2. Alliteration

275

basis would thus have been available for composing P#P#px as a variant of subtype A1s, whether with single or double alliteration. The situation dramatically changed in the Heliand, however. As discussed above, the Heliand allowed for a multiplicity of alliterative patterns for heavy a-verses, axa, axx as well as aax, all attested with demonstrably high frequency. Correspondingly, the pattern axa that we are concerned with here was realised by an array of metrical types, types A1, A2, D*, E, and E*, alongside subtype A1s. This proliferation of alliterative patterns available for heavy verses has been characterised as an innovation that was made possible by the obliteration of the canonical status of the pattern aax, argued above. Placed in such a diachronic context, the scansion of the configuration P#P#px as subtype A1s in Beowulf receives confirmation from the subsequent treatment of it in the Heliand: once the underlying general constraint responsible for avoiding this configuration in the a-verse was removed on independent grounds in the Heliand, it became free to be composed as an a-verse with single and double alliteration, and the whole distribution of the configuration came closest to that of the prototypical variant of subtype A1s, namely, PS#px. The emergence of the configuration P#P#px (exemplified by verse 2870a given in (47) above) in the Heliand, which scans as nothing but subtype A1s, would thus suggest that the same configuration attested in the b-verse in the traditional metre would have been identified as subtype A1s, as I claimed in my earlier work; upon raising of the prohibition against the alliterative patterns axa and axx, the same scansion became fully accessible to the a-verse, and as a consequence the heavy variant of subtype A1s was allowed to occur as an a-verse as well. In other words, the complete absence of the configuration in question as an a-verse with single and double alliteration in Beowulf would not have resided in its inherent incompatibility with scansion as subtype A1s, but stemmed from something extraneous to its internal properties. Let us move onto the second configuration P#Px . . . P, which may be scanned as subtype D2b or type E, as observed in sections 2.7 and 2.9 above. In all, there are fifty-two such instances in the corpus, of which thirty-one is of the syntactic structure [A[BC]], and twenty-one of the structure [[AB]C], exemplified and listed as follows: (53) [A[BC]]; type D, subtype D2b 794a godes êgan barn (also 838a, 1010a, 1335a, 3085a, 326b, 960b, 1135b, 1287b, 2000b, 2291b) 1596a sân aftar thiu (also 2100a, 3108a) 4778a uuacon êne tîd 5170a diap dôdes dalu 174b Bidun allan dag 196b Bêd aftar thiu 369b sunu ôdan uuarð 982b Krist up giuuêt 2250b Self up arês 2632b lisit aftar thiu 2762b gamen up ahuof 4102b Uueros gengun tô

MOSP-03.qxd

5/11/04

276

2:41 PM

Page 276

3. Resolution and alliteration 4241b 4329b 4622b 4824b 4879b 5490b 5803b

sô lioht ôstene quam ferid unmet grôt gramon in geuuitun Uuerod sîðode thô blôd aftar sprang folc ina umbihuarf sân up ahlêd

(54) [[AB]C]; type E 395a godes engil cuman (also 700a) 1117a godes engilo cumen 1774a uuîd strâta endi brêd 3107a godes rîki sehat 3981a tuâ naht endi dagas 4199a sehs nahtun êr 103b That uuerod ôðar bêd 639b endi the cuning selbo gibôd 769b godes engil sprac 926b man ôðar cuman 973b Krist selbo gibôd 2882b Thea gumon alle giuuarð 3074b kumad alle te thi 3144b lioht uuolcan skên 3362b berht lôn antfeng 3612b godes helpe bidun 3902b thes godes barnes uuord 4035b Ik thoh, frô mîn, te thi 4368b suart logna bifeng 4789b Godes engil quam Drawing on the generalisation adduced above, the verses of the structure [A[BC]] may be scanned as subtype D2b, while those with the structure [[AB]C] may be identified as type E. The proposed scansion turns out to be credible on metrical grounds as well. Of forty-one subtype D2b heavy verses with double alliteration (disregarding verse 5170a, which is at issue here; see under Table 2.80, section 2.7 above), only three, c. 7 per cent, contain a resolved disyllable in final position: 661a berht bôcan godes, 2860a, 4611a. By contrast, of a total of twenty-four instances of type E heavy verses with double alliteration (Table 2.89, section 2.9), there are as many as nineteen examples, c. 79 per cent, that end in a resolved disyllable: 510a sibun uuintar saman, 732a, 1159a, 1323a, 1344a, 1471a, 1494a, 1597a, 1707a, 2094a, 2138a, 2171a, 2914a, 3132a, 3257a, 3483a, 3710a, 4002a, 4076a. Such a marked difference in incidence of resolution between the two types seems fully in keeping with what obtains for the two groups of apparently ambiguous verses under consideration here. Specifically, the group of verses given in (53) is distinguished from that in (54) by the comparably low incidence of resolution: the first group, identified as subtype D2b, contains only two instances (c. 7 per cent) that end in a resolved disyllable (5170a, 4622b), whereas we count seven examples (c. 33 per cent)

MOSP-03.qxd

5/11/04

2:41 PM

Page 277

3.2. Alliteration

277

in the second group (395a, 700a, 1117a, 3107a, 3981a, 926b, 3612b), scanned as type E. Furthermore, the proportion of subtype D2b heavy verses to subtype E counterparts is nearly the same in the a-verse with double alliteration and in the b-verse, with the rate of approximately two to one. Since there is no special reason for otherwise, such a balanced distribution may add plausibility to the proposed disambiguation. The distinguishability of the configuration PPxP that rests on syntactic constituency contrasts the Heliand sharply with Beowulf. In it, the two types involved, that is, types D2b and E, were closest to being neutralised in the context at issue. As discussed fully in Suzuki (1996a: 113–17), in the a-verse the configuration was nearly always accompanied with double alliteration in the form of PPxP (type D2b) by virtue of the near uniformity of the alliterative pattern aax; the other two patterns PPxP (type E) and PPxP (ambiguous) were vanishingly rare. Overridden by metrical consideration in this way, syntactic constituency had accordingly little room for functioning. In regard to the b-verse, there are metrical grounds for scanning the entire body of verses involved as type E. Furthermore, resorting to the difference between the constituent structures [A[BC]] and [[AB]C] in disambiguation for the b-verse lacks independent motivation, because it hardly figures as a significant determinant for metrical categorisation of the a-verse. The third configuration to be disambiguated on the basis of syntactic constituency is Px . . . Px . . . P, which may be scanned as type D* (subtype D*2b) or type E*, as observed in sections 2.8 and 2.10 above. We find thirty examples in the corpus, which fall into two classes on the basis of syntactic constituency, [A[BC]] (sixteen examples) on the one hand and [[AB]C] (fourteen examples) on the other, as listed below: (55) [A[BC]] type D*, subtype D*2b 1085a scrîd thi te erðu hinan 4867a uuell imu innan hugi 5327a hreopun thuo alla samad 112b grurios quâmun im 1134b Krist im forð giuuêt 2036b Lârea stôdun thar 2161b Kristes uuârun thô 2467b endi hôrid thar mid is ôrun tô 3228b hala thi thar ôðara tô 3295b Sah imu aftar thô 3822b sâhun manage tô 3998b Duan ûs alla sô 4103b Uuânum up arês 5000b Thes thram imu an innan môd 5755b Thius thiod gilôbit te filo 5775b Uuânom up astuod (56) [[AB]C] type E* 1223a armoro manno filu 1555a Te hlûd ni dô thu it [A[B]C] (see below)

MOSP-03.qxd

5/11/04

278

2:41 PM

Page 278

3. Resolution and alliteration 2232a 3147a 3524a 4245a 4799a 254b 492b 4084b 4547b 5663b 5664b

allaro elithiodo cuman hêlag stemne godes [A[B]C] (see below) uuîse man be mi liudio barno nigên thriddeon sîðu te bedu gôdes cunnies man Listiun talde thô [A[B]C] (see below) fiuuuar naht endi dagos (also 4131b) fundun all sô he sprak harda stênos clubun endi that fêha lacan tebrast

As shown above, the two classes are differentiated by the location of the major constituent break, either after the first word or the second. Based on comparable verses accompanied with double alliteration (sections 2.8 and 2.10), we may be led to scan the verse instances with the syntactic structure [A[BC]] as type D*, subtype D*2b, and those with the structure [[AB]C] as type E*. There are three verses which may be parsed as [A[B]C]: 1555a, 3147a, 492b. Although ambivalent in structural terms, they may be tentatively regarded as belonging to type E* insofar as the first two constituents A and B are integrated analogous to [[AB]C]. There are several apparent or real exceptions to the generalisation that our disambiguation is based on. The following examples, scanned as subtype D*2b by virtue of alliterative patterning, do not embody the structure [A[BC]]; rather, the first and third words constitute a discontinuous construction: (57) 3046a diurlîc drohtines bodo [A(B)C] (also 5806a) 5214a agâbun thi thîna gadulingos mi [A(B)C] These three verses may be subsumed under the class of the syntactic structure [A[BC]], however, inasmuch as the major constituent break comes before the second word, although it and the following word do not form an immediate constituent. On the other hand, we find a single type E* verse with the deviant syntactic constituency [A[BC]], as follows: (58) 978a

dôpte allan dag [A[BC]]

The above mismatch between metrical and syntactic structures reminds us of the well-established fact that in the traditional metre the two organisations were not perfectly isomorphic to and reducible to each other (Suzuki 1996a). What is characteristic of the Heliand metre is the increased extent with which the two structures came to parallel each other, with some residual inconsistencies inherited. And this harmonisation was brought about by the weakening of the canonical alliterative pattern aax for heavy verses. The use of class 1 words as occupants of the verse-internal, first drop of types A1, D*, E, and E* in the Heliand deserves consideration in a wider context than in exclusive connection with the alliterative pattern of heavy verses. Class 1

MOSP-03.qxd

5/11/04

2:41 PM

Page 279

3.2. Alliteration

279

words are apparently used as part of the verse-initial drop as well in the normal verse.31 Thus, we need a broader perspective on the issue by taking into consideration violation of another, but related metrical process, namely, lift formation. This process is also implemented on the basis of the same hierarchy of word classes as is responsible for determining alliterative elements, as observed at the outset of this section (see also Suzuki 1996a: 282, 381). Specifically, class 1 words, which generally occupy stressed metrical positions (lifts and heavy drops), occasionally serve as normal drops. A list of relevant examples is provided below: (59) List of verses in which class 1 words serve as part of the verse-initial drop, and of their position in a clause (compare examples of the violation of the principle of left dominance in (44) above) Type A1; clause-initial 3140a that man thi hêr an thesaru hôhe 1535b sô huat sô man iu an thesoro uueroldi gedôe 3140a, 849b, 1535b, 2456b Type A3; clause-initial 2438a them ôðrun scal man be biliðiun 2782a that man an thea halla 1640a, 1867a, 2142a, 2438a, 2782a, 5036a, 5223a Type B1; clause-initial 1479a that thar man is siuni mugun 1879b that man iu undar themu uuerode ne mugi 1479a, 1866a, 4154a, 4912a, 5244a, 1794b, 1879b, 1898b, 1947b, 2054b, 2191b, 2272b, 2739b, 2788b, 2868b, 3341b, 3433b, 4062b, 4612b, 4812b, 4868b, 4953b, 5149b, 5513b, 5618b, 5733b Type C; clause-initial 1521a ef man huuemu saca sôkea 3069b obar themu stêne scal man mînen seli uuirkean 1448a, 1484a, 1521a, 1839a, 1868a, 2629a, 5135a, 5243a, 5334a, 5623a, 1862b, 1996b, 2227b, 2564b, 3069b, 3194b, 3405b, 3732b, 4080b Type E; clause-initial 345a Hiet man that alla thea elilendiun man With a single exception of stêne ‘stone’ (3069b, given under type C above; cf. Bostock 1976: 315),32 the use of a class 1 word as part of the verse-initial drop in the normal verse is thus almost limited to man ‘man’ in pronominal function: it may appear in the first weak position of a verse, that is in the first drop of types A3, B1, and C, or as part of anacrusis of types A1 and E. The limitation is indeed notable,

31 32

For a similar phenomenon observed in the hypermetric verse, see section 4.1 below. The exceptional use of the full noun stêne as a constituent of the verse-initial drop might strike as less abnormal when we note that man in pronominal function also appears alongside it.

MOSP-03.qxd

5/11/04

2:41 PM

280

Page 280

3. Resolution and alliteration

since the comparable use of class 1 words in heavy verses involves a wide variety of lexical items besides man: e.g., barn ‘child’, craft ‘craft’, lioht ‘light’, uuerk ‘work’ (see (44) and (47) above). It would hardly be warranted to assume, however, that man no longer belongs to class 1 in the Heliand, thus incapable of forming a lift. That man is still qualified for constituting a lift and thereby capable of bearing alliteration is evidenced by the following examples: (60) 3405a 3518a 4262a 4890a 5249a

man fan dôðe (type A1; non-clause-initial) man obar erðu (type A1; non-clause-initial) manno barnun (type A1; non-clause-initial) man adôgen (type A1; non-clause-initial) manno the bezto (type A1; non-clause-initial)

In all these examples, the second lift is occupied by a demonstrable class 1 word such as dôðe ‘death’ and erða ‘earth’, as evidenced for example in verses 5237a dôdes uuirðig and 408a erðun endi himiles; cf. 4349a himiles endi erðun. Accordingly, if man were truly anything but a class 1 word, the second substantive in a verse would serve as a sole alliterating lift, and the resultant verse would scan as type A3. Confronted with the two apparently contradictory behaviour of man, we can hardly avoid noticing a complementarity in distribution that stands in correlation to the location in a verse: man is treated as a drop only when used non-initially in a verse; conversely, only when occurring verse-initially, is man found to be constituting an alliterative lift. By going beyond the domain of a verse and taking into account a broader context of clause structure, however, we may restate the complementary distribution in question more revealingly as being predicated on the opposition between clause-initial and non-clause-initial: man is used as part of a clause-initial drop in a verse, insofar as it does not appear verse-initially; in the remaining contexts, it is treated as a lift. In this respect, punctuation provided by modern editorship will offer a reasonable indication of a clause break, and this practice confirms the proposed generalisation, as noted in parentheses in the above exemplifications. At this point, it will be instructive to make a comparison with the situation in Beowulf. In it, we find four examples attested in which man/mon is treated as part of a normal drop: (61) Beo 1175a Beo 3056a Beo 3175a Beo 2355b

Me man sægde (type A3; clause-initial)33 he is manna gehyld (type B3; clause-initial) þæt mon his winedryhten (type C; clause-initial) þæ¯r mon Hygelac sloh (type B1; clause-initial)

The occurrence of class 1 words in a normal drop is limited to man in Beowulf as in the Heliand, although the process is observed much more frequently in the Heliand than in Beowulf. Furthermore, the demotion takes place only when man 33

Man is treated here demonstrably as equivalent to a class 3 word, since the alliterative element that appears on its right is a class 2 word.

MOSP-03.qxd

5/11/04

2:41 PM

Page 281

3.2. Alliteration

281

appears in clause-initial and non-verse-initial position, much in the same way as in the Heliand. We may therefore be led to assume that the exclusive use of man as part of a normal drop was deep-rooted in traditional metrical practice, rather than an innovation introduced in the Heliand. Characteristic of the Heliand is rather a radical increase of such occurrences, and this feature naturally requires explanation. Of no less significance, Beowulf seems to be regulated by the same rule as is the Heliand with respect to the distribution of alliterating man: in verse-initial position man may occur as a sole alliterative element even when followed by a class 1 word, as exemplified below: (62) Beo 2555a mannes reorde (type A1; non-clause-initial); cf. Beo 3065a mon mid his (ma)gum (emended) Although there is only a single definite example attested in the corpus, the following array of verses are found also compatible with the rule in question, insofar as the verse-initial appearance of man is attended by alliteration: (63) Beo 380a Beo 1729a Beo 2281a Beo 2996a Beo 3181a

manna mægencræft (non-clause-initial) monnes modgeþonc (non-clause-initial) mon on mode (non-clause-initial) mon on middangearde (non-clause-initial) manna mildust (non-clause-initial)

Noteworthy in this regard is that Beowulf does not provide any example in which verse-initial man does not alliterate and the following class 1 word serves as a sole alliterative element in a verse. The dual treatment of man as a lift or as a drop is therefore a feature of the traditional metre. The generalisation that has been proposed above to account for it is predicated on the opposition between clause-initial and non-clause-initial. Implicit behind such a generalisation is the assumption that man counted as a member of class 1 by default (i.e., non-initial) and that only in the marked (i.e., initial) contexts was it treated (for some reason or other; see below) as equivalent to a class 3 word and consequently it had to occur in a drop. Furthermore, by bringing clause structure to bear on the distribution as a significant conditioning factor, it entails the claim that the phenomenon under consideration was partly syntactic in nature. Here we encounter Kuhn’s Laws, his first law in particular, mentioned briefly in section 2.12.1 above: (64) Kuhn’s First Law (Law of Germanic Sentence Particles) Die satzpartikeln stehen in der ersten senkung des satzes, in der proklise entweder zu seinem ersten oder zweiten betonten worte (Kuhn 1933: 8; cf. Momma 1997: 56). By Kuhn’s First Law, sentence particles occur either before or after the first stressed word in a clause. Accordingly, when man is identified as a sentence particle, its placement in a clause has to be regulated by Kuhn’s First Law: it has to occur in the

MOSP-03.qxd

5/11/04

282

2:41 PM

Page 282

3. Resolution and alliteration

first weak position available. In metrical terms, this means that man used as a sentence particle must appear in the first drop, either in the first position of types A1 (with anacrusis), A3, B1, C, D (with anacrusis), D* (with anacrusis), and E (with anacrusis), or in the second position of types A1 (without anacrusis) and D* (without anacrusis). Of these two weak positions available to sentence particles, it is the verse-initial drop that is normally used for accommodating man in pronominal function. In other words, pronominal man appears in the second position of types A1 or D* (without anacrusis) with extreme rarity, as found once for each in the corpus: (65) 3299a ôður mag man olbundeon (type D*; clause-initial) 4813b lêdde man faklon (type A1; clause-initial) Of particular interest, man is used in verse 4813b as a sentence particle on its own, a unique feature in itself that would be unacceptable in the verse-initial drop of types A1 (with anacrusis), A3, B1, C, D (with anacrusis), D* (with anacrusis), and E (with anacrusis). By contrast, occurring in the first, verse-internal drop of type A1 with higher frequency is man used as a full noun, much as in the heavy drop of types E and E*, as follows: (66) 95a 493a 503a 593a 1494a 1710a 2805a

uuîsa man mid uuordun the aldo man an them alaha uuîsas mannas uuord threa man fon thero thiodu suâs man an saca suâses mannes gesiun gôdes mannes forgang

Such a sharply differentiated treatment of the two available positions requires explanation accordingly. As seems most likely, the extended use of anacrusis in the Heliand would have been responsible for the general avoidance of the second position of type A1 as a site for accommodating pronominal man. With relaxation on the traditional strict constraints, anacrusis was rendered capable of serving as a position for realising sentence particles as freely as did the first drop of types A3, B1, and C (section 2.11). As a consequence, type A1 came to be equipped with a means of expressing sentence particles verse-initially, a device that had been totally inaccessible in traditional practice. Type A1 thus had at its disposal two different positions for accommodating sentence particles, anacrusis and the original first drop. Given the two available options, anacrusis would have seemed to have more to recommend it for the following three reasons. First, as Kuhn (1933: 8 n. 1) observes concerning Beowulf and the Heliand, and Momma (1997: 95) reconfirms on the basis of more comprehensive data, the position preceding rather than following the first lift is put to much greater use for hosting sentence particles in Old English and Old Saxon versification. Second, located verse-initially, anacrusis is open to a greater range of language materials for accommodation by virtue of the Linearity-based Metrical Strength Scale (section 1.2), whereby the closer a drop is to the beginning of a verse, the

MOSP-03.qxd

5/11/04

2:41 PM

Page 283

3.2. Alliteration

283

stronger it counts and hence the longer a sequence of unstressed syllables it may be realised by. Third, we should also take into account the unique functional duality of man, nominal and pronominal, a property that has figured centrally in our discussion. Confronted with the two distinct positions available, the Heliand poet would have organised the whole distribution of man in type A1 on the basis of a neat oneto-one correspondence: anacrusis (that is, before the first lift) for pronominal man, the first drop (that is, after the first lift) for nominal man. Thus, the dominant occurrence of man in the first drop of types A3, B1, and C, as well as in initial position (anacrusis) of types A1 and E with anacrusis, arose as consequences of the secondary identity of man as a sentence particle and of the extension of anacrusis to type A1. Insofar as man is treated as a member of class 1 words by default, it normally occurs in a lift, as do other class 1 words in general. Furthermore, when appearing as the first occurrence of class 1 words in a verse, man has to alliterate in conformity with the principle of left dominance. In this way, obligatory alliteration on man in verse-initial, non-clause-initial position, as illustrated in (60) above, is fully attributed to the working of the rule in question. The complete absence of man as a nonalliterating lift in verse-initial position is accordingly a natural consequence of the principle of left dominance, rather than an aberrant property unique to man. Since the category of sentence particles and that of class 3 words largely overlap, as with pronouns, the treatment of man as a sentence particle would appear equal to saying that the item was recategorized as a pronoun, an indefinite pronoun meaning ‘one’, similar to man in German and on in French. Indeed, based on Sehrt’s (1966) readings there are no a-verses attested in the corpus that contain alliterating man used as a pronoun. This means that, apart from a few instances of presumably coincidental alliteration, as in 2444a huô man thea mârien scal; 3425a that man thero manno gihuem; 3429a that man them mannon; 3432a that man im mêra lôn, man in pronominal sense is generally not qualified for carrying alliteration, correctly in accordance with the rule of alliteration that disfavours alliteration on class 3 words, particularly toward the beginning of a verse. We may therefore be justified in concluding that man in pronominal use constitutes a sufficient condition for its use in the verse-initial drop in the a-verse. Much the same applies to the b-verse: of thirty-nine examples in which man is used as an indefinite pronoun and followed by more than two syllables in the same verse (111b, 682b, 746b, 849b, 1109b, 1370b, 1535b, 1794b, 1862b, 1879b, 1883b, 1898b, 1947b, 1996b, 2054b, 2191b, 2227b, 2456b, 2739b, 2774b, 2788b, 2868b, 3069b, 3194b, 3341b, 3433b, 3732b, 4062b, 4080b, 4107b, 4612b, 4812b, 4868b, 4911b, 4953b, 5149b, 5513b, 5618b, 5733b), man is not chosen as an alliterative element, with a single exception (4107b ni mag that man ôðrumu). The extremely low incidence of alliterating man in pronominal use even in the b-verse testifies to its lack of qualification for forming a lift. The use of man as an indefinite pronoun is generally considered to be limited to the nominative singular (cf. Sehrt 1966; Mitchell 1985). The occurrence of man as part of a clause-initial drop, however, extends to other case forms, though small in number: 1479a that thar man is siuni mugun (acc. sg. man); 1484a Ef than thana man is siun uuili (acc. sg. man); Beo 3056a he is manna gehyld (gen. pl. manna). Such examples indicate that man used as a pronoun does not provide a necessary

MOSP-03.qxd

5/11/04

284

2:41 PM

Page 284

3. Resolution and alliteration

condition for its occurrence as part of a drop in verse-initial position. It would accordingly be warranted to assume that metrical demotion of man is not strictly coextensive with its use as an indefinite pronoun: in all likelihood the process in question would have originated from grammaticalization of man, yet it would subsequently have been reconstituted as metrically motivated at least in part. As yet unresolved is the question why man may not appear at the beginning of a verse in the first drop of these metrical configurations. More specifically, why are verses unattested such as *man scal them ôðrun be biliðiun (cf. 2438a them ôðrun scal man be biliðiun)? An exclusive appeal to Kuhn’s First Law does not seem to provide an explanatory account. Two lines of explanation, complementing each other in part, may suggest themselves here, one syntax-based, the other metre-based. From a syntactic perspective, we may posit a syntactic rule which stipulates that man may not appear clause-initially. In fact, Mitchell (1985: §3937) observes: ‘. . . as yet I have recorded no examples of indefinite man ‘one’ beginning a sentence.’ Seemingly limited to a single lexical item in scope, this stipulation might be reducible in the final analysis to a more general rule for placement of sentence particles. On the metrical dimension, we might be prompted to assume that the dual character of man in metrical terms would motivate prohibition of its use as a drop in verse-initial position. This hypothesis would imply that the avoidance of man in verse-initial position is limited to this word, since the dual metrical treatment is unique to it. Specifically, with a default value, man is freely used as a lift in verseinitial position, and in such a capacity it has to bear alliteration in accordance with the principle of left dominance. If man were allowed to appear as a drop verseinitially, it would be necessarily immune to alliteration. As a consequence, we would be confronted with two apparently contradictory behaviour of man in verse-initial position, alliterating man on the one hand, non-alliterating man on the other.34 The non-alliterating verse-initial man, however, would appear to contravene the principle of left dominance, although strictly in structural terms it does not: because it is a drop, it stands categorically beyond the purview of the principle in question. Yet on the surface we could hardly distinguish the two treatments of man upon encountering its appearance in verse-initial position; only by parsing subsequent incoming verse elements could we make a proper scansion. Thus, the occurrence of nonalliterating man even as a drop in verse-initial position would involve complication in processing, and this may possibly have served as an additional, metre-based motivation against use of man as a drop in verse-initial position. At this point, we find ourselves better equipped to make sense of the above unique example, verse 4813b lêdde man faklon, which is doubly exceptional, as remarked above. First, this is the only occurrence of man in pronominal use in the first drop of type A1 (without anacrusis). Second, this is the only instance of man used as a pronoun independently, rather than embedded in a sequence of sentence particles. The clustering of these two exceptional features, however, would seem far from accidental; rather they would have been causally related. Since pronominal man may not occur verse-initially for syntactic and also metrical reasons as argued above, there would have been no room for an alternative composition *man 34

The situation would be further complicated by a possibility of accidental alliteration whereby the alliterating word of the same verse may happen to begin with /m/.

MOSP-03.qxd

5/11/04

2:41 PM

Page 285

3.2. Alliteration

285

lêdde faklon, which would otherwise have come into being given that anacrusis was reserved as a privileged place for realising sentence particles for type A1 verses. Thus in the absence of another sentence particle occurring together, man was precluded from appearing as part of anacrusis. There would accordingly have been no other possibility than to place the word after the first lift as part of the first drop, namely the existing verse. To be sure, this placement was not without a difficulty, because it had to breach the otherwise clear-cut complementary distribution of nominal and pronominal man used as a drop in type A1. Yet weighted against the extent of exceptionality involved in the other composition, this one would have been found less exceptional and chosen accordingly: the alternative *man lêdde faklon would have violated a syntactic rule of much greater coverage that goes well beyond verse-making practice. In summary, the foregoing consideration has identified the following structural constraints imposed on the occurrence in a drop of class 1 words in general and man in particular: (67) Constraints on the occurrence of class 1 words in a drop a. Only man may appear in a verse-initial drop; accordingly, man is the only class 1 word to be used as part of the first drop of types A3, B1, and C. b. Man always alliterates when appearing verse-initially; thus verse-initial man may not serve as a drop, nor may it occur as a non-alliterating lift. c. Man is used as part of a verse-initial drop only when it is preceded by some other like element in the same verse; in other words, man cannot initiate a verse-initial drop. Accordingly, man cannot serve as a drop on its own. d. The dual treatment of man is an integral part of traditional versecraft, rather than an innovation introduced by the Heliand poet. e. A variety of class 1 words other than man may be used as part of a verseinternal first drop. Concerning the specifics of the above generalisations on demotion of class 1 words, the following properties among others require explanation: (68) a. Why was man, rather than any other class 1 word, capable of occurring in a drop most extensively? b. Why did occurrence in verse-initial position require man to serve as an alliterating lift? c. Why did occurrence in clause-initial, non-verse-initial position require use of man as part of a drop? d. Why was the verse-internal drop more favourable to association with class 1 words than was the verse-initial counterpart? e. Why was use of man as part of the verse-initial drop sharply increased in the Heliand? In regard to Question (68a), the original motivation for treating man as a drop would have been its use as an indefinite pronoun through grammaticalisation, as suggested above. By virtue of its very general lexical meaning, this word would have

MOSP-03.qxd

5/11/04

286

2:41 PM

Page 286

3. Resolution and alliteration

been more susceptible than any other class 1 word to grammaticalisation, thereby developing as a pronoun. Consequently, the word came to be allowed to occur in positions which were limited to sentence particles in accordance with Kuhn’s First Law, the first drop of types A3, B1, and C in particular. The next two questions, (68b) and (68c), which concern the correlation between metrical status (lift versus drop) and location in a verse (verse-initial versus nonverse-initial), have been already addressed in part. Syntactically, placement of sentenceparticle man seems subject to a rule which prohibited the word from appearing at the beginning of a clause (cf. Mitchell 1985: §3937). Given this syntactic rule, and given also that a clause break usually corresponded to a verse break (though not vice versa), it may follow that man had to be placed non-verse-initially when used as a sentence particle. Confronted with the dual properties of man, nominal and pronominal, then, verse-initiality, by opposition, came to figure as a place that was allowed to be occupied exclusively by man used as a full noun and therefore as a lift. Furthermore, because appearance of man in verse-initial position was thus limited to the class 1 variant of it, and because the verse-initial position was leftmost in a verse, the word at issue had to alliterate by virtue of the principle of left dominance, whereby the leftmost lift took utmost priority in realising alliteration as far as word class ranking was respected. A supplementary answer to Question (68c) is provided by Kuhn’s First Law as well. This rule required sentence particles to appear as a unit (exclusively) either before or after the first stressed word (corresponding to the first lift in metrical terms), thereby precluding sentence particles from occurring in both positions simultaneously. Accordingly, when man occurred preceded and followed by sentence particles as in many examples, there was no other alternative than to treat it as a sentence particle. Put another way, treating man as a lift in the position that was embedded in a string of sentence particles would have violated Kuhn’s First Law by permitting them to occur in two separate positions (first and second drops) in a verse. Question (68d) may be explained when we recall that the verse-internal, first drop was qualified for association with class 1 words in traditional practice: it was characterised as a heavy drop of types A2a and E. After reorganisation, the Heliand metre came to allow the first drop of type A1 (under which traditional type A2a was now subsumed) to be occupied by a class 1 word in conjunction with non-lexical words (section 2.1.7); similarly, the first drop of type D* was given a wider range of language materials for association, including (though only occasionally) lexicalstressed elements (section 2.8); and the first drop of type E remained in use as a heavy drop (section 2.9). Thus, the occurrence of class 1 words in these verse-internal weak positions did not involve recategorising them as members of a less prominent word class as in the case of man; rather, it was simply a result of permitting a greater variety of language materials to be used as occupants of a drop by metrical reorganisation and harmonisation. Finally, Question (68e) may be solved in general terms in the light of ‘the increase in pronouns, adverbs, and other weak-stressed words that are avoided in earlier alliterative verse’ (Lehmann 1956: 110). Thus, in parallel to a greater demand for these function words, man in pronominal use would have come to be much favoured in the Heliand.

MOSP-03.qxd

5/11/04

2:41 PM

Page 287

3.2. Alliteration

287

3.2.3. The phonological basis of alliteration According to the conventional view, alliteration in Old Germanic metre is divided into two components, consonant alliteration and vowel alliteration, each stipulated as follows: (69) Consonant alliteration When the lift begins with a consonant, the initial consonant (regardless of whether or not it is followed by another consonant) alone is repeated by alliteration, with the exception of /s/-clusters, /sp-/, /st-/, /sk-/, which behave as units by alliterating exclusively with the identical clusters. Vowel alliteration When the lift begins with a vowel, the initial vowel alliterates with any other vowel. This complex of separate rules and conditions, conceptualised in segmental terms as stated above, is reducible to a syllable-based generalisation as follows: (70) Rule for Alliteration (Suzuki 1996a: 295) Alliteration applies to the optimal onset (the optimal syllable margin) that constitutes a proper subset of a given onset. A proper subset of a given onset is defined as a minimal onset structure that shares an overall sonority contour with the onset at issue. Alliteration thus repeats a minimal onset structure that is commensurate with the onset involved in terms of the sonority sequencing. The proposed rule draws on the sonority-based account of the syllable in general and of the onset in particular, such as was presented in Vennemann (1988). Just a brief exemplification will be sufficient here, as a fuller account was provided in Suzuki (1996a: 295–6, 307–9). Given a syllable CV-, the optimal onset that makes a proper subset of it is itself; the remaining candidate V-, the one without an overt onset element, is less preferred because V- counts as a syllable more marked on cross-linguistic grounds. As far as biconsonantal onsets CiCjV- other than /sp-/, /st-/, and /sk-/ are concerned, the optimal onset involved is Ci-, rather than Cj- and ø-: the latter is ruled out as remarked above, while Cj- is less preferable to Ci- by virtue of a greater amount of sonority inherent in it. Moreover, Ci- is compatible with CiCj- in terms of sonority contour: in both, sonority value consistently increases toward the nucleus without interruption. By contrast, the clusters /sp-/, /st-/, and /sk-/ receive distinct treatment. Other things being equal, the least sonorous, hence optimal onsets would be /p-/, /t-/, and /k-/, respectively. These onsets, however, cannot be regarded as proper subsets of the three biconsonantal clusters in question: the overall contour of /sp-/, /st-/, /sk-/ is a fall and rise of sonority, whereas /p-/, /t-/, /k-/ constitutes a simple rise. On the other hand, /s-/ is not chosen as the optimal onset because of a greater sonority of it than /p-/, /t-/, and /k-/, respectively. As a consequence, the clusters as a whole, which constitute the only proper subsets available, are chosen as targets of alliteration. Should /sp-/, /st-/, and /sk-/ be in turn followed by another consonant, /str-/ for example, the first two

MOSP-03.qxd

5/11/04

288

2:41 PM

Page 288

3. Resolution and alliteration

clusters alone are repeated, because these biconsonantal clusters are characterised as minimal structures of the triconsonantal clusters in terms of sonority sequencing. From a slightly different perspective, we may account for the cluster alliterations by invoking an interplay of two parameters underlying syllable marginality. Syllable marginality may be characterised along two dimensions, in terms of sonority value on the one hand and in terms of linearity on the other. Only in the cases of /s/-clusters do syllable marginality (being the least sonorous) and linear marginality (being at the left edge) diverge. In order to satisfy the two requirements, the clusters as a whole have to be involved in alliteration. Finally, a syllable beginning with a vowel (V-) does not contain any overt segment in the onset. It is accordingly this zero onset that is repeated by alliteration. Since the generalisation given above is strictly limited to the onset and thus unaffected by whatever stands after the onset, any syllable-initial vowels would appear free to alliterate with each other. In actuality, however, alliteration concerns unrealised onsets, rather than initial vowels of arbitrary quality. Concerning the specifics of alliteration in the Heliand, there seems nothing particularly notable, as the poem conforms to the Alliteration Rule (70) as fully as Beowulf. More specifically, while Old Saxon was affected to a considerable extent by syllable reduction due to the weakening of stress (section 1.3), traditional practice of alliteration remained undisturbed in the Heliand. This is hardly surprising, however, because syllable reduction largely concerns the coda, the less integrated part of the syllable in typological terms, and because alliteration is strictly confined to the onset as its domain. Yet there remains to be discussed an allegedly novel feature of alliteration in the Heliand, namely, the consonant plus vowel alliteration or ‘extended alliteration’ as Lehmann (1956: 108) put it. According to Lehmann (1953: 25–30; 1956: 107–8), alliteration in the Heliand frequently involves the sequence of the initial consonant and the following stressed vowel in conjunction. In other words, alliteration may concern CV-, rather than simply C-, in the Heliand. Lehmann’s claim is based on his detailed analysis of alliterative patterns involving CV-. His major empirical findings are reproduced as in Table 3.36 in a form slightly different in notation from the original (Lehmann 1953: 26; for further details, see his ‘Chart of alliterative syllables in vowel rimes’). As the table below makes evident, altogether 2278 lines out of 5981 in the corpus, or about 38 per cent of all constituent lines in the Heliand, realise CV-alliteration at least in two lifts. Underlying Lehmann’s analysis are a few working hypotheses that require explicit mention and justification. First, Lehmann dwells on the assumption that Old Saxon has a vowel system as follows: (71) Old Saxon vowel system according to Lehmann (1953: 27–8) short vowels: i, e, a, o, u long vowels: î, ê, â, ô, û diphthongs: io, iu, ei, au Accordingly, vocalic signs other than those listed above are interpreted as graphic variants of the above: for example, was equated with <ê> as a variant representation of /ê/, as was with <ô> as an alternative representation of /ô/.

MOSP-03.qxd

5/11/04

2:41 PM

Page 289

3.2. Alliteration

289

Table 3.36. Distribution of alliterative patterns involving the identical CVa-verse (double alliteration) first lift second lift

b-verse first lift

number of lines

1 1 1 0 0

1 0 1 1 0

177 514 601 558 1987

a-verse (single alliteration)

b-verse

number of lines

1 0

1 0

428 1716

1 1 0 1 0

total

5981

Key: 1 ⫽ involving identical vowel; 0 ⫽ involving non-identical vowel

Second, in determining the identity of vowels for CV-alliteration, Lehmann regards vowel length as immaterial, so that he subsumes a pair of short and long vowels of the same quality under a single class. The above two underlying assumptions, however, demand reconsideration. Beginning with the first working hypothesis, Lehmann makes no distinction between PGmc. */e/ (⫽ */e2/) and */ai/; nor does he distinguish between PGmc. */o/ and */au/. These pairs of front and back mid vowels, however, are treated rather differently, as indicated most tellingly by their orthographic differences. PGmc. */e/ and */o/ have graphic variants besides <ê>, and besides <ô>, respectively, particularly in the C manuscript (Rauch 1992: 129–30). Such a highly consistent distinction seems at odds with the systematic reduction of the two pairs of vowels to the single pair. Following the standard phonemic analysis, therefore, I postulate two kinds of mid long vowels, higher-mid /e/ (⬍ PGmc. */e2/) and /o/ (⬍ PGmc. */o/) on the one hand and lower-mid /␧¯/ (⬍ PGmc. */ai/) and /ɔ¯/ (⬍ PGmc. */au/) on the other. Thus, for example, while Lehmann analyses line 303 huô he sie sô forlêti, / sô iru thar ni uurði lêðes uuiht as containing two identical alliterative vowels /ê/ (forlêti, lêðes), I consider them distinct, on the strength of -lêt- ‘let’ ⬍ PGmc. *-le2t(cf. forlieti in C) and lêð ‘evil’ ⬍ PGmc. *laið (OE lað, OHG leid). Furthermore, primarily on the same graphic evidence, OS /e/ (<e>) and /o/ () are paired with OS /␧¯/ (<e>) and /ɔ¯/ (), respectively, rather than with /e/ (<e>/) and /o/ (/). Therefore, we should better postulate the following vowel system for Old Saxon (Rauch 1992: 87; notice that /a/ and /â/, and the like, are notational variants): (72) Old Saxon vowel system (revised) /i/ /u/ /i/ /u/ /io/ /e/ /o/ /e/ /o/ /␧¯/ /ɔ¯/ /eu/ /a/ /a/ /ai/

/iu/

/au/

MOSP-03.qxd

5/11/04

2:41 PM

Page 290

290

3. Resolution and alliteration

As regards the second assumption underlying Lehmann’s data, while his disregard of vowel length may be justified, his differentiation of simple vowels from diphthongs seems contestable. A systematic phonological thinking would rather require us to extend a similar analysis to diphthongs: just as /ii/ (⫽ /i/) is identified with /i/, so /io/ and /iu/ may well be reducible to /i/, since they begin with the same vocalic element; by the same token, /ei/ should be grouped with /e/ and /␧¯/, as should /au/ with /a/ and /a/. Making the necessary corrections as above, we put the whole corpus to a re-examination and arrive at the following result. Table 3.37. Distribution of alliterative patterns involving the identical CV(revised) a-verse (double alliteration) first lift second lift

b-verse first lift

1 1 1 0

1 0 1 1

1 1 0 1

a-verse (single alliteration)

b-verse

1

1

total of lines involving the identical CV-

number of lines 167 485 567 554 number of lines 369 2142

Key: 1 ⫽ involving identical vowel; 0 ⫽ involving non-identical vowel

We count 5460 lines which involve consonant alliteration in the Heliand. Of these, we find a total of 2142 lines, c. 39 per cent, that realise CV-alliteration at least in two lifts. As it turns out, then, the overall figure obtained through the re-examination confirms Lehmann’s result. The significance of the above figure, however, remains unrevealed until it is placed in wider contexts. Such contextualisations may involve comparison along two dimensions: intratextual on the one hand and intertextual on the other. On the intertextual dimension, Lehmann (1953: 28) provides valuable comparative data, thereby taking us nearer to a proper appreciation of the Heliand poet’s practice. According to Lehmann, the Heliand outranks other Old Germanic alliterative poems in its high proportion of alliterating CVs: in the Edda, less than 10 per cent of lines manifest alliterating CVs; the Hildebrandslied shows a somewhat higher proportion (less than 20 per cent), as does the first two hundred lines of Beowulf, yet these figures are far lower than the one for the Heliand. Since Lehmann takes into account only a fragment of Beowulf, and since Beowulf features prominently in our comparative examinations, we need to analyse the whole corpus in regard to CV-alliteration. For the purpose of counting CV-alliterations in Beowulf, in keeping with the analysis of the Heliand we make no distinction of short and long vowels, and monophthongs and diphthongs, insofar as they share the same first vocalic elements. The major findings of our examination are given in Table 3.38.

MOSP-03.qxd

5/11/04

2:41 PM

Page 291

3.2. Alliteration

291

Table 3.38. Distribution of alliterative patterns involving the identical CVin Beowulf a-verse (double alliteration) first lift second lift

b-verse first lift

number of lines

1 1 1 0

1 0 1 1

35 178 185 179

a-verse (single alliteration)

b-verse

number of lines

1

1

175

1 1 0 1

total of lines involving the identical CV-

752

Key: 1 ⫽ involving identical vowel; 0 ⫽ involving non-identical vowel

Table 3.39. Distribution of vowel alliteration involving the identical vowels a-verse (double alliteration) first lift second lift

b-verse first lift

1 1 1 0

1 0 1 1

1 1 0 1

a-verse (single alliteration)

b-verse

1

1

total of lines involving the identical V-

number of lines 20 29 41 40 number of lines 75 205

Key: 1 ⫽ involving identical vowel; 0 ⫽ involving non-identical vowel

The lines with alliterative CVs amount to 752 in number, or 28 per cent of a total of 2668 lines that realise consonant alliteration. The figure accordingly proves to be considerably lower than the one obtained for the Heliand, thus confirming Lehmann’s original observation by and large. The intertextual comparison accordingly brings to light the significantly high proportion of CV-alliteration in the Heliand. Turning next to the intratextual dimension of comparison, we are most interested in the proportion of vowel alliteration involving the identical vowels to the one with distinct vowels. As discussed above, the so-called vowel alliteration constitutes nothing but a special case of consonant alliteration, involving as it does a zero onset. Given this characterisation, we would naturally predict that the figure for the identical vowel alliteration may hardly be distinguished from the one for the identical CV alliteration. And our prediction is empirically borne out, as shown in Table 3.39 (cf. Mayer 1904: 413–19).

MOSP-03.qxd

5/11/04

2:41 PM

Page 292

292

3. Resolution and alliteration

Table 3.40. Distribution of vowel alliteration involving the identical vowels in Beowulf a-verse (double alliteration) first lift second lift

b-verse first lift

1 1 1 0

1 0 1 1

1 1 0 1

a-verse (single alliteration)

b-verse

1

1

total of lines involving the identical V-

number of lines 11 32 26 28 number of lines 45 142

Key: 1 ⫽ involving identical vowel; 0 ⫽ involving non-identical vowel

A total of 523 lines realise vowel alliteration, of which 205 lines, c. 39 per cent, repeat the identical vowels at least in two lifts. Beowulf provides a similar picture, in which the figure for the identical vowel alliteration, as indicated in Table 3.40, is no different from the one for CV-alliteration: 142 lines out of 514 involving vowel alliteration, c. 28 per cent, have at least two identical vowels alliterating. Recapitulating on the major findings of the foregoing comparative examinations, first of all, CV-alliteration and V-alliteration are reducible to a single process: the proportions involving the identical CV- on the one hand and the identical V- on the other are the same. Consequently, the process may be characterised as a repetition of the optimal syllable structure (rather than the onset as in conventional practice) that constitutes a proper subset of given syllables. The optimal syllable to be repeated consists of a prototypical syllable margin, i.e., the least sonorous onset element, and a prototypical syllable core, i.e., the most sonorous nucleus element. Second, the Heliand is distinguished from Beowulf and other works by the high frequency with which the optimal syllable is repeated by alliteration: approximately four out of every ten lines in the Heliand contain the identical alliterative syllables. The Heliand thus shows a pronounced preference for syllable alliteration in comparison with other poems. If the syllable, rather than the onset, may be involved in alliteration as the foregoing considerations make it plausible, it would be tempting to assume that the Heliand metre would have undergone a radical break with the traditional practice of alliteration, which is exclusively limited to the onset, as prescribed in the Rule for Alliteration given in (70) above. What deserves emphasis, however, is that the syllable alliteration as characterised above is an optional process, in contrast to the obligatory onset alliteration. In other words, implemented in a variable manner, the process presumably served as an embellishment to, as opposed to a structural constituent of, alliterative verse. In this light, what is characteristic of the Heliand is an ornamental use of what would hitherto have been regarded as an accidental feature, rather than a radical restructuring of what constituted a mandatory device in the traditional metre.

MOSP-03.qxd

5/11/04

2:41 PM

Page 293

3.2. Alliteration

293

We are then confronted with the same question as Lehmann (1956) raises: what motivated the Heliand poet in making more frequent use of syllable alliteration? According to Lehmann (1953: 37; 1956: 107–8), the lifted syllables lost much of their original prominence on account of the weakening of stress in Old Saxon; correspondingly, alliterating initial consonants were found to be inadequate to bind the line on their own; as a consequence, the poet extended the domain of alliteration to integrate the following vowels as its constituents. Lehmann’s observations, stated in very general terms as they are, seem essentially correct. Elaborating on Lehmann’s insight, then, will bring us to a deeper understanding of the extensive use of syllable alliteration characteristic of the Heliand. What remains to be explained among other things is the relation between the weakening of stress and the inadequacy of consonants as sole alliterating elements. In order to provide a principled account, we need to call attention to the mechanism of lift formation as discussed in the previous sections in this chapter. As may be recalled, lift formation is based on an interplay of two complementary processes, one primary, the other subordinate (Suzuki 1996a: 310–12, 382–4). The two processes at work are foot construction and alliteration. Of the two processes involved, foot construction provided a primary basis of lift formation. In the traditional metre, the lift came to being through a linguistic-metrical association with the bimoraic foot, a supersyllabic prosodic unit. By contrast, the (normal) drop, the structural opposite of the lift, was occupied by unfooted material. Since the foot depended on the rime for construction, and since the rime in turn determined syllable weight by mora counting (Suzuki 1996a: 310–11), we may conclude that the foot supplied foundational material for effecting prominence. In Old Saxon, however, this sharp categorisation between the lift and the drop as predicated on foot construction became blurred and fuzzy (see particularly sections 2.1.7, 2.4.2.4, 2.12.1, and 3.2.2 above), largely because foot construction was subject to proliferation as a result of the weakening of stress in the phonology (section 1.3). Correspondingly, the drop was increasingly filled by footed materials, as with the first drop of type A1 and the second drop of type B1, to the point where even class 1 words were permitted to occur as occupants of the normal drop (section 3.2.2). This decategorisation contributed to diminishing the significance of the foot as a privileged foundation of the lift: in the Heliand metre, being a foot constituted no more than a weak indication of functioning as a lift. Put another way, the lift lost much of its distinguishability founded on the foot that had served as a primary source of its prominence in the traditional metre. Complementing with the foot, alliteration played a subordinate part as a component of lift formation. Building on the foundation that the foot furnished, alliteration created a superstructure by repeating sound in the onset. The subordinate status of alliteration as against the foot is shown most clearly by the fact that not all lifts alliterated, while all lifts constituted feet or a part thereof. The foot and alliteration were thus brought together to create the lift as a position of enhanced prominence in a verse: realised by the foot (i.e., a long stressed syllable or its resolved equivalent), the lift constituted a maximally weighty position, with the added prominence effected by the alliterating onset element. However, in parallel to the dysfunctioning of the foot as a firm basis of the lift, alliteration was in part dissociated from the lifted syllable. At stake is a reconfiguration of

MOSP-03.qxd

5/11/04

294

2:41 PM

Page 294

3. Resolution and alliteration

double alliteration in the a-verse. Earlier in the traditional metre, double alliteration had primarily been a function of added metrical strength on the lift. More specifically, double alliteration had been realised prototypically when the lift needed reinforced prominence in opposition to a heavy drop found in the same verse, a drop that was associated with a foot. In the Heliand, however, double alliteration came to serve largely as an marker of a relatively long verse, rather than as a marker of an additionally strengthened lift as in the traditional metre (section 3.2.1.4). This change was occasioned by the blurred distinction between heavy and normal drops, a metrical reorganisation that was induced once again by the weakening of stress in the phonology (section 2.12.2). Thus, reconfigured as an attribute of lengthy verses, double alliteration can no longer be adequately characterised as a transparent indicator of a strengthened lift. Faced with the weakened empirical foundation of the lift, the poet would have felt the need to secure a means of restoring endangered prominence to the lift. Striking him as particularly problematic would have been the so-called vowel alliteration lacking in overt alliterative element. The poet would then have been urged to improve alliteration on vowel-initial syllables by making vowels susceptible to the metrical device in question. Incorporating vowels into alliteration if as a variable means, however, would in turn have had consequences for consonant-initial syllables as well. Since alliteration was a metrical device founded on syllable structure in the traditional metre, the vowel following the overt onset element had to be treated analogously as a constituent of the syllable, rather than taken in itself purely in segmental terms. In this way, in order to reinforce the weakened basis of the lift, the poet firmly seized upon the traditional mechanism of alliteration and explored more fully its structural potential by extending its domain of operation from a subsyllabic unit, the onset, to the minimal syllable, that is, the unit comprising the onset and the nucleus. By bringing the optimal syllable structure CV- rather than its minimal part C- to bear on alliteration, and accordingly by bringing forth higher resonance, the poet attempted to keep the lift outstandingly prominent and thus to save the whole metrical system from collapsing. A further comment on Lehmann’s account concerns a more general issue. In evaluating the overall significance of syllable alliteration in the Heliand, Lehmann (1953: 28, 30) emphasises the poet’s break with the Old Germanic alliterative tradition. As I see it, Lehmann places more emphasis than warranted on the innovative character of syllable alliteration at the cost of failing to see continuity in tradition underlying the apparent break, largely because he considers syllable alliteration on its own, in isolation from vowel alliteration, as well as in sharp opposition to consonant alliteration. Thus, Lehmann (1953: 30) invokes Latin poetry as possible sources of influence on the new alliterative principles adopted by the Heliand poet. Although external influences may have been in force, in our view the Heliand poet, and also the Genesis poet following him (section 5, Appendix 2), stood firmly in the Old Germanic alliterative tradition. As with other metrical properties, the Heliand poet put to reorganisation certain inherited specifics of alliteration that he found at odds with the new linguistic environments in which he composed. In so doing, however, he implemented innovations by minimally departing from traditional practice and making most of the inherited properties for balanced reorganisation.

MOSP-04.qxd

5/11/04

2:42 PM

Page 295

4 Hypermetric verses and lines: diversification and restructuring

Old Germanic alliterative metre has at its disposal two distinct kinds of verse and line, the normal and the hypermetric (section 1.2), presumably as an inherited feature of Indo-European metre (Suzuki 1988; 1992). In this chapter, I shall be concerned with the synchrony and diachrony of the hypermetric verse and line in the metre of the Heliand, with emphasis on the diversification and restructuring that this metrical category underwent.

4.1. The composition of hypermetric verses: a synchronic perspective One of the defining features of the normal verse is that the maximal number of lifts does not exceed two. However, there are a number of verses that have more than two lifts. These overlong or overweighty verses are called hypermetric verses. Hypermetric verses thus by definition cannot be reduced to any of the existing normal verses with two lifts. More specifically, a hypermetric verse contains a lexical-stressed syllable that cannot be scanned as a drop in contrast to a heavy verse, which is fully reducible to one of the independently occurring forms of normal verses by demoting that extra strong syllable in metrical terms. Formal irreducibility to a normal verse is thus identified as a primary structural feature of a hypermetric verse. Some implications of this working hypothesis will be explored in due course below. Despite the presence of an extra lift, the number of alliterating lifts per verse is, with rare exceptions, two in the a-verse and one in the b-verse, as with the normal verse; the additional, third lift in the hypermetric verse is thus immune to alliteration.1 Distinctively characteristic of the hypermetric verse is accordingly that double alliteration is obligatory in the a-verse in sharp contrast to heavy normal verses, which may be attended by single alliteration (see section 3.2.2 above).

1

There are only two examples of hypermetric verses with triple alliteration (one of which, moreover, contains a foreign word, and is thus excluded from systematic consideration in this book; see section 1.4 above): 1315a thie hebbiad iro herta gihrênod; 3062a sâlig bist thu Simon, quað he, sunu Ionases.

MOSP-04.qxd

5/11/04

296

2:42 PM

Page 296

4. Hypermetric verses and lines

With the above initial observations in mind, we proceed to an examination of structural properties of hypermetric verses in the Heliand. In the whole corpus, we count 158 definite examples of hypermetric verses, and more or less a similar number of possible ones.2 Disregarding the latter group for the time being, we now concentrate on the verse structure of these clear instances. Of these 158 examples, as many as 128 instances are a combination of type A1 verses (type A1A1), as exemplified below: (The underlined words are those shared by the first and second constituents of hypermetric verses, and they will be referred to as linking words since they serve to integrate the two constituents into an organised whole. For detailed discussion on this notion, see below.) (1) Type A1A1 557a erlos fon ôðrun thiodun 1311a thie rincos, thie hîr rehto adômiad a-verse with double alliteration: 557a, 558a, 559a, 560a, 600a, 601a, 602a, 603a, 605a, 881a, 899a, 900a, 990a, 991a, 992a, 993a, 1271a, 1301a, 1306a, 1307a, 1308a, 1309a, 1310a, 1311a, 1312a, 1313a, 1314a, 1315a, 1316a, 1317a, 1318a, 1320a, 1321a, 1512a, 1542a, 1556a, 1667a, 1682a, 1684a, 1685a, 1686a, 1688a, 1689a, 1833a, 2208a, 2209a, 2210a, 2211a, 2212a, 2213a, 2214a, 2215a, 2358a, 2614a, 2615a, 2822a, 2823a, 2824a, 2825a, 2903a, 2904a, 2985a, 2987a, 2988a, 2989a, 2990a, 3037a, 3038a, 3065a, 3066a, 3067a, 3072a, 3125a, 3241a, 3493a, 3494a, 3495a, 3496a, 3497a, 3498a, 3499a, 3500a, 3501a, 3502a, 3503a, 3505a, 3506a, 3560a, 3677a, 3971a, 3990a, 4251a, 4265a, 4392a, 4393a, 4394a, 4395a, 4396a, 4413a, 4418a, 4568a, 4986a, 5228a, 5354a, 5379a, 5551a, 5609a, 5663a, 5722a, 5732a, 5812a, 5813a, 5916a, 5917a, 5921a, 5922a, 5923a, 5924a, 5925a, 5927a, 5928a, 5930a, 5931a, 5932a, 5933a, 5975a; b-verse: 1685b, 3067b At this point, we need to provide comments on several notable verses. Verse 899a faran an fern that hêta appears aberrant in that the second drop of type A1 is constituted by a function word that. Although not very common, similar verses are found in independent use, as exemplified below (section 2.2.1): (2) 5104a gihid that he god sî (similarly 5191a) 46b Ên uuas iro thuo noh than 490b drohtin frô mîn (similarly 2109a, 971b, 4861b, 5017b) Verse 3067b hugiskefti sind thîne stêne gelîca might be alternatively scanned as A1⫹A1 ([hugiskefti sind thîne][stêne gelîca]), that is, a juxtaposition of two type A1 verses without a linking word.3 Yet it should better be scanned as type A1A1, with

2

3

I follow Sievers (1878), Hofmann (1991: 53), and others in aligning drôgun in line 2858 to the b-verse, so that we obtain verse 2858a is geba gerno (type C) and verse 2858b drôgun gumono gihuemu (type B1). Furthermore, with Hofmann (1991: 55), I relineate verses 5112b and 5113a by regarding mahtin as belonging to verse 5113a; accordingly, these two verses should read huat sie themu godes barne (type C), rather than huat sie themu godes barne mahtin, and mahtin sô haftemu mêst (type B1), rather than sô haftemu mêst. See in this connection the treatment of verse 1687a (given in (5)) below.

MOSP-04.qxd

5/11/04

2:42 PM

Page 297

4.1. The composition of hypermetric verses

297

stêne serving as a link. A similar structure with the second element of a compound serving as part of the extended first drop is embodied in a demonstrable type A1A1 verse, 2215a uuârsagon an thero uueroldes rîki. Anomalous about verse 3067b is rather that this is one of the two isolated instances of type A1A1 in the b-verse, the other example being verse 1685b god uuili is alles râdan.4 (See also related discussion on verses 1308b, 1312b, 1314b, and 2825b further below.) A vast majority of type A1A1 examples have their third lift realised by a primarystressed syllable, as illustrated so far. Yet there are eight instances (605a, 1301a, 1833a, 2210a, 2212a, 4568a, 5379a, 5609a) in which the third lift is occupied by a secondary-stressed syllable, that is, the second element of a compound, as shown below: (3) 605a thana cuning an thesumu kêsurdôma 1301a man an thesoro middilgardun 4568a uuîti endi uunderquâle5 In these verses, the constituent break that terminates the first component verse coincides usually with a word-internal break (e.g. kêsur-dôma ‘kingdom’, uunder-quâle ‘monstrous torture’), rather than with a word boundary #. Exceptional in this respect, however, is verse 1833a allun them ungelîco: the verse-internal break decomposes ungelîco ‘unlike’ into unge- and -lîco, in defiance of the morphologically motivated analysis un-gelîco. Type A1A1 differs from type A1 in manners of realisation of the first drop: it shows preference for a larger number of syllables as an occupant of the first drop, as indicated in the overall distribution pattern given in Table 4.1. Further, while the maximal number of syllables occupying the first drop is seven for type A1 (321a, 611a, 1637a, 1662a, 2706a, 4440a, 4491a, 4822a, 4974a), it amounts to ten for type A1A1, as illustrated below: (4) Type A1 611a sâligoron undar them gisîðea (seven syllables) 1637a sô ferhtlîco undar thesumu folke (seven syllables) 4974a an thînun uuordun endi an thînaru uuîson (seven syllables) Type A1A1 5917a thena hêrron, thar iro uuârun at thia helpa gilanga (eight syllables) 3497a grimmes than lango, the he môste is iuguðeo neoten (nine syllables) 3037a the iungaron the he imu habde be is gôde gicorane (ten syllables) 4

5

Characterised in formal terms, these two isolated examples do not follow the requirement for the linking word to alliterate (see below), in favour of another generalisation, namely that given two lexical words of the same class (class 1), the first lift takes precedence over the second in realising alliteration, a corollary of the principle of left dominance (section 3.2.2). I follow the conventional practice of reading -quâla/-quâle ‘torment’, contrary to Hofmann’s -quala/-quale (with a short stem vowel). Accordingly, I differ from Hofmann in the scansions of verses 4568a, 5066a, 5379a, 5590a, 5609a, 2249b, and 5687b, which all contain the word uuunderquâla ‘monstrous torture’. The other verses containing the same item as the second element of compounds (ferah-: 5174a, 5396a; firin-: 4918b; thiod-: 4463a, 4795a), however, are unaffected by this difference as far as metrical type categorisation is concerned.

MOSP-04.qxd

5/11/04

2:42 PM

Page 298

298

4. Hypermetric verses and lines

Table 4.1. Distribution of types A1 and A1A1 according to the number of syllables in the first drop type A1 (a-verse) 1 2 3 4 5 6 7 8 9 10 total

651 26% 883 35% 450 18% 316 13% 155 6% 47 2% 9 less than 1% 0 0% 0 0% 0 0% 2511 100%

type A1A1 20 16% 29 23% 28 22% 29 23% 10 8% 5 4% 4 3% 1 less than 1% 1 less than 1% 1 less than 1% 128 100%

Table 4.2. Distribution of type A1A1 according to the number of syllables in the first and second drops first drop 1 2 3 4 5 6 7 8 9 10 total

20 16% 29 23% 28 22% 29 23% 10 8% 5 4% 4 3% 1 less than 1% 1 less than 1% 1 less than 1% 128 100%

second drop 94 34 0 0 0 0 0 0 0 0

73% 27% 0% 0% 0% 0% 0% 0% 0% 0%

128 100%

Comparing the two constituent verses of hypermetric verses, the first constituent tends to be longer than the second, because the first drop (of the first verse) is usually filled by a longer sequence of syllables than is the corresponding position of the second counterpart, as shown in Table 4.2. Such an asymmetry in composition accords fully with the linearity-based scaling of metrical strength, whereby the positions of the same metrical status (the normal drop in this case) are realised in more weighty form as they are farther removed from the end of a verse (section 1.2). Conversely, limited to shorter strings of syllables for association, the second drop is subject to a far narrower range of variation: only one or two syllables are allowed to occupy it; and the great majority (approximately three out of every four verses) contain nothing but a single syllable in their second drop.

MOSP-04.qxd

5/11/04

2:42 PM

Page 299

4.1. The composition of hypermetric verses

299

Table 4.3. Distribution of the numbers of syllables used as anacrusis in type A1A1 in comparison with type A1 number of anacrustic syllables 0 1 2 3 4 5 6 7 8 total

type A1A1 79 38 8 0 2 1 0 0 0

62% 30% 6% 0% 2% 1% 0% 0% 0%

128 100%

type A1 (a-verse) 1802 72% 503 20% 113 5% 58 2% 22 1% 6 less than 1% 5 less than 1% 2 less than 1% 0 0% 2511 100%

Finally, it will be of interest to examine use of anacrusis in type A1A1 through comparison with that in type A1 (a-verse), as summarised in Table 4.3. As should be clear from Table 4.3, the distribution pattern of varying numbers of anacrustic syllables is essentially the same for types A1A1 and A1: more than nine out of every ten verses have zero or at most one anacrustic element, and verses without anacrusis constitute the majority in either type. While type A1A1, counted 128 examples in the corpus, constitutes the majority of hypermetric verses in the Heliand, the remaining thirty examples are structured in widely different ways. According to metrical types of constituent verses as well as the manner in which they are combined, these atypical instances may be grouped into nine classes. In what follows, we provide comments on each class and/or specific members of it as appropriate. (5) Type A1⫹C (one example) 1687a Gerot gi simbla êrist thes godes rîkeas (Hofmann’s scansion: S2g-av/ CIIv52/x’1.2) The notation ‘A1⫹C’ means a juxtaposition of, rather than an overlapped combination of, types A1 and C. The whole verse constituency may accordingly be represented as follows: (6) [Gerot gi simbla êrist][thes godes rîkeas] In my view, this constitutes the only example that is composed by a juxtaposition of two normal verses, in disagreement with Hofmann (1991: 178–9). The scansion suggested above, following Hofmann (1991), seems better founded than the alternative reading as type D*, primarily because the additional drop between the two lifts, whether in subtypes D*1, D*2a, or D*2b, appears incompatible

MOSP-04.qxd

5/11/04

300

2:42 PM

Page 300

4. Hypermetric verses and lines

with a sequence containing a relatively weighty word like simbla and êrist. In fact, the position at issue is realised in its maximal form by a concatenation of light words, as illustrated below: (7) Maximal realisations of the extra drop of type D* 2948a gâhon te themu godes sunie 5214a agâbun thi thîna gadulingos mi 5298a alêdian hiet ina lungra mann 5473a Hiet im thuo te is handon dragan 2467b endi hôrid thar mid is ôrun tô There are three classes that begin with type B1: types B1A1, B1C, and B1D. We may consider them in turn. (8) Type B1A1 (seven examples) 954a He dôpte sie dago gihuuilikes 1096a gibrengen uppan ênan berg then hôhon 1561a ne galpo thu far thînun gebun te suuîðo 3063a gimarcon an thînun môdgithâhtiun 3068a sô fast bist thu sô felis the hardo 5918a that uuîf ni mahta uuôp forlâtan 5919a gimerrid uuârun iro thes muodgithâhti There are two outstanding features about the first constituent verse (cf. Hofmann 1991: 159). First, the first drop is minimal in realisation: it is occupied by a single syllable, not infrequently by a prefix. Such a minimal realisation of the first drop of type B1 stands in sharp contrast to type B1 in independent use. Comparing the like with the like and therewith concentrating on the a-verse,6 the first drop of type B1 (a-verse) is occupied most frequently by two or three syllables, together accounting for more than the half of all type B1 occurrences, followed by four syllables, as demonstrated in section 2.4.1 above. The second notable feature is the converse of the first: given verse 5919a with six syllables standing between the first two lifts, the second drop of the first constituent verse exceeds what seems to be the limit on the corresponding position of type B1 in normal use. The maximal number attested is five, as witnessed by the following unique example: (9) 3097a githolos undar thesaru thiod Yet the authenticity of the above verse seems somewhat in doubt: while the manuscript reading (C) is githolos undar theson thioda, Hofmann emends it githolos undar thesaru thioda, thereby scanning it as type A1 with anacrusis. Being on the safer side, we might better regard verses with four syllables in the comparable position

6

We are reminded here that the a-verse shows preference for shorter realisations of the first drop of type B1 than the b-verse.

MOSP-04.qxd

5/11/04

2:42 PM

Page 301

4.1. The composition of hypermetric verses

301

as maximal in the way the second drop of type B1 is realised, such as verses 2231a and 4483a: (10) 2231a an is uuord endi an is uuerc 4483a âno uuîg endi âno uurôht With such a stricter interpretation, noteworthy turn out to be not only verse 5919a but also verses 1096a and 1561a. In any event, we can hardly dispute the oversize of the second drop of type B1 when used as part of a hypermetric verse. As regards the realisation of the drops of type B1 serving as part of a hypermetric verse, we are thus faced with the two deviations that go in the opposite directions: the too small number of syllables for the first drop on the one hand, and the too large number for the second drop. This doubly aberrant pattern, however, becomes subject to understanding when we relate it to the prototypical composition of the hypermetric verse, namely type A1A1.7 Earlier we pointed out that the first drop of type A1A1 is generally distinguished by the greater number of syllables associated with it, while anacrusis tends to be confined to a single syllable at the most as with the normal type A1 verse. This means in effect that type A1A1, too, is characterised by the relatively large number of syllables for the verse-internal first drop at the expense of the verse-initial position, namely anacrusis. In this light, the apparently peculiar pattern of type B1 in hypermetric use proves to be highly analogous to the prototypical relation that obtains for the corresponding positions in type A1A1. The deviant pattern in question is accordingly attributed to the general property of the hypermetric verse: minimise the weak position before the first lift, maximise the weak position after the first lift. Given the nearly exclusive occurrence of the hypermetric verses at issue (i.e., types A1A1 and B1A1) in the a-verse, this trade-off relation may in turn be explained as a corollary of the generalisation, formulated in section 2.12.3.2 above, that the a-verse shows preference for expanding the drop standing after the first lift at the expense of the verse-initial counterpart. We move onto a second type beginning with type B1, namely type B1C: (11) Type B1C (one example) 604a that uui ina selbon gisehan môstin This verse seems unnatural in three respects. First, in the light of the above discussion, the first drop (of type B1) ought to be minimal in length, which is not true here, however. Second, the first drop of the second constituent (type C) is found minimal, while in independent use it tends to be longer (most frequently, two or three syllables; see section 2.6.1 above). Third, the resulting pattern brought together seems to contradict the tendency recurrently observed, namely minimisation of the verse-initial drop (before the first lift) and maximisation of the verse-internal drop (after the first lift).

7

On the status of type A1A1 as a prototype of the hypermetric verse, and a structural account for it, see below.

MOSP-04.qxd

5/11/04

302

2:42 PM

Page 302

4. Hypermetric verses and lines

A third class is constituted by type B1D: (12) Type B1D (two examples) 2290a uuið selban thene sunu drohtines 4415a thiu helpe quam te hebencuninge At first glance, scanning verse 4415a as type B1D would strike as rather implausible, because the second drop of type B1 when used independently is not filled by a string of words containing a finite verb (quam). Yet other class 2 words such as adverbs occur in the second drop of type B1 (see section 2.4.2.4 above). Moreover, in the light of the foregoing discussion, we may also understand this apparently aberrant feature of verse 4415a as another manifestation of the more prominent realisation of the drop after the first lift in hypermetric verses in general and in type B1A1 in particular.8 We recognise two types that begin with type D, each attested once as listed below in (13) and (14), respectively. (13) Type DD* (one example) 1318b thie môtun uuesan suni drohtines genemnide (14) Type DA1 (one example) 5590a uuan uuunderquâla Verse 5590a strikes us as peculiar, because the ending of the first constituent verse (type D, subtype D2b) does not correspond with the immediate constituent break of the compound word involved, which ought to be analysed as uuunder-quâla ‘monstrous torture’. Actually, this morphological constituency is respected by the verse internal break in the following verse (type A1A1): (15) 4568a uuîti endi uunderquâle On closer consideration, however, the isolation of underquâl- as the linking element would appear less peculiar (though still marked): the final syllable -a functions as an inflectional marker (acc. sg.), and as such it may be separated out from the remaining body of the word. A substantial number of verses belong to type D*A1, as given below: (16) Type D*A1 (thirteen examples) 989a uuonoda im obar them uualdandes barne 1681a lilli mid sô lioflîcu blômon 1683a liudi sint im lioboron mikilu 1796a up te themu êuuinom rîkea 8

Extending the present argument to Beowulf, one might incline to reinterpret Beo 1166a æt fotum sæt frean Scyldinga as type B1D1 (i.e., [æt fotum sæt [frean] Scyldinga], contrary to my earlier scansion (Suzuki 1996a: 363–4) based on the juxtaposition of types B1 and D1 (i.e., [æt fotum sæt][frean Scyldinga]).

MOSP-04.qxd

5/11/04

2:42 PM

Page 303

4.1. The composition of hypermetric verses 2596a 3127a 3298a 3562a 4211a 4411a 5720a 5811a 5892a

303

obana mid is engilo craftu uuânamo fan themu uualdandes barne unôði ôdagumu manne hlûdo te themu hêlagon Criste thea idisa mid is orlobu gôdu them mannun, the hêr minniston sindun darnungo uuas hie ûses drohtines iungro thiu uuîf uppan them giuuendidan stêne ahebbian be than hêlagan drohtin

Finally, there are two groups (EA1 and ED) that contain type E as the first constituent, as shown in (17) and (18): (17) Type EA1 (two examples) 2597a liudi, the io thit lioht gisâun 3066b Diurlîco scalt thu thes lôn antfâhen As pointed out by Hofmann (1991: 159), the first constituent (identified as type E) of verse 2597a would appear unparalleled because the second drop involved is realised by three syllables (the io thit). The association with three syllables in itself, however, is far from unique, as observed in 2735a heritogono an that hûs, for example, although the metrical contexts involved are not exactly the same. We may then attribute the apparently unique feature in question to the general extension of the first drop of the first constituent verse, recurrently remarked in the earlier discussion. It should be noted that the occupation by an extended string of syllables concerns the first normal drop, whereas the heavy drop is by its very nature limited to a single stressed syllable or its resolved equivalent. (18) Type ED (two examples) 901a gerno thes gramon anbusni 5723a thingon uuið thena thegan kêsures The variant of type E involved in these two verses (as well as in verse 2597a given in (17) earlier) is the least typical manifestation of type E (/ \ ⫻ /): the first drop, which is heavy (as represented by the symbol \), is realised by a word-final short unstressed syllable. Although extremely rare in occurrence, such a peripheral manner of realisation is attested by the following four examples: (19) 279a 1526b 2482b 4291b

scadouuan mid skimon ôðres ni uili nahtes endi dages (emended) dôdun endi quikun

In light of the definition of the hypermetric verse provided at the outset of this chapter, we identify as hypermetric only those verses that defy proper scansion as one of the existing normal verse types. Such a structurally strict characterisation excludes an array of verses that are treated as hypermetric by other metrists,

MOSP-04.qxd

5/11/04

304

2:42 PM

Page 304

4. Hypermetric verses and lines

Hofmann (1991) in particular. It will be in order to consider in detail these contestable verses that are open to two competing scansions. These ambiguous verses fall into two major groups according to two different non-hypermetric readings that they may receive, type D* on the one hand, and type A1 with anacrusis on the other. Furthermore, these two groups are largely complementary in distribution: the first group occurs predominantly in the a-verse, and the second one overwhelmingly in the b-verse. The first group consists of twenty-six examples, as listed in (20) below. Common to this group of verses is that they can be scanned as type D* (subtype D*2b) or hypermetric of some sort, as subclassified in (20) below. In fact, Hofmann scans all these examples as hypermetric in one way or another, as given at the end of each example. In the following list, the entire group of verses is divided into five subgroups for the sake of discussion, according to plausible hypermetric scansions, which are not always identical to Hofmann’s, however: (20) List of verses that are open to two alternative scansions, as type D* (subtype D*2b) on the one hand, and as hypermetric on the other Subgroup 1 (five examples): type D* or type A1B1 347a quâmi te them cnôsla gihue (Hofmann’s scansion: Se-A5B52v.x22.x1) 1451a frâhon is friunda gehuuane (Hofmann’s scansion: Se-A5B52.x2.x’1) 1730a that sie gehôrean iuuua hêlag uuord (Hofmann’s scansion: S2e-a/ BII32/x22.1) 3344a fan themu hêroston, the thes hûses giuueld (Hofmann’s scansion: S2e-da/B5x33a/x22.x1) 5755b Thius thiod gilôbit te filo (Hofmann’s scansion: Se-*AB5x1.x2.x’1) Subgroup 2 (one example): type D* or type A2A1s 743a fremidun firinuuerc mikil (Hofmann’s scansion: Se-AA’2v.22.23) Subgroup 3 (five examples): type D* or type D*B1 1107a hôrean the hêlago Crist (Hofmann’s scansion: Se-Da6B52.x3a.1) 2856a gaf it is iungarun forð (Hofmann’s scansion: Se-davEIIv33a.1) 3646a the man te themu mahtigon gode (Hofmann’s scansion: Se-*Da5B5x1. x33a.’1) 5420a hâhan that hêlaga barn (Hofmann’s scansion: Se-Da6B52.x3a.1) 5690a that sia thia haftun thuru thena hêlagon dag (emended) (Hofmann’s scansion: S2e-B/B5x32.1/x43a.1*) (emended) Subgroup 4 (fourteen examples): type D* or type D*E 266a thes hôhon hebancuninges suno (Hofmann’s scansion: Se-BDax2.’1:3a4*) 903a up te them alomahtigon gode (Hofmann’s scansion: Sg-Da6*E51. x2’ 4da.’1) 961a diurlîc drohtines sunu (Hofmann’s scansion: Se-Da5E2.3a.’1) 1005a diurlîc drohtines suno (Hofmann’s scansion: Se-Da5E2.3a.’1) 1110a up te them alomahtigon gode (Hofmann’s scansion: Se-Da6*E51. x2’ 4da.’1) 1144a nu is it giu ginâhid thurh thes neriandan craft (Hofmann’s scansion: S2e-a/B5II52/x23a.1)

MOSP-04.qxd

5/11/04

2:42 PM

Page 305

4.1. The composition of hypermetric verses

305

1429a that ic feldi thero forasagono uuord (Hofmann’s scansion: Se-*Da6B5 x22v.x2’3a3⫹.1) 1973a thar uppe far them alouualdan fader (Hofmann’s scansion: Sz-*Da6* Ex2.x2’3a2.’1) 3046a diurlîc drohtines bodo (Hofmann’s scansion: Se-Da5E2.3a.’1) 3994a diurlîc drohtines thegan (Hofmann’s scansion: Se-Da5E2.3a.’1) 4972a frâgodun fîundo barn (Hofmann’s scansion: Sz-Da6E3av.3a.1) 5654a hlûdo te them himiliscon fader (Hofmann’s scansion: Sz-Da6B52. x2’3a.’1) 5806a diurlîc drohtines bodo (Hofmann’s scansion: Se-Da5E2.3a.’1) 3228b hala thi thar ôðara tô (Hofmann’s scansion: Se-davEIIv43a.1) Subgroup 5 (one example): type D* or hypermetric (unclassifiable) 1348a gibidig grimmora thing (Hofmann’s scansion: Se-Da5E’2.3a.1) All the instances of subgroup 1 are readily scannable as type D* without invoking any discernible deviations from ordinary incontestable examples of this type. The association of the penultimate drop (/ ⫻ / ⫻ ⫻) with unstressed syllable sequences such as are used in the verses at issue, namely -x#x- and -x#x#, is independently attested, as exemplified in 1826a haldid hêlag gebod and 3802a uuîsis uuâres sô filu, respectively. Further support for scanning these five examples as normal verses (type D*) is provided by the fact that they all occur in isolation even according to Hofmann’s system of scansion, whereby ‘Se’ designates a hypermetric verse used in isolation (‘einzelner Schwellvers’; Hofmann 1991: 92, 156; cf. Russom 1998: 156 n. 96). The single member of subgroup 2 has to be regarded as unique in either scansion. The degrees of deviation involved, however, seem different. On the A2A1s scansion, the verse in question would be the only example that consists of the verses that are both otherwise unattested as constituents of hypermetric verses: as indicated above, there is no demonstrable composition of a hypermetric verse that involves type A2 or subtype A1s. And this will be hardly surprising when we argue below that the prototype of the hypermetric verse contains a maximally unmarked metrical type, that is, type A1. On the type D* scansion, on the other hand, the verse under consideration would constitute the only instance in which the second drop is realised by the second element of a true compound. Yet we find examples in which the same position is occupied by a heavy derivational suffix, -lîc (4751a drôp is diurlîc suêt; 5678a sâuuun seldlîc thing). Given the minimal difference in metrical terms between the second members of compounds and those of quasi compounds (see section 2.12.5 above), the existence of verses like 4751a may rather be regarded as lending plausibility to scanning the verse in question as type D*. As with subgroup 1, none of subgroup 3 verses are found anomalous in comparison with clear-cut cases of type D*, particularly in the manners of realisation of the drops and anacrusis involved (on anacrusis, see further below). Scanning them as hypermetric (type D*B1), however, will run into difficulty: for one thing, the alleged configuration D*B1 is unknown; for another, there is no definite example of a hypermetric verse that ends in type B1 in general. We may therefore be justified in concluding that all the instances of subgroup 3 constitute type D*.

MOSP-04.qxd

5/11/04

306

2:42 PM

Page 306

4. Hypermetric verses and lines

Subgroup 4 may fall into two classes according to the composition of the second drop. The first class, which comprises verses 266a, 903a, 1110a, 1429a, and 1973a, may more plausibly be identified as hypermetric in the absence of definite examples of type D* in which the second drop is realised by -Sx(x)#. However, since the same position is capable of being occupied by -xx# and -Xx# in type D* in its independent use (e.g., 946a forlâtad lêðaro drôm; 414a Afhôbun thô hêlagna sang), there seems no compelling reason for not scanning the other nine verses (961a, 1005a, 1144a, 3046a, 3994a, 4972a, 5654a, 5806a, 3228b) as type D*.9 Moreover, these remaining verses do not seem to differ markedly from clear instances of type D* in terms of the composition of the extra drop located between the lifts: up to four weak syllables are attested in it (section 2.8). Of related significance, these nine examples all occur in isolation, rather than embedded in a group of hypermetric verses, in contrast to the first subgroup (266a, 903a, 1110a, 1429a, and 1973a), some members of which appear in conjunction with demonstrable hypermetric verses, as indicated by ‘Sg’ (‘Teil einer Schwellversgruppe’; Hofmann 1991: 92, 156) in Hofmann’s notation. More specifically, seven out of these nine are not paired with hypermetric verses according to Hofmann’s system of scansion, as indicated by ‘Se’; and for the remaining two verses, their corresponding b-verses, 4972b (huilikes he folkes uuâri) and 5654b (ik an thina hendi befilhu, quathie), although scanned as hypermetric by Hofmann, are indistinguishable from clear instances of type A1 with anacrusis (on this issue, see below). Since the hypermetric verse is obviously marked in opposition to the normal counterpart, it seems safer to conclude in the absence of evidence to the contrary that the nine verses in question constitute type D*, rather than hypermetric examples. Finally, on subgroup 5. One might scan the second constituent grimmora thing as type E, thereby characterising the whole verse as type D*E. Yet type E does not allow its second position (heavy drop) to be occupied by a short medial syllable -x-; in other words, the concatenation Pxx#P is unmetrical (section 2.9). Because the second constituent is irreducible to any of the well-formed normal verses attested independently in the corpus, there seems to be no other choice than to scan it as type D*, an independently occurring normal verse. Scanned this way, the verse proves to be fully conforming to the standard properties of type D*. At this point, we may note a further instance that might appear susceptible to two scansions, normal (type D) and hypermetric (type DE): (21) 5929a bi namen neriendero best This verse is unparalleled as regards the composition of the penultimate position, which is occupied by a sequence of three syllables, -Xxx#. Hofmann (1991: 220) scans this verse as an equivalent of type DE (Sg-*DaEx’1.3a⫹.1), a unique type of its own. More plausible, however, would seem the scansion as type D, subtype D2b. 9

Of related interest, Hofmann (1991) characterises as normal, rather than hypermetric, the following verse that is minimally removed from the ones under discussion, particularly in regard to the realisation of the second drop (-Xx: -agna): 414a Afhôbun thô hêlagna sang.

MOSP-04.qxd

5/11/04

2:42 PM

Page 307

4.1. The composition of hypermetric verses

307

In support of the proposed scansion as a normal verse, we may draw attention to the following comparable if not identical verses: (22) 5802a lîf langerun huîl (type D, subtype D2b) 1144a nu is it giu ginâhid thurh thes neriandan craft (type D*, subtype D*2b) The above examples have their penultimate drop associated with a string of medial and final syllables, -X/xx#. To be sure, verse 5929a contains a trisyllabic sequence, but the final two syllables constitute a disyllabic ending (-xx), which behaves more like a single syllable for metrical purposes, as pointed out above recurrently. In this light, verses 5802a and 1144a may be regarded as closest to verse 5929a and accordingly lend plausibility to the characterisation of it as a normal verse. We now turn to consider the other major group of verse that appears ambiguous in scansion, hypermetric or normal (type A1 with anacrusis). The hypermetric scansion has been long practised in earlier scholarship, most recently followed by Hofmann (1991). This group may be divided into three subgroups (listed below in (23), (25), and (26), respectively) on the basis of specific hypermetric scansions according to Hofmann, as treated individually below. The first subgroup may be broadly scanned as type A3A1 as an alternative to the characterisation as type A1 with anacrusis (cf. Hofmann 1991: 161–2): (23) Type A1 with anacrusis or type A3A1 (Hofmann’s notation: S-aA) (104 examples) 1554a the ina iu an thesoro uueroldi ne lônon 600b sô ic uuêt, that it hêlag drohtin 3499b endi uuôpit sie mid bittrun trahnun a-verse with double alliteration: 5926a; a-verse with single alliteration: 1319a, 1554a; b-verse: 557b, 558b, 559b, 560b, 561b, 599b, 600b, 601b, 603b, 604b, 605b, 881b, 898b, 901b, 902b, 990b, 991b, 993b, 1301b, 1305b, 1307b, 1308b, 1310b, 1311b, 1312b, 1314b, 1315b, 1317b, 1319b, 1321b, 1542b, 1553b, 1555b, 1681b, 1682b, 1683b, 1684b, 1686b, 1687b, 1688b, 1689b, 2208b, 2209b, 2211b, 2212b, 2213b, 2214b, 2215b, 2596b, 2614b, 2821b, 2822b, 2823b, 2824b, 2825b, 2903b, 2985b, 2988b, 2990b, 2991b, 3036b, 3037b, 3038b, 3062b, 3063b, 3064b, 3071b, 3493b, 3494b, 3495b, 3496b, 3497b, 3499b, 3502b, 3503b, 3504b, 3505b, 3507b, 3562b, 3563b, 4393b, 4394b, 4395b, 4396b, 4411b, 4413b, 4414b, 5732b, 5811b, 5916b, 5917b, 5918b, 5921b, 5922b, 5923b, 5924b, 5925b, 5929b, 5930b, 5931b, 5933b The two competing analyses are graphically distinguished by means of bracketing in the following example: (24) As type A1 with anacrusis: 3499b [endi uuôpit sie mid bittrun trahnun] As type A3A1: 3499b [endi uuôpit sie mid [bittrun] trahnun] In evaluating these alternative scansions, we need to explore their divergent empirical consequences. First, we may examine the size of anacrusis involved (or of

MOSP-04.qxd

5/11/04

2:42 PM

Page 308

308

4. Hypermetric verses and lines

its equivalent in type A3A1) through comparison with the situation observed in definite instances of type A1 with anacrusis, particularly in the b-verse, given the nearly exclusive appearance of the group in question in the b-verse. At issue here is the implication that is entailed by the anacrustic type A1 scansion in opposition to the hypermetric scansion: other things being equal, the group of verses in question will be indistinguishable from clear cases of type A1 with respect to the number of syllables realised as anacrusis. The above prediction turns out to be incorrect, however. As Table 4.4 shows, a kind of complementary distribution obtains between the two classes being compared: incontestable instances of type A1 with anacrusis tend to limit anacrusis to the first three numbers of syllables (one through three), with a single syllable being the most frequent in occurrence; by contrast, the problematic cases show marked preference for larger numbers of syllables, ranging from four to seven (cf. Hofmann 1991: 170; Russom 1998: 155). In other words, more than 60 per cent of anacrustic type A1 b-verses fall under the zone between one and three syllables; on the other hand, with rare exceptions that same area is avoided by the other group, and instead a comparably high percentage of its examples belong to a higher zone comprising five through seven syllables. Although the two classes partially overlap, it seems hardly disputable that their respective areas of concentration are clearly demarcated. Put another way, the two classes in question are distinguished as far as their core members are concerned, and it is accordingly largely on the border area (primarily the cases of four syllables) that is affected by categorial ambiguity.

Table 4.4. The varying number of anacrustic syllables in type A1 (or of the initial part before the first lift of type A3A1), compared with that of type A1 in the a-verse and the b-verse A1 or A3A1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 total

0 0 2 16 31 20 19 11 2 1 1 0 0 1

0% 0% 2% 15% 30% 19% 18% 11% 2% 1% 1% 0% 0% 1%

104 100%

A1 (a-verse)

A1 (b-verse)

503 71% 113 16% 58 8% 22 3% 6 1% 5 1% 2 less than 1% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0%

148 64 46 38 40 27 24 5 2 2 0 0 0 0

709 100%

396 100%

37% 16% 12% 10% 10% 7% 6% 1% 1% 1% 0% 0% 0% 0%

MOSP-04.qxd

5/11/04

2:42 PM

Page 309

4.1. The composition of hypermetric verses

309

The foregoing consideration then brings us to a second consequence of related interest, which is drawn from the hypermetric scansion and falsifies the anacrustic type A1 scansion. Since on this view the group of verses at issue is scanned as type A3A1, one would predict that the first constituent is essentially no different from ordinary instances of type A3. More specifically, the verse-initial part consisting of unstressed syllables in the alleged A3A1 verses should be formally analogous to the corresponding part of type A3, other things being equal. To this end, we now need to compare the number of syllables standing between the beginning of a verse and the first lift in type A3 on the one hand and in the problematic class on the other. As shown in Table 4.5, while analogous in overall shape, the distribution pattern of type A3A1 seems to be shifted by one row up: the greatest concentration comes at five rather than four syllables, and the first two syllables, rather than one, are unattested. Such an imprecise correspondence between the two groups, however, is amenable to an explanatory account. While type A3 occurs exclusively in the a-verse, the so-called type A3A1 appears almost always in the b-verse (Hofmann 1991: 162). This distributional discrepancy may be held responsible for the partial distinction at issue. As pointed out from time to time, the b-verse favours longer sequences of unstressed syllables in verse-initial position (section 2.12.3.2), as in the first drop of types B1 (section 2.4.1) and C (section 2.6.1), as well as in anacrusis (section 2.11.1). Given such recurrent preference patterns, it seems a matter of course that type A3A1 shows preference for a larger number of syllables in its initial part than does type A3. Thus, the hypermetric scansion turns out to be empirically tenable in this respect.

Table 4.5. The varying number of syllables before the lift in type A3, compared with that before the first lift of type A3A1 (or with that of anacrusis of type A1) type A3 1 2 3 4 5 6 7 8 9 10 11 12 13 14

0 7 25 134 108 75 32 11 5 0 0 0 0 0

0% 2% 6% 34% 27% 19% 8% 3% 1% 0% 0% 0% 0% 0%

total

397 100%

type A3A1 (or aA1) 0 0 2 16 31 20 19 11 2 1 1 0 0 1

0% 0% 2% 15% 30% 19% 18% 11% 2% 1% 1% 0% 0% 1%

104 100%

MOSP-04.qxd

5/11/04

2:42 PM

310

Page 310

4. Hypermetric verses and lines

A third consequence of empirical interest concerns the pairing of these two classes with a-verses. In view of the alleged strong tendency of hypermetric verses to be combined in lines (subject to re-examination later), we are interested in examining whether there is any notable difference in the pattern of pairing between the definite instances of type A1 with anacrusis on the one hand and the ambiguous instances open to the two scansions on the other. The scansion of the latter group as normal verses would imply that they are hardly distinguishable from the first group in this regard, an empirical claim that would be rejected by the alternative scansion. As Tables 4.6 and 4.7 substantiate, the two classes show converse patterns: while the vast majority of type A1 b-verses with anacrusis are paired with normal verses, the structurally ambiguous verses exhibit the opposite pattern of predominantly being linked with hypermetric verses. Reducing the problematic group to type A1 with anacrusis would accordingly seem empirically suspicious. Summarising the above arguments, the apparently ambiguous group of verses open to the normal as well as the hypermetric scansions seems more compatible with the hypermetric interpretation, first for its similarity to type A3 and conversely for its dissimilarity to type A1 with anacrusis as regards the composition of the verse-initial unstressed part, and second for its overwhelmingly frequent association with hypermetric a-verses in forming lines. We may thus be justified in concluding that the verses under discussion are hypermetric and scanned as type A3A1. Characteristic of the second group of ambiguous verses is that the first constituents involved may be analysed as type B3 (section 2.5). For the same reasons adduced earlier in section 2.5 above, we raise an objection to an alternative hypermetric Table 4.6. Pairing of anacrustic type A1 (or type A3A1) b-verses with a-verses according to the distinction between hypermetric and normal verses paired with hypermetric verses normal verses total

number

percentage

83 19

81% 19%

102a

100%

a

Since verse 1319a is in itself scannable as type A1 with anacrusis and type A3A1 alike, verse 1319b is disregarded from the statistics here

Table 4.7. Pairing of anacrustic type A1 b-verses with a-verses according to the distinction between hypermetric and normal verses paired with

number

percentage

hypermetric verses normal verses

1 395

less than 1% nearly 100%

total

396

100%

MOSP-04.qxd

5/11/04

2:42 PM

Page 311

4.1. The composition of hypermetric verses

311

scansion, followed by Hofmann (1991) among other metrists, whereby the verses begin with type E3 (cf. Hofmann 1991: 163–4). (25) Type A1 with anacrusis or type B3A1 (S-eA in Hofmann’s notation) (fourteen examples) 899b Thes uuirðid sô fagan an is môde 1541b ne rôkead, huueðar gi is ênigan thanc antfâhan 4418b sagad im that sie sculin thea dâd antgelden 899b, 900b, 989b, 1541b, 2597b, 2826b, 2987b, 2989b, 3065b, 3498b, 3501b, 3677b, 4418b, 5228b Much the same arguments as with the first group extend to the above verses as well. The initial part preceding the first lift consists of more than two syllables, with a concentration on four syllables, as indicated in Table 4.8. Type A1 verses with anacrusis, on the other hand, show marked preference for smaller numbers of syllables, a single syllable in particular, as pointed out above and reproduced in Table 4.8. The pattern characteristic of the ambivalent group under consideration, however, is found highly similar to that of type B3, which tends to fill the corresponding part with a relatively large number of syllables (ranging between four and six). Bringing the first and second arguments together, then, we may be justified in claiming that this group of verses should be scanned as hypermetric, consisting of types B3 and A1 in that order, namely type B3A1. The conclusion reached above receives added support from the third argument: as substantiated in Table 4.9, the group of ambivalent verses at issue shows a strong preference for hypermetric a-verses to form lines, a preference pattern that is more pronounced than that of type A3A1 discussed above.

Table 4.8. The varying number of anacrustic syllables in type A1 (or of the initial part before the first lift of type B3A1), compared with that of type A1 in the b-verse and that of type B3 A1 or B3A1 1 2 3 4 5 6 7 8 9 10 11 total

0 0 2 6 1 1 1 2 0 1 0

0% 0% 14% 43% 7% 7% 7% 14% 0% 7% 0%

14 100%

type A1 (b-verse)

type B3

148 64 46 38 40 27 24 5 2 2 0

0 0 0 1 5 3 0 0 0 0 0

37% 16% 12% 10% 10% 7% 6% 1% 1% 1% 0%

396 100%

0% 0% 0% 11% 56% 33% 0% 0% 0% 0% 0%

9 100%

MOSP-04.qxd

5/11/04

2:42 PM

312

Page 312

4. Hypermetric verses and lines

Table 4.9. Pairing of anacrustic type A1 (or type B3A1) b-verses with a-verses according to the distinction between hypermetric and normal verses paired with

number

percentage

hypermetric verses normal verses

12 2

86% 14%

total

14

100%

We proceed to the third group of structurally ambivalent verses, which may be scanned as normal (type A1 with anacrusis) or hypermetric (type CA1; cf. Hofmann 1991: 163). We count twenty-one relevant examples in all: (26) Type A1 with anacrusis or type CA1 (S-daA in Hofmann’s notation) (twentyone examples) 1300b quað that thie sâlige uuârin 1554b endi rômot te iuuues uualdandes rîkea 3072b Ik fargibu thi himilrîceas slutilas 988b, 1300b, 1302b, 1304b, 1306b, 1313b, 1316b, 1320b, 1554b, 1973b, 2210b, 2986b, 3072b, 3500b, 3506b, 3990b, 4392b, 4415b, 5722b, 5927b, 5928b The second lift of the first constituent verse (type C) is realised by a short unstressed medial syllable -x- in the vast majority of these verses (988b, 1300b, 1302b, 1304b, 1306b, 1313b, 1316b, 1320b, 1973b, 2210b, 2986b, 3500b, 3506b, 3990b, 4392b, 4415b, 5927b, 5928b). Verse 1973b endi for allumu is engilo crafte is noteworthy because of the apparent alliteration on allumu, which would create the erroneous alliterative pattern aax for the b-verse; in this light, the word should be scanned as part of the first drop. Arguments analogous to those provided above concerning the previous two groups may be explored here in favour of the hypermetric scansion. First, as shown in Table 4.10, relatively larger numbers of syllables are involved in verse-initial sequences of unstressed syllables in the group of verses at issue than in anacrusis in type A1, thereby throwing doubt on their formal identification. The second argument, however, appears less conclusive. The composition of the verse-initial unstressed part is found hardly identical to that of the first drop of type C in the b-verse, as substantiated in Table 4.11. While examples with one or two syllables are unattested for the problematic group, there are numerous instances of type C with a single syllable or two serving as its first drop. Furthermore, the range of distribution seems different: the group under consideration is distinguished from type C by virtue of the more compact range it is distributed over. Yet by relative weighting, the pattern characteristic of the ambiguous group has more in common with that of type C than with that of type A1 with anacrusis, which diverges from the former two by its unique preference for a single syllable. The third argument relates to the pairing with a-verses in regard to the distinction between hypermetric and normal verses. Although not as outstanding as with types A3A1 and B3A1, a marked preference for hypermetric verses is still shown by this group in the ways of association with a-verses (Table 4.12).

MOSP-04.qxd

5/11/04

2:42 PM

Page 313

4.1. The composition of hypermetric verses

313

Table 4.10. The varying number of anacrustic syllables in type A1 (or of the initial part before the first lift of type CA1), compared with that of type A1 in the b-verse A1 or CA1 1 2 3 4 5 6 7 8 9 10 11

0 0 6 3 2 5 4 1 0 0 0

total

0% 0% 29% 14% 10% 24% 19% 5% 0% 0% 0%

21 100%

A1 (b-verse) 148 64 46 38 40 27 24 5 2 2 0

37% 16% 12% 10% 10% 7% 6% 1% 1% 1% 0%

396 100%

Table 4.11. The varying number of anacrustic syllables in type A1 (or of the initial part before the first lift of type CA1), compared with that of the first drop of type C in the b-verse A1 or CA1 1 2 3 4 5 6 7 8 9 10 11 total

0 0 6 3 2 5 4 1 0 0 0

0% 0% 29% 14% 10% 24% 19% 5% 0% 0% 0%

21 100%

C (b-verse) 124 10% 258 21% 248 20% 219 18% 169 14% 97 8% 60 5% 24 2% 11 1% 10 1% 2 less than 1% 1222 100%

In summary, the group of verses open to the scansions as type A1 with anacrusis and as type CA1 seems to be implicated in a higher degree of indeterminacy in categorisation than are the first two groups treated above. Yet weighted in overall terms, the hypermetric scansion appears more convincing than the alternative. The foregoing investigations into the composition of hypermetric verses have brought to light the whole range of their structural variants, listed in Table 4.13. The first lift of the second constituent verse serves also as the second lift of the first constituent verse. This may be called the linking lift. Normally, the whole word

MOSP-04.qxd

5/11/04

2:42 PM

Page 314

314

4. Hypermetric verses and lines

Table 4.12. Pairing of anacrustic type A1 (or type CA1) b-verses with a-verses according to the distinction between hypermetric and normal verses paired with

number

percentage

hypermetric verses normal verses

16 5

76% 24%

total

21

100%

Table 4.13. Composition of hypermetric a-verses/b-verses with respect to the identity of constituent normal verses, and their occurrences in number second first

A1

A1 A3 B1 B3 C D D* E E*

126/2 3/103 7/0 0/14 0/21 1/0 13/0 1/1

B1

C

D

D*

E

E*

1 (A1⫹C)/0 1/0

2/0

0/1 5/0 2/0

that contains the linking lift serves as part of the two constituent verses. Put another way, the two constituents of a hypermetric verse share a lexical word (called a linking word) whose head (i.e., primary-stressed syllable) serves as the linking lift. Because the linking word plays a pivotal role in organising a hypermetric verse by providing a common structural basis, the entire structure of the hypermetric verse largely rests on the specific structure of the linking word involved. Since the linking word is by definition capable of serving as a lift, it is usually a content word, most typically a substantive (class 1 word; for details on the tripartite categorisation of word classes in connection with lift formation, see section 3.2.2 above). It thus ranges between monosyllabic and trisyllabic in prosodic structure: schematically represented, (x)P, (x)Px, (x)Pxx, (x)PS, (x)PSx, (x)PxS.10 The first constituent of a hypermetric verse accordingly has to end in one of these word forms. This formal restriction imposed on the verse coda works in turn as a constraint on the structure of the first constituent, thereby legitimating the following metrical types on this capacity: types A1 and A3 (ending in Px or PS), types B1 and B3 (ending in (x)P), some variants of type C (ending in Pxx or PSx), type D (ending in Pxx, PSx, or PxS), type D* (ending in (x)Pxx, (x)PSx, or (x)PxS), and type E (ending in (x)P). Yet these word structures 10

For the sake of exposition, we are disregarding here difference in syllable length; S and x are thus designed as cover symbols for S/s and X/x, respectively; (x) before P stands for a prefix, with the parentheses designating its optionality.

MOSP-04.qxd

5/11/04

2:42 PM

Page 315

4.1. The composition of hypermetric verses

315

Table 4.14. The inventory of well-formed compositions of hypermetric verses with respect to the identity of constituent normal verses second first

A

B

C

D

D*

E

E*

A B C D D* E E*

⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹

(⫹) (⫹) (⫹) (⫹) (⫹) (⫹) (⫹)

⫺ (⫹) ⫺ ⫺ ⫺ (⫹) (⫹)

⫺ ⫹ ⫺ ⫺ ⫺ ⫹ ⫹

⫹ (⫹) (⫹) (⫹) (⫹) (⫹) (⫹)

⫺ ⫺ ⫹ ⫹ ⫹ ⫺ ⫺

⫹ (⫹) (⫹) ⫹ ⫹ (⫹) (⫹)

Key: ‘⫹’ ⫽ well-formed; ‘(⫹)’ ⫽ highly marked; ‘⫺’ ⫽ ill-formed

are far from equal in use: compound words (PS, PSx, PxS) are not very common in occurrence; nor is the appearance of an unstressed prefix (x) very frequent, particularly in connection with substantives, lexical items most likely to be encountered here. Analogously circumscribed is the structure of the second constituent verse: it has to begin with one of those well-qualified words. Accordingly, the possible candidates include type A1 (beginning with P, Px, PS, Pxx, or PSx), types B1 and C (beginning with xP), type D (beginning with P), type D* (beginning with P, Px, PS, Pxx, or PSx), and type E (beginning with P, Px, Pxx, PS, or PSx). As above, compounds are relatively infrequent in occurrence, as are prefixed words. On the basis of the above formal consideration, we are now capable of providing an inventory of possible hypermetric verses, represented in Table 4.14. While the above arguments drawing on the structures of linking words may throw some light on the structural organisation of hypermetric verses, we find ourselves still far removed from a full understanding of the issue. As yet unexplained among other things is the maximal use of type A1 as either constituent (first and second), which would go so far as to prompt us to characterise it as the prototype of the constituent of the hypermetric verse, for the a-verse and the b-verse alike. In this connection, the occurrence of type A3 rather than A1 for the first member of the hypermetric b-verse may well be regarded as a consequence of the specific alliterative pattern as argued below. To provide a principled account for the issue raised above, we need to bring a wider context to bear on the consideration, namely properties of metrical types rather than of linking words taken in isolation. First, type A1 constitutes the most frequently occurring metrical type. Combining the most readily available type together thus yields type A1A1. Second, type A1 is the most varied in internal structure, largely by virtue of the wide range of language materials available to the first drop. Such structural flexibility may as well contribute to its extensive use. Third, type A1 has a reduced variant, type A3, which is nearly the only reduced type, apart from scattered occurrences of type B3. Along with this reduced variant, type A1 is capable of serving as a common structural basis for the a-verse and the b-verse. Fourth, type A1 is the only type that can be freely combined together to form a well-formed hypermetric verse: the remaining types are scarcely subject to such a coupling, as represented in

MOSP-04.qxd

5/11/04

316

2:42 PM

Page 316

4. Hypermetric verses and lines

Table 4.14. Thus, only type A1 can be used without restriction as any part of the hypermetric verse, as the first or the second constituent, as well as for the a-verse or the b-verse. Furthermore, the consistent use of the same metrical type for the whole hypermetric verse and line may have been highly appreciated as in keeping with the poet’s profound sense of balance and harmony in composition. In this way, unparalleled by other types, type A1 may be isolated as the most efficient formal basis for composing hypermetric verses, and this may be held responsible for the status of type A1 as the prototypical constituent of the hypermetric verse. Worth questioning further is, despite the prototypical status of type A1 for the hypermetric verse as a whole, why the first constituent, as opposed to the second, is amenable to a greater range of variation in choosing constituent types, as detected from Table 4.14. This question harks back to the relative difference in size between the two constituent verses, discussed above: the first constituent is relatively larger than the second in accordance with the general principle of decreasing metrical strength along the horizontal axis (the Linearity-based Metrical Strength Scale; section 1.2), as most clearly witnessed by the fact that the positions of the same metrical status are realised in less prominent form as we come closer to the end of a verse. Correspondingly, we may assume that the first constituent verse is accorded relatively greater metrical strength than is the second counterpart. This would then give access to a wider range of language materials for use, thereby making available a greater variety of metrical types to choose from. In contrast to the structural complexity characteristic of the hypermetric verse as detailed above, its alliterative patterning is strictly regulated by the same rule that governs the normal verse. Given that the linking lift figures centrally in forming a hypermetric verse, it is only natural that it is invariably endowed with corresponding prominence, namely alliteration: the linking lift has to alliterate. And it is at this point where asymmetry emerges between the a-verse and the b-verse, motivated as it is by the categorical distinction in alliterative patterning, a fundamental categorisation that permeates the organisation of the normal verse and line. We may begin with the b-verse, the unmarked opposite that categorically manifests single alliteration. Since the first lift alone is involved in alliteration in the b-verse, the linking lift may not be preceded by another lift; otherwise, the linking lift would be deprived of its qualification for obligatory alliteration (cf. Hofmann 1991: 163). Therefore, the first constituent of the hypermetric b-verse may not contain more than one lift: the single lift involved constitutes the linking lift; accordingly, only the configurations that contain one lift are found eligible for serving as the first member of a hypermetric b-verse. On this formal criterion, types A3 (⫻ ⫻ / ⫻) and B3 (⫻ ⫻ ⫻ /) are legitimated as qualified constituents. In addition, the type C variants with the second lift occupied by a non-primary-stressed syllable may be acceptable, because only one primary-stressed syllable occurs in them, which is chosen as the linking lift; in the absence of another lifted word in the preceding positions, these type C variants may be used as the first constituent of a hypermetric b-verse. As a result, the hypermetric verse may contain two (if beginning with types A3 or B3) or three lifts (if beginning with type C), consisting of six metrical positions in all in either variety.

MOSP-04.qxd

5/11/04

2:42 PM

Page 317

4.1. The composition of hypermetric verses

317

In this connection, we should draw attention to the following four verses (cf. Rieger 1876: 57–8; Sievers 1893: §122.2; Heusler 1956: §140; Hofmann 1991: 164–5): (27) 1308b 1312b 1314b 2825b

Sâlige sind ôc, the sie hîr frumono gilustid Sâlige sind ôc them hîr mildi uuirðit Sâlige sind ôc undar thesaro managon thiodu Nâh sind hêr gesetana burgi

These verses would appear exceptional: the verse-initial words involved, sâlige ‘blessed’ and nâh ‘near’, obviously class 1 words, do not alliterate. Unlike man, these words do not seem so highly grammaticalised as to be susceptible to metrical demotion, which would make it possible for them to be used as part of anacrusis or the verseinitial drop. Furthermore, the sharing by these four verses of the same syntactic construction adjective ⫹ copula would have been no coincidence. On the basis of these observations, rather than invoking grammaticalisation of an individual lexical item as with man, we would attribute these seemingly exceptional examples to the interplay of two forces, metrical on the one hand and syntactic on the other. The metrical force involved is the canonical structure of the hypermetric b-verse in which alliteration takes place in verse-medial position, while the syntactic counterpart resides in the construction of adjective ⫹ copula in the main clause. The proposed explanation has in support the following three facts. First, generally type A3A1 a-verses do not contain comparable lexical words before the first lift; their initial part consists solely of weak-stressed items, that is, sentence particles and the like. Second, and closely related to the first point, when a comparable word (a trisyllabic non-compound adjective) occurs verse-initially in a similar syntactic construction in the a-verse, it bears alliteration, serving thereby as the first lift of type A1A1, as exemplified in 3502a sô êgrohtful is, the thar alles geuueldid. Third, when the adjective ⫹ copula construction appears in the subordinate clause introduced by quað that ‘he said that’ and the like, alliteration falls on the predicate adjective, which now stands in verse-internal position. As observed by Berron (1940: 12), sâlige bears alliteration in the same passage of the Beatitudes when used in indirect speech, as in 1304b, 1306b, 1316b, and 1320b Quað that ôc sâlige uuârin. Thus, type A3A1 with the adjective ⫹ copula construction appears in the b-verse, but not in the a-verse, whereas type A1A1 with the comparable construction occurs in the a-verse, but not in the b-verse. Such a complementary distribution between type A3A1 and type A1A1 finds a credible account when we claim that the b-verse in contrast to the a-verse is strictly regulated by the canonical placement of the sole alliterative element in verse-internal position. Because class 3 words are unqualified for a lift, only they are allowed to appear before the first lift of type A3A1 in the a-verse; by contrast, content words – under which adjectives are subsumable – have to be invariably alliterative (see further below), and accordingly they are compatible exclusively with the alliterative pattern aax, which is totally unacceptable for type A3A1 in the a-verse. Furthermore, serving as predicate rather than as modifier of a noun, the adjectives under discussion (27) occur in the contexts that would have motivated them for not fully realising their lexical salience characteristic of class 1 words (substantives; cf. Suzuki 1996a: 288). Put another way, these adjectives functioning as predicate would have come closest to being treated as finite verbs (Hofmann 1991: 165), that is,

MOSP-04.qxd

5/11/04

318

2:42 PM

Page 318

4. Hypermetric verses and lines

class 2 words, which may give way to class 1 words in carrying alliteration (section 3.2.2). By virtue of this syntactically based demotion, these adjectives are made exempt from manifesting alliteration otherwise expected, and as a consequence the canonical alliteration xax is allowed to occur in keeping with the rule. To be contrasted with these four verses with adjectival predicates in this regard are the two type A1A1 verses already mentioned, 3067b hugiskefti sind thîne stêne gelîca and 1685b god uuili is alles râdan. In these verses, the verse-initial lexical words are both nouns, serving as subject, and therewith realising their inherent lexical potential to the full. Defying being treated as equivalent to finite verbs (class 2 words), these two words have to realise alliteration, thereby resulting the pattern axx, in violation of the canonical alliterative pattern xax for the b-verse. Turning now to the a-verse: in default of the special requirement for single alliteration, the first constituent verses for hypermetric a-verses are chosen from the metrical types that are unmarked with respect to the number of lifts contained; accordingly, types A3 and B3 are normally excluded. We are thus left with types A1, B1, D, D*, and E. Consequently, the hypermetric a-verse in its entirety contains three lifts, the middle one serving as the linkage. As for alliterative patterning, double alliteration constitutes the norm, as naturally expected. For one thing, even for the normal verse, double alliteration is regarded as an unmarked option for the a-verse (section 3.2.1.1). For another, even as heavy verses, provided with three lexical words inherently qualified for alliteration, typically manifest double alliteration (section 3.2.2), hypermetric verses with a comparable or greater complexity in composition are expected to realise double alliteration to a fuller extent. The extensive implementation of double alliteration for hypermetric a-verses thus accords with the expectation. Yet being extensive should not be equal to being obligatory. Mandatory double alliteration for the hypermetric a-verse is thus in need of explanation. Moreover, two patterns of alliteration would be logically possible, aax and xaa, both conforming to the obligatory alliteration on the linking lift. The actually occurring pattern, however, is limited to aax, and we are required to account for this limitation as well. Addressing the second question first, we may regard the selection of aax over xaa as a logical consequence of the alliteration rule that applies to the normal verse: this rule concerns the first and second lifts of the verse, simply because there is no further lift involved in the normal verse. The third lift, available only in the hypermetric verse, falls outside the purview of the alliterative rule in question, and accordingly a formal basis is lacking in the existing rule for making alliteration accessible to the third lift. Concerning the first question of obligatory operation of double alliteration, we are reminded that the principle of left dominance was postulated to explain the alliterative pattern of the normal verse in section 3.2.2 above. This principle stipulates that the first lift or the leftmost one in a verse has to alliterate. Accordingly, given the lift concatenation A B C in the hypermetric a-verse (where B, the second lift, links the two constituent verses), the first lift, A, must always alliterate. Even when A ranks lower than B in the lexical hierarchy, it is required to alliterate. Otherwise, it would fail to be unambiguously identifiable as the first lift, and would be in danger of being misinterpreted as a part of a drop. Hence, the resulting verse, containing only two demonstrable lifts, would deviate from the prototypical organisation of the hypermetric a-verse, and might be scanned as non-hypermetric. The principle of left

MOSP-04.qxd

5/11/04

2:42 PM

Page 319

4.1. The composition of hypermetric verses

319

dominance and the privileged status of the second lift, the linking lift, thus conspire to demand obligatory occurrence of double alliteration on the hypermetric a-verse. In evaluating the competing scansions of the so-called hypermetric b-verses above, we were implicitly assuming that they were distinguishable on independent grounds from structurally similar, normal b-verses, notably type A1 with anacrusis. More specifically, the comparison between type A1 with anacrusis and the ambivalent verses (type A1 with anacrusis or hypermetric) with regard to the varying number of verse-initial unstressed syllables as summarised in Tables 4.4, 4.8, and 4.10 might mislead us into believing that structural ambivalence concerns only the class of the so-called hypermetric b-verses, and that the other class, that of type A1 with anacrusis, was clearly bounded, exempt from ambiguity in scansion, and thus capable of serving as a firm basis of comparison. This assumption does not hold good, however. To be sure, reference to the class of type A1 a-verses with anacrusis, which is well-defined with clear-cut boundaries, may provide some justification, particularly for those instances with a single syllable at the beginning. Yet the two classes, type A1 a-verses with anacrusis and type A1 b-verses with anacrusis, are far from identical concerning the very property at issue, as shown in Table 4.4, and we have to admit therefore that the argument based on analogy with the a-verses is of limited force. Increasingly uncertain in categorisation become groups of verses with relatively long anacrustic elements, such as are exemplified below: (28) 1718b sîðor mag hi mid is lêrun uuerðan (with six anacrustic syllables; Hofmann’s scansion: *Ax62.2) 3611b sâtun im thô bi theru drohtines strâtun (with seven anacrustic syllables; Hofmann’s scansion: *A5x73a.2) 5555b quâðun that hie im sô an is uuilleon sprâki (with eight anacrustic syllables; Hofmann’s scansion: *Ax82.2) Given the above verses and the like, we may reasonably ask whether we have at our disposal any principled way of telling that they are type A1 with anacrusis as were assumed in the above discussion, rather than hypermetric verses (type A3A1). The answer is in the negative, as Hofmann (1991: 169–70) states categorically. There seems to be no formal basis for distinction; there exist no structural properties internal to the verses in question that would keep us from scanning them as hypermetric (type A3A1), as equivalent to verses like 3036b Thar gifragn ik that he is gesîðos grôtte (Hofmann’s scansion: Sg-aAII82.2). This would mean that the boundary between hypermetric b-verses and type A1 with anacrusis is inherently fuzzy, defying sharp dichotomisation. Thus, the two classes in question merge at their boundary areas, and as a consequence structural ambiguity resides in the intersection of the two classes: both classes may be reducible to each other particularly as regards their periphery members, such as verse 601b Uui gisâhun morgno gihuilikes with four initial unstressed syllables, in addition to those listed in (28) above; by contrast, the central members of their respective categories may be distinguished with a reasonable amount of credibility on probabilistic grounds on the basis of their prototypical properties (i.e., the number of verse-initial unstressed syllables). In this way, we can hardly fail to recognise that a fundamental asymmetry exists between hypermetric a-verses and b-verses: the hypermetric b-verses in general,

MOSP-04.qxd

5/11/04

320

2:42 PM

Page 320

4. Hypermetric verses and lines

in contrast to hypermetric a-verses, cannot be identified as such on their own structural terms alone through clear differentiation from anacrustic type A1 verses. The hypermetric a-verses, on the other hand, are identifiable as such independently on definitive grounds, that is, with exclusive reference to their own structure. The hypermetric b-verses may indeed be suspected on the basis of their own structural properties, but the final identification becomes viable only after making verse-external considerations, namely the a-verse with which the verse being examined is paired, and more significantly a group of adjacent verses and lines, as pointed out by Hofmann (1991: 170–1). Accordingly, while the identification of hypermetric a-verses is absolute and categorical, determined unequivocally by their structural properties, that of hypermetric b-verses is relative and probabilistic, and thus implicated in inherent uncertainty and indeterminacy in scansion. The foregoing consideration then induces us to re-examine in a new light the verse-external contexts that have potential bearing on the identification of hypermetric b-verses. Since hypermetric a-verses are sharply delineated on structural grounds as remarked above, we may isolate under strict control the category of hypermetric b-verses on the evidence of their pairing with demonstrable hypermetric a-verses. Such a reconsideration is all the more needed, because, following the rigid criterion of structural irreducibility to normal verses, my scansions of hypermetric a-verses are narrower in scope than Hofmann’s, which provided the initial database for investigation. In the absence of its own defining properties, the class of hypermetric b-verses with inherent fuzzy boundaries should be susceptible to reconfiguration in accordance with the recategorisation of hypermetric a-verses proposed above. The following list contains all examples of hypermetric a-verses that we have identified above, and a collection of hypermetric b-verses that are taken with minor change from Hofmann (1991: 182) on the other: (29) Distribution of hypermetric a-verses and b-verses (those occurring in isolation are underlined) 266a 557a, 557b; 558a, 558b; 559a, 559b; 560a, 560b; 561b 599b; 600a, 600b; 601a, 601b; 602a; 603a, 603b; 604a, 604b; 605a, 605b 881a, 881b 898b; 899a, 899b; 900a, 900b; 901a, 901b; 902b; 903a 954a 988b; 989a, 989b; 990a, 990b; 991a, 991b; 992a; 993a, 993b 1096a 1110a 1271a 1300b; 1301a, 1301b; 1302b 1304b; 1305b; 1306a, 1306b; 1307a, 1307b; 1308a, 1308b; 1309a; 1310a, 1310b; 1311a, 1311b; 1312a, 1312b; 1313a, 1313b; 1314a, 1314b; 1315a, 1315b; 1316a, 1316b; 1317a, 1317b; 1318a, 1318b; 1319a, 1319b; 1320a, 1320b; 1321a, 1321b 1429a 1512a 1541b; 1542a, 1542b 1553b; 1554a, 1554b; 1555b; 1556a

MOSP-04.qxd

5/11/04

2:42 PM

Page 321

4.1. The composition of hypermetric verses

321

1561a 1667a 1681a, 1681b; 1682a, 1682b; 1683a, 1683b; 1684a, 1684b; 1685a, 1685b; 1686a, 1686b; 1687a, 1687b; 1688a, 1688b; 1689a, 1689b 1796a 1833a 1973a, 1973b 2208a, 2208b; 2209a, 2209b; 2210a, 2210b; 2211a, 2211b; 2212a, 2212b; 2213a, 2213b; 2214a, 2214b; 2215a, 2215b 2290a 2358a 2596a, 2596b; 2597a, 2597b 2614a, 2614b; 2615a 2821b; 2822a, 2822b; 2823a, 2823b; 2824a, 2824b; 2825a, 2825b; 2826b 2903a, 2903b; 2904a 2985a, 2985b; 2986b; 2987a, 2987b; 2988a, 2988b; 2989a, 2989b; 2990a, 2990b; 2991b 3036b; 3037a, 3037b; 3038a, 3038b 3062b; 3063a, 3063b; 3064b; 3065a, 3065b; 3066a, 3066b; 3067a, 3067b; 3068a 3071b; 3072a, 3072b 3125a 3127a 3241a 3298a 3493a, 3493b; 3494a, 3494b; 3495a, 3495b; 3496a, 3496b; 3497a, 3497b; 3498a, 3498b; 3499a, 3499b; 3500a, 3500b; 3501a, 3501b; 3502a, 3502b; 3503a, 3503b; 3504b; 3505a, 3505b; 3506a, 3506b; 3507b 3560a 3562a, 3562b; 3563b 3677a, 3677b 3971a 3990a, 3990b 4211a 4251a 4265a 4392a, 4392b; 4393a, 4393b; 4394a, 4394b; 4395a, 4395b; 4396a, 4396b 4411a, 4411b 4413a, 4413b; 4414b; 4415a, 4415b 4418a, 4418b 4568a 4986a 5228a, 5228b 5354a 5379a 5551a 5590a 5609a

MOSP-04.qxd

5/11/04

2:42 PM

Page 322

322

4. Hypermetric verses and lines 5663a 5720a 5722a, 5722b; 5723a 5732a, 5732b 5811a, 5811b; 5812a; 5813a 5892a 5916a, 5916b; 5917a, 5917b; 5918a, 5918b; 5919a 5921a, 5921b; 5922a, 5922b; 5923a, 5923b; 5924a, 5924b; 5925a, 5925b; 5926a; 5927a, 5927b; 5928a, 5928b; 5929b; 5930a, 5930b; 5931a, 5931b; 5932a; 5933a, 5933b 5975a

The above list might induce us to generalise that hypermetric b-verses occur in groups, in contrast to hypermetric a-verses, which are frequently used in isolation; and accordingly that hypermetric b-verses are immune to independent use, as seems to be testified by the absence of such examples. However, we should not be misled into thinking that this apparent asymmetry between the hypermetric a-verse and b-verse is a real problem to be solved. On the contrary, the alleged asymmetry derives from our characterisation of hypermetric b-verses given above and recapitulated in the following paragraph. Thus, the condition of co-occurrence with demonstrable hypermetric a-verses simply crept into the identification of hypermetric b-verses above. A given verse is singled out as a candidate for a hypermetric b-verse if it is scannable as types A3A1, B3A1, or CA1, as discussed earlier. The candidate then has to satisfy one of the following two verse-external criteria to be ultimately identified as hypermetric. First, the candidate verse is paired with a hypermetric a-verse to form a line. Second, even if not directly matched with a demonstrable hypermetric a-verse, the candidate b-verse precedes or follows without a break in terms of line sequencing a minimal cluster of hypermetric verses. In lieu of a better definition, the minimal cluster of hypermetric verses may be defined here, partly drawing on Hofmann’s (1991: 171) insight, as a pair of hypermetric verses that are found in the same line or in two consecutive lines. Accordingly, given five consecutive lines n⫺2, n⫺1, n, n⫹1, n⫹2 (n ⫽ any line number), each consisting of two verses ((n⫺2)a, (n⫺2)b; (n⫺1)a, (n⫺1)b; na, nb; (n⫹1)a, (n⫹1)b; (n⫹2)a, (n⫹2)b), and nb being questioned as to its hypermetricality, the verse nb is identified as hypermetric if it appears adjacent to one of the following four minimal clusters of hypermetric verses. Table 4.15. The minimal clusters of hypermetric verses with the candidate verse nb as a point of reference line verse

n⫺2 a ⫹

b

n⫺1 a

b

⫹ ⫹

⫹

n a

b

n⫹1 a

⫹ ⫹ Key: ‘⫹’ ⫽ hypermetric verse

b

n⫹2 a

⫹ ⫹

b

MOSP-04.qxd

5/11/04

2:42 PM

Page 323

4.2. Diversification and restructuring of the verses

323

Thus feeding on the first condition, the second one determines a given candidate b-verse as hypermetric if it stands in immediate adjacency in terms of line sequencing to a minimal cluster of hypermetric verses. On the basis of these two conditions, then, we may distinguish probable hypermetric b-verses from questionable ones, which, occurring in isolation, fail to meet either of the above necessary conditions. For the sake of clarity, the isolated occurrences are underlined in the above list. We should recall, however, that hypermetric a-verses are identifiable as such on their own formal grounds. Accordingly, the hypermetric a-verses occurring in isolation are nonetheless hypermetric beyond doubt; by virtue of single appearance, however, these instances have to be regarded as less prototypical examples. As far as hypermetric b-verses are concerned, the foregoing re-examination requires us to add to the above list (29) the following three verses (printed in boldface) that were treated as non-hypermetric in the previous discussion. The three verses in question, however, all exhibit structural resemblance to hypermetric verses by virtue of an extended number of verse-initial syllables (more than four syllables), and most significantly they are incorporated in an existing block of hypermetric verses through adjunction to a minimal cluster of well-established hypermetric verses, as represented below: (30) 897b huuô sea sculin iro gilôbon haldan 897b; 898b; 899a, 899b; 900a, 900b; 901a, 901b; 902b; 903a 2820b Thô gengun is gesîðos tuelibi 2820b; 2821b; 2822a, 2822b; 2823a, 2823b; 2824a, 2824b; 2825a, 2825b; 2826b 5920’b êr than hie ina cûðian uuelda 5920’b; 5921a, 5921b; 5922a, 5922b; 5923a, 5923b; 5924a, 5924b; 5925a, 5925b; 5926a; 5927a, 5927b; 5928a, 5928b; 5929b; 5930a, 5930b; 5931a, 5931b; 5932a; 5933a, 5933b We thus count a total of 305 definitive hypermetric verses, 162 a-verses and 143 b-verses. This means that less than 3 per cent of all verses in the corpus (11448 verses) are hypermetric. Of these, thirty-three (as underlined in the above list) or slightly over one out of every ten are used in isolation. These isolated occurrences are all a-verses, as expected, because hypermetric b-verses, lacking their own defining properties, have to occur by definition in combination with another hypermetric verse or two, to say the least.

4.2. Diversification and restructuring of hypermetric verses: a diachronic perspective The structure and distribution of hypermetric verses in the Heliand that we investigated from a purely synchronic perspective in the preceding section needs now to be placed in the wider context of the Old English metrical tradition as epitomised in Beowulf. In doing so, we expect to identify distinctive characteristics of the hypermetric verse in the Heliand metre and explore the motivations and sources of innovation. To this end, we may draw on Suzuki’s (1996a) scansions of hypermetric verses in

MOSP-04.qxd

5/11/04

2:42 PM

Page 324

324

4. Hypermetric verses and lines

Beowulf for descriptive concerns on the one hand and his analytical framework for theoretical considerations on the other. In comparing between the Heliand and Beowulf as regards the composition of hypermetric verses, we are struck first of all by the excessive rarity of hypermetric verses in Beowulf:11 hypermetric verses account for only about 0.3 per cent of all verses in Beowulf, while the proportion rises to 2.6 per cent in the Heliand. And all but one example appear in groups in Beowulf, the sole exception being Beo 2297a ealne utanweardne. It should be noted that this isolated verse is minimally removed from the normal verse (heavy verse): the third lift is realised by the second element of a compound. Such an atypical feature may have made the exceptionality concerned less pronounced. It may be warranted to conclude then that hypermetric verses are under strict control in Beowulf and generally may not occur singly. This nearly exceptionless avoidance of using hypermetric verses in isolation in Beowulf stands in marked contrast to the liberal use of single occurrences in the Heliand, as detailed in section 4.1 above. Of the small set of hypermetric verses in Beowulf, type A1A1 and type A3A1 occur predominantly in the a-verse and the b-verse, respectively, the distribution patterns common to both traditions, as shown in Tables 4.16 and 4.17. The apparently similar concentrations on the two types in question, however, differ in points of detail in the two works. First, considered in overall terms, the degree of concentration on the two types is greater in Beowulf (86 per cent) than in the Heliand Table 4.16. Distribution of two major hypermetric verse types a-verse

b-verse

total

type A1A1 type A3A1

126 3

78% 2%

2 104

1% 73%

128 107

42% 35%

subtotal

129

80%

106

74%

235

77%

total of hypermetric verses

162 100%

143 100%

305 100%

Table 4.17. Distribution of two major hypermetric verse types in Beowulf a-verse

b-verse

total

type A1A1 type A3A1

9 0

82% 0%

0 10

0% 91%

9 10

41% 45%

subtotal

9

82%

10

91%

19

86%

total of hypermetric verses

11

11 100%

11 100%

22 100%

List of hypermetric verses in Beowulf (Suzuki 1996a: 360) A1A1: 1163a, 1164a, 1165a, 1705a, 1706a, 1707a, 2297a, 2995a, 2996a; A2aA1: 1168a; B1D1: 1166a; A3A1: 1163b, 1164b, 1165b (A3A1s), 1166b, 1167b, 1705b, 1706b, 1707b, 2995b, 2996b; B3D1: 1168b.

MOSP-04.qxd

5/11/04

2:42 PM

Page 325

4.2. Diversification and restructuring of the verses

325

(77 per cent). Second, seen relative to the distinction between the a-verse and the b-verse, Beowulf shows stronger preference for each favourite type, types A1A1 (a-verse) and A3A1 (b-verse); especially outstanding is the concentration on type A3A1 in the b-verse, which accounts for more than 90 per cent of all hypermetric b-verses in the corpus. Third, Beowulf is distinguished from the Heliand by the perfect complementary distribution of types A1A1 and A3A1 in correlation to the verse distinction: type A1A1 occurs exclusively in the a-verse, while type A3A1 is limited to the b-verse in occurrence. By contrast, the Heliand allows both types A1A1 and A3A1 to occur in the b-verse and the a-verse, respectively, although their distribution comes closest to being complementary. The first and second points raised above bring us in turn to the use of other types of hypermetric verses. In Beowulf we find a single instance each of types A2aA1 and B1D1 in the a-verse, and of a questionable configuration, possibly type B3D1, in the b-verse. In other words, Beowulf uses only three other types, one instance for each, besides the two favourite types, A1A1 and A3A1. By contrast, the Heliand uses a far wider spectrum of hypermetric verse types, as treated in detail in the previous section. Two of these three minor types found in Beowulf are attested in the more inclusive range of hypermetric verses in the Heliand: type A2aA1 (Beo 1168a arfæst æt ecga gelacum) is integrated as a variant of type A1A1, in response to the reduction of the form PS# to PX# (section 2.1.7.2); and type B1D1, exemplified by Beo 1166a æt fotum sæt frean Scyldinga, remains as such (B1D) in the Heliand metre as well. Finally, Beo 1168b Spræc ða ides Scyldinga poses some difficulty. While the second constituent (ides Scyldinga) unquestionably scans as type D1, the first counterpart (Spræc ða ides) apparently lacks one metrical position, whether scanned as type A3, type B3, or type E3. In the absence of conclusive evidence, we might tentatively scan this as type B3D1 on the strength of the argument adduced in section 2.5 above. In any event, this verse finds no parallel in the Heliand, in which all hypermetric b-verses end in type A1. The existence of the hypermetric verse ending in type D1 in Beowulf shows accordingly that in this respect the Heliand is more restrictive in the composition of hypermetric verses: it completely levels out variation on the second constituent of hypermetric b-verses. Apart from the above noted three additional configurations, Beowulf offers no instances of the minor configurations attested in the Heliand. Of particular instance is the complete absence in Beowulf of even types B1A1 and D*A1, which occur relatively frequently in the Heliand. The extremely narrow range of hypermetric verses allowed in Beowulf, virtually limited to the two types (types A1A1 and A3A1), led to a prototype-based formal account of the hypermetric verse, as proposed in Suzuki (1996a): the hypermetric verse constitutes an overlapped combination of two normal verses whereby the middle two positions serve simultaneously as part of the two constituent normal verses, as represented below (Suzuki 1996a: 356–64): (31) Derivation of the prototypical hypermetric verse in Beowulf first constituent / ⫻ / ⫻ / ⫻ / ⫻ second constituent hypermetric verse / ⫻ / ⫻ / ⫻

MOSP-04.qxd

5/11/04

326

2:42 PM

Page 326

4. Hypermetric verses and lines

(32) Metrical scheme for the hypermetric verse in Beowulf [/ ⫻ [/ ⫻] / ⫻] Key: / ⫽ lift; ⫻ ⫽ drop; [ ] ⫽ verse boundaries, as above Without going into detail, the verse structure invoked for conjunction, namely the configuration / ⫻ / ⫻, is one of the three metrical schemes postulated at the most abstract level of analysis: the metrical schemes are characterised as further irreducible formal foundations underlying the construction of metrical types (Suzuki 1996a: 132–4, 376–7). Since the specific mode of conjunction underlying the composition of hypermetric verses requires the identity of the overlapped positions involved, the other two metrical schemes available, / / ⫻ ⫻ and / ⫻ ⫻ /, are unqualified for it: in neither configuration are the first two positions and the last two found isomorphic to each other. Thus, only the scheme / ⫻ / ⫻ is demonstrated to be capable of providing a structural basis for the construction of hypermetric verses. Because the scheme / ⫻ / ⫻ serves as a common basis for all realisation variations of types A1, A2, and A3 to the exclusion of other metrical types, the prototypical composition of hypermetric verses may concern only variants of class A as their constituents. Verses 1166a and 1168b would then have to be treated as pure exceptions to this generalisation. Confronted with the multiplicity of hypermetric verse types encountered in the Heliand, as well as with the correspondingly weaker concentration there on types A1A1 and A3A1, we cannot expect to apply the above account to this poem without radical revision. Particularly problematic would seem the formal restriction that we imposed on the manner of conjunction of constituent verses in Beowulf: the sharing of the middle two positions by the two constituents and the implied identity of these overlapped positions in terms of the lift versus drop distinction. In the Heliand, the number of positions involved in overlapping is variable, ranging between one and three: e.g., B1 (⫻ / ⫻ /) A1 (/ ⫻ / ⫻); D* (/ ⫻ / ⫻ ⫻) A1 (/ ⫻ / ⫻). Moreover, the number of positions serving as a link is not always identical between the two constituent verses, as with D* (/ ⫻ / ⫻ ⫻) A (/ ⫻ / ⫻) for type D*A, and D (/ / ⫻ ⫻) D* (/ ⫻ / ⫻ ⫻) for type DD*, with the resulting mismatch in terms of the quantity of metrical positions. The strict formal restriction on the composition of hypermetric verses that we devised for Beowulf would accordingly have to be relaxed for the Heliand. In the light of the variability of linking elements in the hypermetric verses in the Heliand, we should best assume that language materials, rather than metrical positions, serve as a link between two constituent verses of a hypermetric verse, as proposed in section 4.1 above. The new analysis shown viable for the Heliand is thus predicated on the sharing of a lexical item (linking word), rather than metrical positions, by the two constituent verses. The restructuring of hypermetric verses on the basis of linking words would have been favoured by the metrical revaluation of an array of words including PS#, Psx#, PSx#, and PXx# as a sequence of lift and drop (i.e., / ⫻) for the composition of type A1, the most frequently used metrical configuration. Earlier in the traditional metre, these words were endowed with the invariable metrical values / \ (PS, Psx) and / \ ⫻ (PSx, PXx), respectively, and as such they were excluded from use as part of type A1. For the reasons detailed in section 2.1.7 above, however, the word forms in question came to acquire an additional value / ⫻, on top of the traditional

MOSP-04.qxd

5/11/04

2:42 PM

Page 327

4.2. Diversification and restructuring of the verses

327

one that they still retained. Accordingly, the concatenations PSx and PXx (the location of word boundaries being immaterial here) became ambivalent, bipositional (/ ⫻) or tripositional (/ \ ⫻), depending on context. The one-to-one correspondence between metrical configuration and syllable sequence was disrupted as a result. Legitimated now as part of type A1, these metrically ambivalent syllable sequences made themselves prominent in the composition of hypermetric verses, for which type A1 played a key role as a prototypical constituent. By virtue of their contextsensitive variability in metrical value, the trisyllabic sequences in question proved to be incongruent with the traditional scheme for the prototypical hypermetric verse with the fixed number of metrical positions. This structural incongruity would in turn have encouraged the poet to restructure the formal basis of hypermetric verses by giving priority to the specific syllable concatenations realised by concrete word forms over the abstract configurations of metrical positions. The notion of the linking word accordingly came to figure centrally in reorganising the hypermetric verse in the Heliand. Although types A1A1 and A3A1 remain in use as typical hypermetric verses in the Heliand, they are not identical to their counterparts in Beowulf in manners of their realisation, particularly with regard to their length when used as the first constituent verses. Specifically, the two most favourite types tend to be much longer than their corresponding types in Beowulf primarily by virtue of the longer sequences of the first drop of types A1 and A3, respectively, as indicated in Tables 4.18 and 4.19. We may find it hardly surprising for the first drop to show such marked preference for longer strings of unstressed syllables in the Heliand than in Beowulf, when we are reminded of the generally extended realisations of the same position in independent use (sections 2.1.5 and 2.3). It is noteworthy in this connection that the length of the second drop shows no significant difference between the two works.

Table 4.18. Distribution of type A1A1 according to the number of syllables in the first and second drops in Beowulf, compared with that in the Heliand Beowulf first drop

second drop

1 2 3 4 5 6 7 8 9 10

5 3 1 0 0 0 0 0 0 0

7 2 0 0 0 0 0 0 0 0

total

9 100%

56% 33% 11% 0% 0% 0% 0% 0% 0% 0%

78% 22% 0% 0% 0% 0% 0% 0% 0% 0%

9 100%

Heliand first drop 20 16% 29 23% 28 22% 29 23% 10 8% 5 4% 4 3% 1 less than 1% 1 less than 1% 1 less than 1% 128 100%

second drop 94 34 0 0 0 0 0 0 0 0

73% 27% 0% 0% 0% 0% 0% 0% 0% 0%

128 100%

MOSP-04.qxd

5/11/04

2:42 PM

Page 328

328

4. Hypermetric verses and lines

Table 4.19. Distribution of type A3A1 according to the number of verse-initial unstressed syllables in Beowulf, compared with that in the Heliand Beowulf 1 2 3 4 5 6 7 8 9 10 11 12 13 14 total

0 1 3 3 3 0 0 0 0 0 0 0 0 0

0% 10% 30% 30% 30% 0% 0% 0% 0% 0% 0% 0% 0% 0%

10 100%

Heliand 0 0% 0 0% 2 2% 16 15% 33 31% 20 19% 20 19% 11 10% 2 2% 1 less than 1% 1 less than 1% 0 0% 0 0% 1 less than 1% 107 100%

Obviously inseparable from the lesser extension of the first drop in Beowulf is the minimal use of the heavy counterpart of type A1, namely type A2a, as the first constituent of a hypermetric verse: the sole example attested is Beo 1168a arfæst æt ecga gelacum. By contrast, comparable forms are not very rare in the Heliand (e.g., 2215a uuârsagon an thero uueroldes rîki; 1317a, 2614a, 2822a, 3067b). A further difference in realisations of type A1A1 concerns use of anacrusis. None of the nine examples of type A1A1 in Beowulf are provided with anacrusis; in fact, there is not a single instance of a hypermetric verse with anacrusis in the whole corpus of Beowulf. In the Heliand, on the other hand, forty-nine out of 128 instances of type A1A1, c. 38 per cent, contain anacrusis. Such frequent occurrences of anacrusis in hypermetric verses in the Heliand may be ascribed to the extensive use of anacrusis and the resulting reconfiguration of it as an optional but full metrical position in the Heliand metre (section 2.11). In the Old English metrical tradition, on the other hand, anacrusis constitutes a derived entity on the surface, arising as it does through association with particular language materials under rigid conditions (Suzuki 1996a: 315–40). Although there seems no principled reason for having to avoid anacrustic hypermetric verses, anacrusis in itself is a highly marked phenomenon: anacrusis occurs in eighty-nine out of 6364 verses, c. 1 per cent. A combination of unusual features (hypermetricality and anacrusis) would accordingly have to be rare in the extreme. By far the most significant difference between the two traditions, however, seems to lie in the status of hypermetric b-verses, type A3A1 in particular. In the Old English tradition, this type is identifiable on its own formal terms with exclusive reference to itself, the defining property lacking in the same type in the Heliand

MOSP-04.qxd

5/11/04

2:42 PM

Page 329

4.2. Diversification and restructuring of the verses

329

metre. In it, type A3A1, along with other hypermetric types limited to the b-verse, can be differentiated from normal verses with anacrusis only in relative terms on probabilistic grounds by essential reference to adjacent verses. In other words, in the Heliand the opposition between normal and hypermetric verses is partly neutralised in the b-verse, while it is clearly demarcated in the a-verse. By contrast, in Beowulf the category of type A3A1 is clearly bounded, hardly susceptible to ambiguity in scansion. Anacrusis and type A3A1 are largely in complementary distribution in the Old English tradition, and perfectly so in Beowulf. Appearance of unstressed syllables preceding the metrical configuration / ⫻ / ⫻, therefore, would scarcely cause difficulty in categorisation. Furthermore, the number of initial unstressed syllables occurring in type A3A1 in Beowulf largely falls under the range between three and five, with a single exception, as substantiated above. Anacrusis, however, is disyllabic at most, with a vast majority being instantiated by monosyllabic prefixes. Accordingly, the range of unstressed syllables used is distinguished discretely between the two categories, apart from a few marginal cases. Moreover, we may assume that anacrusis would have been hardly compatible with hypermetric verses in the absence of anacrustic hypermetric verses in the corpus. These two categories are thus not only unambiguously distinguishable but rarely intersect in distribution. The structural indeterminacy that inheres in type A3A1 and other hypermetric types limited to the b-verse in the Heliand would presumably have stemmed from the reorganisation of anacrusis in the metre that led to its legitimate use in the b-verse on the one hand and to the enlargement of unstressed syllable sequences on the other. As a consequence, the original near complementarity in distribution between the two categories was lost, resulting in their partial merge and subsequent indistinguishability at large. In this way, type A3A1 came to be deprived of its structural autonomy and suffered inherent indeterminacy. The repercussions of the reconfiguration of anacrusis, however, did not stop here with the partial disruption of type A3A1. Concomitant with the extension of anacrusis to the b-verse, type A1 was rendered capable of occurring with anacrusis in the b-verse. As was true with normal verses in general, type A1 with anacrusis was capable of appearing on its own without regard to the nature of neighbouring verses. Given that type A1, when provided with a relatively long sequence of syllables as anacrusis, was intrinsically susceptible to a hypermetric rescansion, the configurations that did not occur in group with other hypermetric verses but were scannable as type A3A1 in structural terms would have given rise to a false (in the light of tradition) generalisation that hypermetric verses at large may be used in isolation. This would then have lent motivation for composing hypermetric a-verses in independent context in the Heliand, thereby contributing in part to the extensive use in this work of hypermetric verses beyond the traditional scope.

MOSP-05.qxd

5/11/04

2:43 PM

Page 330

5 The remaking of alliterative tradition: gradation and harmonisation

In composing a Christian heroic poem in the traditional alliterative metre, the Heliand poet encountered a serious challenge that was scarcely known to his Anglo-Saxon colleagues: he had to work in a significantly different prosodic environment, namely a weakening of stress in Old Saxon. The reduced force of stress in Old Saxon may be demonstrated by the following three phonological processes: (i) restoration of syncopated vowels (section 1.3.1); (ii) development of svarabhakti vowels (section 1.3.2); (iii) retention of /j/ that induced West Germanic Gemination (section 1.3.3). Syncopated vowels were restored in Old Saxon primarily by reintroduction of feet in word-final position: unstressed vowels were accordingly precluded from loss by virtue of their resultant incorporation into foot organisation. This reintroduction of feet came into effect by removing Defooting, the phonological rule that had been largely responsible for syncopation, and the removal at issue may in turn be ascribed to the weakened dominating power of preceding stressed syllables. Svarabhakti vowels developed in Old Saxon as a primary strategy for dealing with liquid reduction in prefinal position in the coda of stressed syllables. An alternative solution by diphthongisation (breaking) such as was implemented in Old English was hardly viable in Old Saxon given the lowered attracting power of stress: the complex coda consequent on breaking could not have been subsumed under stressed syllables in Old Saxon. In similar fashion, the West Germanic Gemination inducer /j/ was retained in Old Saxon on account of the reduced attracting power of stressed syllables: the output of West Germanic Gemination [-VC][CjV-] failed to be resyllabified [-VCC][jV-] in Old Saxon in contrast to Old English because the stressed syllable was found incapable of dominating such a cluster in the coda; /j/ was therefore immune to loss as it was not located in syllable-initial position in Old Saxon. Thus, generalising in terms of van Coetsem’s (1996) useful accent typology, Old Saxon is distinguished from its earlier stage, West Germanic, as well as from its sister language, Old English, by a lower degree of accent prominence. The weakening of stress in Old Saxon and the consequent relative levelling of stressed and unstressed syllables constituted a fundamental factor in reshaping the traditional alliterative metre. This prosodic change shook the foundations of the metre by hitting directly on the association between metrical positions and language material: it necessitated reorganisation of the original linguistic-metrical association based on footing. In the earlier metre, the normal drop was associated with unfooted material, whereas the lift and the heavy drop were occupied by a foot; the latter two

MOSP-05.qxd

5/11/04

2:43 PM

Page 331

5. The remaking of alliterative tradition

331

positions were in turn differentiated in prototypical terms by the status of the feet involved, primary versus non-primary. Such a sharp categorisation predicated on the presence and nature of feet, however, became vulnerable to loss and obscuration in Old Saxon largely because of the proliferation of feet occasioned by the weakened force of stress, as unstressed syllables came to be incorporated into foot structure in large measure. Consequently, the normal drop was permitted to be filled by any language material without regard to foot organisation in the Heliand metre, and this led to a blurred distinction between the normal and the heavy drops on the one hand, and between the normal drop and the lift on the other. The fuzzy categorisation of the originally distinct metrical positions resulted in a lesser differentiated use of primary-, secondary-, and weak-stressed syllables for verse composition, and therewith brought about an expansion in the range of language materials available for each kind of metrical position. As a consequence, traditionally marginal configurations came to be accepted as less peripheral, on the one hand, and new metrical structures were introduced that had been disallowed in the traditional metre, on the other. Specifically, the drop was increasingly filled by footed materials (stressed syllables), as with the first drop of type A1 (/ ⫻ / ⫻), the second drop of type B1 (⫻ / ⫻ /), and the final drop of type A3 (⫻ ⫻ / ⫻), to the point where even independent words of class 1, namely nouns and adjectives, were allowed to occur as an occupant of the normal drop, which had normally been reserved to the lowest ranking class of words (class 3) in the traditional metre. And type E (/ \ ⫻ /) exhibited a markedly increased incidence of an independent lexical word in the second position (\; e.g., 1323a godes uuang forgeben). Most important of all, having abandoned the original restriction on linguistic-metrical association based on the foot, the Heliand metre permitted the strings -Sx# and -Xx# to appear in contexts in which they had earlier been prohibited or deterred from occurring: -Sx# (e.g., 2492a mancunnie mildie) and -Xx# (e.g., 106a uualdandes uuilleon) were now capable of constituting the first drop of type A1; and -Xx# (e.g., 2458a undar fîundo folc) was increasingly compatible with filling the second drop of type B1, while the traditionally exceptional occurrence of -S# (e.g., 221a Thô sprac ên gêlhert man) and -sx# (e.g., 5727b Im ni uuelda thie folctogo thuo) became less marginal; -Sx# (e.g., 2181a ênan lîflôsan lîchamon) and -Xx# (e.g., 3414a the hêrosto thes hîuuiskeas) were allowed to realise the first drop of type D*, as did -S# (4215a hetelîc hardburi). Furthermore, -S# (e.g., 2062a is thit folc frômôd) occasionally constituted the second drop of type C (⫻ / / ⫻). A more dramatic consequence of the increasing decategorisation between normal and heavy drops concerns the status of metrical types: the traditional opposition between basic and increased metrical types was deprived of its structural basis, precisely because the opposition at issue resided in the well-articulated distinction between normal and heavy drops. Accordingly, the traditional types A2a (/ \ / ⫻), D2a (/ / \ ⫻), and D2b (/ / ⫻ \) no longer constituted independent metrical types in the Heliand: the corresponding configurations were now re-evaluated as realisation variants of types A1 and D (/ / ⫻ ⫻), respectively. By contrast, the opposition between basic and reduced metrical types (A3 ⫻ ⫻ / ⫻; B3 ⫻ ⫻ ⫻ /) was left intact, since it had little to do with the opposition between normal and heavy drops. The reorganisation of the original basic and increased metrical types as variant configurations of the same types would have been motivated and implemented through

MOSP-05.qxd

5/11/04

332

2:43 PM

Page 332

5. The remaking of alliterative tradition

the following design of the poet. As regards the reorganisation of types A1 and A2a, the poet, through his familiarity with the prototype-based gradient distinctions of the normal drop as keyed to the varying preference for the a-verse with double alliteration in traditional practice, would have been keen to recognise the redundancy and ambiguity inherent in the traditional organisation of the first drop of these two types: the verse form with a maximal preference for the a-verse with double alliteration (type A2a) could be interpreted as a most peripheral member of type A1. Capturing on such a conceptual redundancy, the poet would have simplified and reintegrated the traditional system by reducing the more specific metrical category (the heavy drop) to the more general (the normal drop). As a consequence, the first drop of type A1 came to be organised consistently by a single category, the normal drop, on a gradient basis. Accordingly, type A2a ceased to be an independent metrical form; instead, it became just a particular realisation of type A1. Much the same may be said about the integration of types D1, D2a, and D2b into a single metrical type, type D. More specifically, the configurations PsX/xx, PSX/xx, P#PX/xx, P#PS/sx, P#P#P/px, P#Px . . . P, and P#Px . . . S were all reduced to the common metrical type, type D, as its realisation variants. Rather than grouped into the three clearly bounded categories of separate metrical types as in the traditional metre, these variants were subjected to a finer gradation along the single scale of varying preference for the a-verse with double alliteration in accordance with a complex of parameters, as shown below. Yet by far the most outstanding consequence of the obscured distinction of metrical positions is the carefully controlled distribution of the enriched variety of configurations by a wholesale reorganisation of the metre on the basis of gradation: a majority of configurations, new and old alike, are implicated in graded distribution pattern, whereby relevant realisation variants of metrical types are ordered along the scale of preference for the a-verse with double alliteration (or conversely the b-verse) by a complex of parameters. The major variants of type A1 (/ ⫻ / ⫻) fall into three groups, which are ranked along the scale of increasing order of preference for the a-verse with double alliteration, as follows (section 2.1.7.1): (i) Psx#Px, Pxxx#Px, Px#Px, Pxx#Px; (ii) PX#Px, PS#Px, Pxx#Px, P#xPx; (iii) Px#xPx, P#xxPx, PXx#Px, PSx#Px, PX#xPx, PXxx#Px, P#xxx . . . Px, PS#xPx. Group (i) is characterised by the avoidance of the a-verse with double alliteration; characteristic of group (ii) is the absence of marked avoidance of or preference for the a-verse with double alliteration; and group (iii) is distinguished by the predominance of the a-verse with double alliteration. This graded patterning in distribution is in turn conditioned by the following five parameters: (i) long versus short syllables; (ii) the presence of an unfooted syllable; (iii) monosyllabic versus disyllabic endings; (iv) the number of syllables; (v) independent words versus prefixes. The parameters all concern various properties of syllables in such ways that a long syllable, the existence of a larger number of syllables, and the presence of a disyllabic (rather than monosyllabic) ending or of an independent word (rather than a prefix) contribute to increased preference for the a-verse with double alliteration in their different ways. The parameters are binary in value, marked versus unmarked; and whatever configuration is specified as marked along a given parameter has a stronger tendency to occur in the a-verse with double alliteration than does its unmarked counterpart.

MOSP-05.qxd

5/11/04

2:43 PM

Page 333

5. The remaking of alliterative tradition

333

The variety of medial and final syllables that may realise the second drop of type B1 (⫻ / ⫻ /) is broadly ranked along the differing incidence of the a-verse with double alliteration (section 2.4.2.3). As responsible for the ranking may be identified the following two parameters with the first one outranking the second: (i) presence versus absence of lexical stress, and (ii) presence versus absence of a word-final unstressed syllable following the lexical foot. Accordingly, a sequence of a lexical-stressed long syllable and a following syllable (-Sx#) is categorically ruled out as a possible realisation of the second drop of type B1. By contrast, the minimal stress bearing units standing alone (-S#, -sx#) may realise this position; yet these units, which count as marked by virtue of the presence of lexical stress, are distinguished from their minimally distinct unmarked sequences -X# and -xx#, respectively, by the significantly low incidence in occurrence and also by the relatively high frequency of the a-verse with double alliteration. Type C (⫻ / / ⫻) is subject to gradation in fine points of distribution primarily according to the varying ways the second lift is realised, as follows (section 2.6.2): (i) x . . . P#Px; (ii) x . . . P#px; (iii) x . . . PSx; (iv) x . . . Psx; (v) x . . . PXx; (vi) x . . . Pxx. When this position is filled by a primary-stressed syllable, type C occurs in the b-verse with overwhelming frequency. In sharp contrast, when the corresponding position is occupied by a syllable lacking lexical stress, this metrical type appears in the b-verse with minimal incidence. Between these two extremes is found the variant in which the second lift is realised by a secondary-stressed syllable: this variant is more or less evenly distributed in the a-verse and the b-verse. Furthermore, the pairs of the short and long syllables sharing the same kind of stress are subject to subtler distinctions on the basis of syllable length in highly context-sensitive manners. We may thus identify two parameters as governing the distribution pattern: (i) nature of stress, and (ii) syllable length. The major realisation variants of type D (/ / ⫻ ⫻) are ranked as follows in the increasing order of the incidence of the a-verse with double alliteration (section 2.7): (i) PS/sX/xx; (ii) P#PX/xx; (iii) P#P#P/px; (iv) P#Px . . . P; (v) P#PS/sx; (vi) P#Px . . . S. Most important, the occupation of one of the two drops by a lexicalstressed syllable correlates to a greater incidence of the a-verse with double alliteration. The whole variation in distribution is largely determined by the three prosodic properties of the drops, which may thus be identified as three parametric conditions on the above ranking: (i) nature of stress (lexical versus non-lexical); (ii) location of stress (penultimate versus ultimate); (iii) syllable length. In addition, heavy verses, that is, the configurations in which a drop is realised by an independent word, are subject to a further condition that has to do with a wider variety of alliterative patterns available to them (see below). While predominantly occurring in the a-verse with double alliteration, type D* (/ ⫻ / ⫻ ⫻) is also involved in gradation as regards fine detail of its distribution (section 2.8). As with type D, yet in a less outstanding way, the distribution of type D* is patterned in gradient terms depending on the two parameters: (i) stress degrees, and (ii) syllable length of the last three positions, particularly of the two drops. By the first parameter of stress degrees, the configuration PxSxx, a minor variant of subtype D*1, stands out from the rest of type D* by virtue of the maximal representation of the b-verse. By the same token, the remaining variants of subtype D*1 (Px . . . Pxx and Px . . . PXx) are characterised by a weaker preference for the

MOSP-05.qxd

5/11/04

334

2:43 PM

Page 334

5. The remaking of alliterative tradition

a-verse with double alliteration than the configurations Px . . . PS/sx (subtype D*2a) and Px . . . PxS (subtype D*2b). By the second parameter of syllable length, the pair of subtype D*1 with the second lift occupied by a primary-stressed syllable is further divided into two variants, Px . . . Pxx and Px . . . PXx, and the shorter variant Px . . . Pxx is distinguished from the longer one Px . . . PXx by its lower incidence of the a-verse with double alliteration. As far as heavy verses are concerned, a further condition comes into play, as for type D: heavy verses of type D* (subtypes D*2a and D*2b) are differentiated from those equipped with a secondary-stressed drop by their weaker preference for the a-verse with double alliteration, by virtue of a larger selection of alliterative patterns exclusively accessible to such verses. Unique to type D* is still another manifestation of graded distribution patterning in regard to the varying number of syllables that occupy the first drop: as the number of syllables involved increases, so does the relative proportion of the a-verse with double alliteration. The major variants of type E (/ \ ⫻ /) are involved in the following ranking in the increasing order of the incidence of the a-verse with double alliteration (section 2.9): (i) PX/x#x . . . P; (ii) PXx(x)#(x)P; (iii) P#Px(x)#(x)P; (iv) PSx(x)#(x . . .)P; (v) PS#x . . . P; (vi) P#P#x . . . P. The occupation of the second and third positions by the sequences -Sx#, -S#x, #Px#, and #P#x shows far stronger preference for the a-verse with double alliteration than does the realisation of the same positions by the sequences -Xx# and -X#x. The association of the second position with a syllable without lexical stress (x and X) favours occurrence of b-verses; by contrast, the realisation of the same position by a lexical-stressed syllable (S or P) brings about a higher incidence of a-verses with double alliteration. Furthermore, as shown by the two minimally distinct pairs of strings, -Sx# and -S#x, and #Px# and #P#x, the occurrence of the word-initial #x as opposed to -x# immediately after the second position favours the a-verse with double alliteration more strongly, a pattern which applies to the first drop of type A1 as well. On the other hand, the association of the same position with a primary-stressed syllable in heavy verses (P#Px(x)#(x)P) leads to a lower frequency of the a-verse with double alliteration than with a secondarystressed counterpart, as with types D and D*, again due to the multiplication of alliterative patterns available to such verses, as considered in due course below. We may thus identify the following three parametric conditions on the gradation for the first drop of type E: (i) nature of stress; (ii) syllable length; (iii) location relative to the word boundary of the syllable immediatley following the second position. A further gradation patterning is implemented in relation to the size of the second drop: as the number of syllables occupying the second drop (⫻) increases, the relative frequency of the b-verse decreases in favour of the a-verse with double alliteration. This general pattern obtains equally for each of the subtypes identified above. Although the particular parameters involved and their specific manners of co-ordination and ranking are not exactly the same across different metrical types, they all seem to be motivated by the unified design; in other words, they may be viewed as different strategies adopted as appropriate to individual contexts but driven by the same objective. Because the parameters largely concern the quality and quantity of syllable sequences, we may generalise that the underlying design is based on varying prominence of the metrical position as it gets associated with diverse language material for its realisation. The presence of stress, particularly of the one lexically assigned, the long syllable, and a larger number of syllables all contribute to an enhanced

MOSP-05.qxd

5/11/04

2:43 PM

Page 335

5. The remaking of alliterative tradition

335

prominence of the position involved; the presence of an additional unstressed syllable may reinforce the salience of the immediately preceding foot through contrast on the syntagmatic axis; and the opposition between prefix and independent word is also characterised as pertinent to relative prominence: since a prefix is incapable of occurring independently of the word it is attached to, it is regarded as less autonomous and hence less prominent than a word. The correlation between prominence and distribution pattern may therefore lend itself to the following generalisation: the more prominent the position is in realisation, the more frequent is the a-verse with double alliteration. The apparently distinct parameters and variation on their combination accordingly constitute different manifestations of the same underlying property, namely, prominence. The reduction in attracting power of the stressed syllable in Old Saxon would have entailed vital consequences in phrasal phonology as well: the primary-stressed syllable was rendered scarcely capable of accommodating unstressed syllables standing outside its own word domain. Cliticisation became correspondingly less prevalent. The diminished role of cliticisation in the phonology would in turn have led the poet to a reappraisal of the conventional practice based on this phrase-phonological process, namely the derivation of subtype D* from type D through encliticisation (whereby the first lift of increased metrical types is augmented with the following unstressed syllable or syllables) on the one hand, and use of anacrusis through procliticisation (whereby the first lift of increased metrical types is realised in extended shape by a primary-stressed syllable in conjunction with one or more preceding unstressed syllables) on the other. In the traditional metre, these unstressed syllables did not constitute independent positions of their own, rather they simply served as surface increments of the lift concerned. While having lost substantive support from the phonology, this traditional composition based on cliticisation was deprived of its original formal underpinnings in the Heliand metre by the obscured distinction between basic and increased metrical types. In the face of a rich body of relevant verses available for use in the inherited versecraft, however, the Heliand poet could hardly have afforded to ignore and abandon the traditional mode of composition altogether. Thus challenged to revitalise the metrical practice in question on a new basis, the poet seized upon a solution by providing these surface increments with the status of full positions: on the one hand, subtype D* was reconfigured as an independent metrical type, type D* (/ ⫻ / ⫻ ⫻), consisting of five full positions with the extra drop standing between the two lifts (section 2.8); on the other hand, anacrusis was promoted to a verse-initial optional drop that may be attached to the head of any metrical types beginning with a lift (section 2.11), that is, types A1, A2 (/ ⫻ / \), D, D*, E, and E* (/ ⫻ \ ⫻ /). The establishment of these optional elements as independent metrical positions in the Heliand resulted in their extended use in terms of frequency, distribution, and language materials. The incidence of type D* and anacrustic verses underwent a radical increase in the Heliand. No longer bounded by the original restriction in force in the traditional metre, an optional drop became capable of appearing as a second position after the first lift. Consequently, the light variant of type D*, subtype D*1, Px . . . PX/xx, became widespread in use; and proclitics, as well as enclitics, was permitted to occupy the additional drop in question. The marginal variant configuration of type E in the earlier metre, PxSxP, was reanalysed as containing an extra drop

MOSP-05.qxd

5/11/04

336

2:43 PM

Page 336

5. The remaking of alliterative tradition

in the second position, and an autonomous metrical type, type E* (/ ⫻ \ ⫻ /), came into being accordingly (section 2.10). And anacrusis was allowed to involve a long sequence of independent words, in ways comparable to the first drop of types B1 and C, rather than limited to prefixes. A further consequence of the promotion of anacrusis to a full position was the reorganisation of hypermetric b-verses, in particular of their prototypical type, type A3A1 (e.g., 1684b Bethiu ne thurbon gi umbi iuuua geuuâdi sorgon; section 4.2). Anacrusis and type A3A1 were largely in complementary distribution in the Old English tradition, and perfectly so in Beowulf. The two categories, anacrustic normal verses and hypermetric b-verses, were thus unambiguously distinguished and scarcely intersected each other in Old English metre. In the Heliand, by contrast, this and other types of hypermetric verses limited to the b-verse were differentiated from anacrustic normal verses only in probabilistic terms by reference to adjacent verses, rather than categorically on the basis of their own structural properties. This inherent indeterminacy of type A3A1 and other hypermetric types limited to the b-verse in the Heliand would presumably have been caused by the reorganisation of anacrusis whereby anacrusis was allowed to occur in the b-verse as freely as in the a-verse and realised by a long sequence of unstressed syllable as was the first drop of types B1 and C. As a consequence, the appearance of relatively long sequences of unstressed syllables in the b-verse became ambiguous in scansion, and the original near complementary distribution of the two categories of verse collapsed accordingly. As a result of the extension of anacrusis to the b-verse, type A1 became qualified for occurring in the b-verse accompanied with anacrusis. Since anacrusis was a property of individual verses, type A1 with anacrusis was capable of appearing freely regardless of the metrical contexts constituted by adjacent verses. Because type A1 with anacrusis was also scannable as hypermetric due to the structural indeterminacy of these two categories as observed above, those verses that were scannable as type A3A1 and occurred on their own without being paired with definite hypermetric a-verses would have given rise to a reanalysis whereby hypermetric b-verses were allowed to occur singly. Thereupon this reanalysis would in turn have prompted independent use of hypermetric a-verses in analogous ways. Hypermetric verses therefore came to enjoy extensive use in the Heliand going far beyond their original restrictions in traditional practice. Implementation and suspension of resolution is a device for controlling degrees of prominence falling on a metrical position by associating it with two different prosodic units, the foot and the syllable (section 3.1). This metrical device is based on the phonological generalisation that the bimoraic foot constitutes a minimal domain for encoding stress in the language. Depending on the length or moraic status of the stressed syllable involved, the foot falls into two types: (i) the monosyllabic foot, that is, the long stressed syllable (P or S); (ii) the disyllabic foot, or a sequence of the short stressed syllable and a following unstressed syllable (px or sx). Accordingly, stress is carried by the long stressed syllable on its own (monosyllabic foot) or by the short stressed syllable in conjunction with a following unstressed syllable (disyllabic foot). Since the long stressed syllable is invariably qualified for bearing stress by its inherent property of being minimally bimoraic, it is the short stressed syllable that is amenable to manipulation by means of implementation and suspension of resolution. In other words, as far as the long stressed syllable is concerned,

MOSP-05.qxd

5/11/04

2:43 PM

Page 337

5. The remaking of alliterative tradition

337

the opposition between implementation and suspension of resolution lacks manifest realisation on the surface: whether associated with the level of the foot or that of the syllable, it remains unchanged in its outer form. Resolution associates a given metrical position with a disyllabic foot, the sequence of a short stressed syllable and another syllable (px or sx); on the other hand, suspension of resolution associates it with a short stressed syllable alone (p or s). Implementation and suspension of resolution thus concern the two distinct prosodic units, the foot on the one hand and the syllable on the other. Since the foot is located higher up in the prosodic hierarchy, resolution realises a correspondingly greater degree of prominence than suspension of resolution. Because the lift prototypically corresponds with the foot in linguistic-metrical association (Suzuki 1996a: 379–81), suspension of resolution constitutes a marked option, and as such should be more susceptible to manipulation and reorganisation. As the original sharp distinction between lift and drop was obliterated and the drop became accordingly capable of association with footed material, the poet would have felt a growing need for reorganising resolution. He solved this problem by putting suspension of resolution under stricter control. Of particular importance are the following three contexts that he subjected to reorganisation. First, the configuration PS#px (subtype A1s) normally occurs while the resolved sequence PS#pxx (type A1) is all but absent. Second, the unresolved configuration Psxx (type D) appears, but the resolved counterpart Psxxx does not; conversely, the resolved configuration P#pxxx (type D) is attested, whereas the unresolved counterpart P#pxx is not. Third, the unresolved configuration PsxP (type E) tends to be avoided in favour of the resolved counterpart PsxxP. By redefining the domain for suspension of resolution more specifically, the poet succeeded in removing the apparently ambiguous contexts that had been subject to suspension of resolution as well as to implementation of resolution in the traditional metre. This firmer control of suspension of resolution in the Heliand metre seems inseparable from lack of significance of Kaluza’s Law in it, which figured centrally as a major constraint on implementation and suspension of resolution in Beowulf. The law referred to the moraic value of the second syllable of the disyllabic foot in determining whether resolution is implemented or suspended. And the law was of crucial importance in regulating the treatment of the second position in the context of Psx#Px among other metrical environments in the traditional metre: type A2a (/ \ / ⫻) provided a principal context in which normally the short disyllable (-VCi/u) was exclusively qualified for resolution to the exclusion of the long one (Suzuki 1996a: 223). The demise of type A2a as an independent metrical type, however, weakened the basis of the law, thereby endangering the distinguishability of the short disyllable in itself: upon loss of type A2a, the configuration Psx#Px became scannable without appealing to resolution, since the disyllable involved was matched with the first drop of type A1 without regard to resolution. A further difficulty in sustaining Kaluza’s Law came from the phonology. Old Saxon is characterised by a wealth of svarabhakti vowels, the unstressed vowels that developed secondarily due to vowel epenthesis (section 1.3.2). As a consequence, a wide variety of disyllabic forms emerged as candidates for resolution. Their original shape before operation of vowel epenthesis, however, was -VCC, a long syllable, and as such they had stood outside the purview of resolution in traditional metrical practice.

MOSP-05.qxd

5/11/04

338

2:43 PM

Page 338

5. The remaking of alliterative tradition

By virtue of their derived status, Kaluza’s Law as an archaic constraint on resolution would have been incapable of regulating the newly introduced disyllables. These novel word forms thus continued to be used as freely as had been their earlier monosyllabic forms (-VCC) without paying regard to the traditional opposition between implementation and suspension of resolution: these new long disyllables -VCV- were accordingly allowed to be resolved alongside the short counterparts even in the remaining minor contexts that required use of the short disyllables for resolution, particularly the second position of types D and E. It may be appropriate to recall that implementation of resolution constituted an unmarked option. In this way, the phonological change of these monosyllabic stems into resolvable disyllabic forms contributed further to disrupting Kaluza’s Law by reducing the original distinguishability of the short disyllables through the massive influx of derived long disyllables into the contexts where the original long disyllables were not permitted to occur. These two phenomena combined together, one metrical and the other phonological in motivation, deprived the short disyllables of the basis for their identification, and with it the identity of the long disyllables became ill-defined as well. Kaluza’s Law accordingly ceased to operate in the Heliand metre, and this necessitated a reorganisation of implementation and suspension of resolution. The unavailability of Kaluza’s Law, then, would have motivated the innovation of subsuming the unresolved configuration PS#px (subtype A1s) fully under suspension of resolution by generalising that resolution may be suspended when preceded by a stressed syllable. The loss of Kaluza’s Law had a further devastating consequence: the originally regular distribution of resolved and unresolved sequences for the second position of types D and E proved to be nearly out of control as the opposition between short and long disyllables became no longer available as a basis of generalisation. As an alternative generalisation, the poet captured on the difference in degrees of stress that served as a significant parameter elsewhere, and integrated it as part of the new rule for suspension of resolution. Occurrence of heavy verses was tightly controlled in the traditional metre (section 3.2.2). As observed above, the traditional metre was firmly founded on the sharp categorisation of the lift and the drop. Because the composition of heavy verses involves an independent word rather than the second element of a compound as a heavy drop, particularly one of the most prominent class (class 1), it seems reasonable that such special verses had to be permitted to appear only in strictly restricted contexts; otherwise, the very foundation of the metre would have faced the danger of falling into disintegration. Serving as a principal means to this end was the canonical alliterative pattern aax for heavy a-verses, which may be derived from the principle of left dominance. That is, given three (or more) words in a verse that were qualified for carrying alliteration on lexical grounds, the first two had to be alliterative. This restriction excluded as unacceptable the configurations PPPx (type A2a), PPpx (subtype A1s), and PPxP (type E) among others in favour of PPPx (type D2a), PPpx (type D2a), and PPxP (type D2b), respectively. Furthermore, single alliteration was prohibited from occurring on heavy a-verses of whatever composition, such as PPPx, PPpx, and PPxP. The blurring of the distinction in question in the Heliand, however, undermined the traditional restriction on heavy a-verses. Use of a prominent word as a drop became correspondingly more acceptable, and was accordingly found no longer disrupting

MOSP-05.qxd

5/11/04

2:43 PM

Page 339

5. The remaking of alliterative tradition

339

to the metrical system. Such a general decategorisation of metrical positions, notably between the lift and the heavy drop, would have encouraged the poet to construct new verse forms by analogy, that is, by substituting the second elements of compounds (S/s) with independent words (P/p). The traditionally canonical alliterative pattern aax would not have kept him from making such an innovation, because it had lost its original motivation in the meantime. Specifically, the sequence PPpx (2870a grôt craft godes) came to being partly by analogy to the well-established subtype A1s variant PSpx, and partly by analogical extension of the traditionally available b-verse variant of subtype A1s, PPpx (Suzuki 1996a: 89–90). The configuration PPxP (2138a gôd lioht mid gode) was created as a heavy verse variant of type E by the analogical substitution of the secondary-stressed syllable with the primary-stressed one; the traditional use of the string PPxP as an acceptable variant of type E in the b-verse also contributed to this analogical extension, as with the PPpx variant of subtype A1s discussed above. A further consequence of the proliferation of heavy verses was the introduction of the configuration PxPxP (2170a liudeo barnum leof) as a heavy variant of type E* on analogy with the newly established sequence PxSxP. Affected by the decategorisation of metrical positions to no lesser extent was the distinction between lift and normal drop. Prominent words, notably of class 1, were allowed to occupy not only the positions that had been a heavy drop in earlier metre, but also those that had been a normal one. As a consequence, such lexical words occurred in the first drop of types A1 (2234a thie godes suno mid is iungron) and D* (2285a gôd uuerk mid is iungeron), and in the final drop of type A3 (3805a endi an thene godes uueg), in the Heliand, positions that had been reserved for word-internal constituents and less salient words. In the traditional metre, the lift was sharply distinguished from the drop by its enhanced prominence, which was provided in complementary ways by a pair of processes operating on the two distinct constituents of the syllable: lift formation (the rime) and alliteration (the onset). Lift formation constituted a metrical analogue of foot construction: the lift was created on the basis of the bimoraic foot as determined by the rime. On the other hand, alliteration exclusively concerned the onset by repeating its representative segment. The original unity of the lift with these two processes became disintegrated in the Heliand metre, however. Insofar as the sharp distinction was maintained between stressed and unstressed syllables as in the traditional metre, the foot served as a firm basis of the lift by supplying a primary source of its prominence. As the distinction became blurred in Old Saxon largely upon proliferation of feet caused by the weakening of stress, however, the involvement with the foot in linguistic-metrical association ceased to be the privilege of the lift, which thus came to be deprived of one of its material bases of prominence. Meanwhile, the other basis of the lift met with danger of disintegration as well. In the traditional metre, double alliteration was primarily a function of extra metrical strength falling on the lift: double alliteration was prototypically realised when the lift needed to have its prominence reinforced through opposition to a heavy drop appearing in the same verse, a drop that was also associated with a foot. The blurred distinction between heavy and normal drops in the Heliand, however, led to a recharacterisation of double alliteration: rather than serving as a marker of an additionally strengthened lift as in the traditional metre, it was reconstituted as a marker of a

MOSP-05.qxd

5/11/04

340

2:43 PM

Page 340

5. The remaking of alliterative tradition

relatively long verse. And since verses tended to be composed in longer form in the Heliand, double alliteration figured as an unmarked alliterative pattern of the a-verse. Thus, reconfigured as an attribute of lengthy verses and hence dissociated from the lift, double alliteration can no longer be characterised in the Heliand as a privileged property of a lift supplied with added strength. Faced with such a weakened basis of prominence of the lift, the poet sought after some measure of compensation (section 3.2.3); the alliterative metre in itself would fall apart if he left it as it was, the poet may have reasoned in apprehension. As especially threatening would have struck him the so-called vowel alliteration: in the absence of an overt alliterative element, the lift appeared lacking in needed prominence. The poet would thus have devised an optional means of reinforcing alliteration on vowel-initial syllables by directly subjecting those vowels to alliteration. Such a decision would seem reasonable, given that the traditional perfect complementarity between the rime (lift formation) and the onset (alliteration) was no longer in force; accordingly, extending the domain of alliteration beyond the onset to involve part of the rime would have appeared practical. Integrating vowels into alliteration, however, had in turn immediate consequences on consonant-initial syllables. Since alliteration was a syllable-based metrical device, the vowels following the overt onset elements had to be treated as syllable constituents correspondingly, rather than autonomous segments standing on their own. In this way, in order to reinforce the weakened basis of the lift, the poet, building on the existing mechanism of alliteration in the traditional system, set out to make fuller use of it by extending its domain of operation from the onset to the syllable if as an optional embellishment, so that the lift would be provided with greater degrees of resonance. By bringing in the optimal syllable structure CV- rather than its minimal fraction C- to alliteration, the poet implemented an experimental measure to keep the lift prominent and thus to save the whole metrical system from collapsing. Thus, most if not all of the major changes in the Heliand metre as considered above can be characterised as loss of traditional metrical categorisations that were founded on full distinctiveness of stress, and may therefore be attributed ultimately to the weakening of stress and the consequent levelling of stressed and unstressed syllables in Old Saxon. Subsumable under this generalisation may be the following individual changes among others: the obscuration of the three kinds of metrical positions (the lift, the normal drop, and the heavy drop); the neutralisation of basic and increased metrical types; the reappraisal of parasitic weak derivative elements occurring before and after the first lift as full metrical positions; the increasing ambiguity and partial merge of normal and hypermetric verses; and the disruption of the unity between double alliteration and lift formation including resolution. In view of the foundational and pervasive role of stress in organising the traditional metre, it should be hardly surprising that, with the weakening of such a basis the Heliand metre underwent a radical transformation in its underpinnings. In it, stress was no longer the absolute organiser of the metre; it was demoted to serve as one of the prominence-related parameters for constructing the metre. Specifically, syllable quantity came to the fore as a more central parameter of prominence and accordingly as a more significant organiser of the metre at the expense of stress in regard to a number of metrical phenomena. For one thing, for realisation of the drop, the association with a longer sequence of unstressed syllables figured centrally as an

MOSP-05.qxd

5/11/04

2:43 PM

Page 341

5. The remaking of alliterative tradition

341

exponent of enhanced prominence by outranking the association with a stressed syllable, as with the composition of the first drop of type A1. For another, while increasingly diminished, the distinction between the a-verse and the b-verse that remained in evidence was marked out primarily in quantitative rather than qualitative terms through its greater propensity to expansion by a verse-internal extra weak position, rather than to association with a heavier syllable. As prominence thus became increasingly encoded by quantity-based means of linear extension oriented to the horizontal axis, as opposed to those based on weighting oriented to the vertical axis, the Heliand metre may be characterised as correspondingly removed from prototypically accentual versification. And such a transformation in overall organisation led to a partial disruption of the original isomorphism between the metrical and prosodic hierarchies, and obsoleted metrical representation in terms of grid structure, a mode of representation that we found most adequate for the metre of Beowulf (Suzuki 1996a). For all its significance, the diminished distinction of metrical positions would hardly have given rise to all relevant metrical reconfigurations as its mechanical consequences: some metrical reorganisations seem to defy full reduction to simple phonological change. Furthermore, even those metrical changes that may broadly be characterised as linguistically motivated would be far from explainable wholly in such terms in their specifics: the fundamental prosodic change at issue would have been compatible with other manners of reorganisation including absence of appreciable pattern at all, and accordingly these changes would also have been shaped in points of detail non-mechanically, as shown in the preceding paragraphs. Rather than fully committed to linguistic determinism, we would thus have to assume the Heliand poet’s active and creative role in reorganising the metre. A prime manifestation of the poet’s artistic creativity seems to lie in the gradation and harmonisation that a wide spectrum of verse configurations are organised with in their distribution patterns. We should now explore artistic aspects of this gradation patterning and the nature of the grand design presumably underlying the poet’s creativity. While irreducible to linguistic change in full, part of the gradation patterning at issue should be viewed as an inheritance from traditional versecraft, rather than the Heliand poet’s own invention, since we find a similar regularity in Beowulf, for example: variant configurations of some metrical types, type A1 in particular, show a comparable scalar distinction of varying preferences for the a-verse with double alliteration depending on an array of similar parameters (Suzuki 1996a: 149–58). In this sense, highly innovative as he was, the Heliand poet was far from revolutionary in practice: he conformed to tradition as closely as possible and created a new order out of the resources available in the inherited system. Yet details of the patterning and the overall principle of design seem to defy wholesale reduction to conformity to tradition; rather they have to be appreciated essentially as the Heliand poet’s own making. The poet achieved the refining and remaking of the traditional gradation in diverse ways. First, he enriched the existing pattern by adding new variant configurations, such as the sequences PSx#Px and PXx#Px for type A1, and x . . . PXx#P for type B1, as pointed out above. Second, he improved on the traditional pattern by integrating into it some configurations that had resisted incorporation so far, such as the configurations PS/sX/xx for type D. Third, he modified the existing gradation by re-evaluating

MOSP-05.qxd

5/11/04

342

2:43 PM

Page 342

5. The remaking of alliterative tradition

and reranking some underlying parameters: for example, for the composition of the first drop of type A1, he reorganised the parameter of syllable length as a primary organising feature at the expense of stress, which had hitherto served as a major parameter; the role of stress accordingly receded to the background, and as a consequence the strings -S# and -sx# were encoded with lower preference for the a-verse with double alliteration. Fourth, he introduced a further articulation to the inherited pattern by making a distinction of the traditionally undistinguished configurations, such as Px#Px and PX#Px for type A1, and P#Pxx and P#PXx for type D. Fifth, and most significant of all, by analogy with the existing pattern, he devised a comparable gradation where there had been none before, as for types C, D*, and E. Thus, while firmly building on traditional conventions and constraints, the poet manipulated, elaborated, and systematised them as befitted his design of creating more refined and integrated gradation patterning. The poet accordingly accomplished a full reorganisation of the traditional pattern in the metre. The gradation patterning that the poet designed and executed had another dimension to it. While redefining the contexts for suspension of resolution more strictly and thus securing a tighter control over it as shown above, the Heliand poet elaborated on implementation of resolution by generalising and reorganising the variation pattern existing in the traditional metre and integrating it more fully on the basis of a finer and more harmonious gradation (section 3.1.4). He started with the observation that the first lift of type C was subject to variation on incidence of resolution as conditioned by the three distinct stress properties of the immediately following syllable (primary-stressed, secondary-stressed, and non-lexical-stressed), whereby the presence of a stronger stress correlated to a higher frequency of resolution on the first lift. The following three configurations were accordingly ranked in the decreasing order of incidence of resolution on the first lift: x . . . P#P/px, x . . . PS/sx, and x . . . PX/xx. Drawing on the patterned variation that he identified within the traditional metre, the poet then set out to regulate variation on resolution for other sets of configurations that were amenable to comparable grouping and differentiation on the basis of analogous stress properties of the following syllable. As a consequence, the configuration P#PS/sx came to be sharply distinguished from the configuration P#Pxx by a markedly greater frequency of resolution on the second lift. Conversely, the configuration Px . . . Pxx, which had been affected frequently by resolution on the second lift, was recharacterized as hardly compatible with the process. The poet also extended the gradation pattern to the pair of type E configurations PSx . . . P and PXx . . . P with respect to resolution on the first lift, which had not been involved in such a sharp differentiation in traditional practice: the configuration PSx . . . P was now treated as more susceptible to resolution on the first lift than the configuration PXx . . . P. A further notable case of the extension of the graded patterning of resolution is the second lift of type A2 (Px . . . PS). This position was not only subjected to higher frequency of resolution, but implicated in the graded variation on its operation differentiated according to the varying degrees of prominence as encoded in the distinct morpholexical status of the language material that occupied the following syllable. Such a variation on resolution had not have anything remotely comparable in the traditional metre, however. The ascendancy of the immediately following stressed syllable as a major determinant of increased incidence of resolution in the Heliand would in turn have heightened the role that the

MOSP-05.qxd

5/11/04

2:43 PM

Page 343

5. The remaking of alliterative tradition

343

immediately adjacent stressed syllable had played in inducing resolution on the second lift in the traditional metre. Correspondingly, the configuration x . . . PS#P, a variant of type B1 that was found increasingly acceptable, came to be integrated into this generalisation, with the resulting extensive operation of resolution on the second lift. This may be counted as another significant innovation of devising well-integrated gradation patterning by the Heliand poet. Gradation distributes members of a given class in relative terms more or less evenly along a scale according to a chosen parameter or parameters; it avoids polarity and binarity in opposition in favour of graded, qualified, and relative difference; it minimises contrast and separation in favour of similarity and continuity; it is thus antithetical to polarisation and dichotomisation that constitutes a sharp division into two diametrically opposed categories. Underlying the poet’s deep commitment to a fuller integration and generalisation of gradation pattern may therefore be identified his profound sense of balance and harmony as an ultimate source of artistic innovation. This inner motivating force may thus be characterised as a grand principle of maximal harmony, to be postulated in contrast to the one of maximal contrast, which I showed to have permeated the metrical organisation of Beowulf (Suzuki 1996a). Based on the foregoing considerations, we may thus assume that the Heliand poet had a firm control of the traditional metre in its structural details; he also had a keen awareness of innovation where it was needed; and he was strongly inspired by aesthetic sensibilities of balance and harmony. Such an image of the poet as highly traditional, largely innovative, and profoundly sensitive to symmetry, would seem to be borne out in other aspects of his verse-making activity. As Murphy (1992: 221–30) has shown based largely on Rathofer’s (1962) findings, the Heliand is distinguished by symmetrical arrangement of fitts (songs).1 By this method of arrangement, Song 38 ‘The Transfiguration on the Mount’ constitutes the centre of the epic, and pairs of songs parallel in content are placed symmetrically before and after that central scene in the form of inclusio, namely a multiple framing with the overall structure a, b, c . . . x . . . c’, b’, a’ (x ⫽ central song; a/a’, b/b’, c/c’ ⫽ pairs of parallel songs). For example, Song 33 ‘Death of John the Baptist’ and Song 43 ‘Death of Jesus Foretold’, or Song 28 ‘The Cure of the Bed-fast Man’ and Song 48 ‘The Raising of Lazarus’ (Murphy 1992: 229) constitute such pairs of parallel songs; and one member (Song 33 or 28) comes before the central song (Song 38), the other after it (Song 43 or 48), placed at an equal distance away from the centre. Furthermore, the poet’s fascination with the framework of inclusio provided motivation for the retelling of individual Gospel stories, as exemplified in Murphy (1989: 68–73). Thus, as far as narrative structure is concerned, the whole epic is permeated by the organising principle based on symmetry and multiple embedding (cf. Rathofer 1962: 543–4). Of vital interest here is that, in order to achieve the symmetrical structure that he strove for, the poet deliberately rearranged and restructured the original stories in the Gospels; such were the poet’s sense of balanced structure and his innovative spirit that he could not restrain himself from reorganising and reconfiguring traditional materials at his hand, when he found appropriate grounds and sources for improvement. 1

‘It seems that there is more than enough evidence for deliberate symmetrical structuring of a significant number of the songs of the Heliand’. (Murphy 1992: 230)

MOSP-05.qxd

5/11/04

344

2:43 PM

Page 344

5. The remaking of alliterative tradition

In the light of the poet’s keen sense of symmetry and balance, and his active engagement with traditional material in the restructuring of Gospel stories, it seems hardly surprising that in dealing with the medium of his narration, the poetic metre, he was also inspired by the same spirit and appealed to a similar principle of organisation. Confronted with the weakening of stress and the consequent levelling of stressed and unstressed syllables in the language he had to compose in, he had to deal squarely with its immediate metrical consequences: the weakened opposition between the three kinds of metrical positions. He then reorganised and integrated the traditionally discrete categories of basic and increased types as single types that he subjected to fine gradation patterning in distribution; because the increased metrical types had been largely limited to the a-verse, this unification led to harmonising the traditional sharp opposition between the a-verse and the b-verse, and thereupon emerged a more balanced line structure with a close approximation of the two constituent verses. Meanwhile, being aware of the reduced prosodic motivation, the poet reconstituted the two extra metrical weak elements, the first drop of type D* and anacrusis, as optional full metrical positions. By doing so, he came closest to denying the first lift its original superiority in prominence, and made the difference between the first and second lifts less outstanding; on the other hand, he arrived at a lesser distinction and greater overlapping of the two clearly bounded categories of the normal and hypermetric verses in the traditional metre. Generalising the partly patterned variation on resolution inherited from traditional practice, the poet gave a more articulate gradation to the metrical device by making it sensitive to stress properties of the following syllable. The enriched alliteration patterns of heavy verses were another achievement that the poet made by turning to advantage the obscured distinction between lift and drop. Inspired by the sense of maximal harmony and symmetry, and thus in disregard of the canonical alliterative pattern aax, a traditional generalisation founded on the principle of maximal contrast, he explored logical combinatory possibilities thus far disallowed and came out with a multiplicity of patterns. The reorganisation and restructuring that the Heliand poet achieved in the metre as discussed in detail in this book may thus be ultimately attributed to his ingenuity as an epic poet who made innovative use of traditional sources in highly controlled and imaginative ways. And as with the narrative structure, the metrical organisation of the Heliand permeated by balance, harmony, and gradation in its whole and parts may reveal itself to an experienced audience who have comparable structural awareness and aesthetic sensibilities cultivated in the mind through continual engagement in attentive listening and careful reading.

MOSP-Appendix 1.qxd

5/11/04

2:44 PM

Page 345

Appendix 1. Foreign names There are 520 verses in the Heliand in which one or more foreign names appear, as listed below. These verses were largely disregarded from consideration in the text because of uncertainties that many of them present in interpretation, particularly in regard to vowel length, as remarked in section 1.4 above. Such indeterminacy of vowel length poses insurmountable difficulty in identifying metrical types and their realisation variants in precise terms, and also in determining whether resolution is implemented or not. Only with reference to the detailed knowledge that we arrived at through close examinations of metrical treatment of native and fully integrated language materials do we begin to gain insight into the prosodic structure of some of these otherwise intractable foreign names, as will be explored in what follows. (1) List of verses containing at least one foreign name a-verse (317 examples): 13a, 18a, 19a, 54a, 57a, 61a, 63a, 65a, 67a, 69a, 71a, 72a, 75a, 76a, 88a, 96a, 97a, 120a, 139a, 218a, 250a, 254a, 255a, 257a, 259a, 326a, 339a, 357a, 359a, 361a, 363a, 364a, 368a, 370a, 401a, 404a, 452a, 455a, 459a, 461a, 463a, 491a, 505a, 532a, 548a, 606a, 612a, 621a, 625a, 628a, 685a, 711a, 716a, 728a, 731a, 749a, 756a, 757a, 762a, 766a, 768a, 769a, 772a, 776a, 780a, 782a, 788a, 791a, 803a, 833a, 859a, 910a, 914a, 920a, 930a, 951a, 952a, 960a, 965a, 994a, 1020a, 1031a, 1036a, 1046a, 1081a, 1108a, 1127a, 1130a, 1135a, 1151a, 1153a, 1166a, 1175a, 1181a, 1256a, 1258a, 1263a, 1270a, 1675a, 1952a, 1995a, 2002a, 2072a, 2075a, 2089a, 2126a, 2134a, 2135a, 2177a, 2221a, 2291a, 2321a, 2340a, 2586a, 2648a, 2664a, 2687a, 2699a, 2705a, 2711a, 2733a, 2737a, 2794a, 2800a, 2832a, 2842a, 2933a, 2983a, 2986a, 2991a, 3000a, 3006a, 3054a, 3062a, 3069a, 3088a, 3109a, 3142a, 3171a, 3183a, 3184a, 3187a, 3196a, 3201a, 3210a, 3306a, 3318a, 3352a, 3360a, 3365a, 3375a, 3380a, 3387a, 3389a, 3396a, 3398a, 3521a, 3543a, 3547a, 3557a, 3563a, 3595a, 3625a, 3635a, 3655a, 3672a, 3684a, 3691a, 3702a, 3708a, 3716a, 3717a, 3719a, 3783a, 3794a, 3809a, 3828a, 3928a, 3952a, 3955a, 3957a, 3965a, 3967a, 3972a, 3973a, 3977a, 3982a, 3985a, 3993a, 4006a, 4012a, 4014a, 4016a, 4019a, 4028a, 4034a, 4066a, 4067a, 4070a, 4079a, 4104a, 4107a, 4110a, 4126a, 4135a, 4142a, 4147a, 4162a, 4170a, 4181a, 4186a, 4189a, 4200a, 4203a, 4206a, 4214a, 4237a, 4263a, 4364a, 4368a, 4370a, 4459a, 4465a, 4478a, 4480a, 4508a, 4516a, 4533a, 4546a, 4598a, 4599a, 4615a, 4624a, 4629a, 4659a, 4700a, 4719a, 4736a, 4826a, 4829a, 4845a, 4848a, 4875a, 4883a, 4937a, 4939a, 4946a, 4951a, 4953a, 4958a, 4960a, 4975a, 4992a, 4994a, 5052a, 5057a, 5107a, 5109a, 5112a, 5126a, 5129a, 5145a, 5150a, 5160a, 5176a, 5203a, 5212a, 5238a, 5240a, 5250a, 5253a, 5256a, 5259a, 5262a, 5270a, 5281a, 5283a, 5297a, 5310a, 5320a, 5345a, 5358a, 5376a, 5387a, 5402a, 5428a, 5429a, 5435a, 5487a, 5516a, 5525a, 5540a, 5552a, 5557a, 5572a, 5606a, 5607a, 5611a, 5617a, 5670a, 5719a, 5765a, 5784a, 5821a, 5838a, 5856a, 5866a, 5874a, 5881a, 5897a, 5906a, 5911a, 5915a, 5955a, 5956a, 5968a, 5981a

MOSP-Appendix 1.qxd

346

5/11/04

2:44 PM

Page 346

Appendix 1. Foreign names

b-verse (203 examples): 60b, 61b, 74b, 97b, 104b, 133b, 198b, 236b, 250b, 252b, 295b, 307b, 313b, 318b, 340b, 424b, 444b, 458b, 468b, 504b, 532b, 640b, 696b, 700b, 704b, 745b, 764b, 766b, 780b, 788b, 791b, 792b, 806b, 830b, 833b, 865b, 873b, 910b, 951b, 965b, 970b, 977b, 1042b, 1114b, 1127b, 1151b, 1159b, 1192b, 1227b, 1262b, 1263b, 1265b, 1269b, 1271b, 1473b, 1591b, 1997b, 2072b, 2075b, 2079b, 2125b, 2134b, 2138b, 2234b, 2273b, 2285b, 2360b, 2664b, 2679b, 2703b, 2728b, 2774b, 2805b, 2982b, 3008b, 3035b, 3039b, 3042b, 3043b, 3045b, 3085b, 3093b, 3108b, 3129b, 3136b, 3141b, 3165b, 3176b, 3183b, 3241b, 3257b, 3335b, 3361b, 3367b, 3521b, 3543b, 3557b, 3625b, 3682b, 3685b, 3702b, 3716b, 3734b, 3742b, 3748b, 3821b, 3833b, 3850b, 3858b, 3871b, 3879b, 3884b, 3896b, 3982b, 3985b, 4016b, 4074b, 4094b, 4096b, 4126b, 4148b, 4152b, 4190b, 4198b, 4200b, 4214b, 4236b, 4239b, 4469b, 4476b, 4552b, 4562b, 4577b, 4673b, 4720b, 4724b, 4736b, 4810b, 4811b, 4815b, 4837b, 4847b, 4850b, 4866b, 4913b, 4925b, 4948b, 4955b, 5089b, 5104b, 5109b, 5127b, 5129b, 5133b, 5136b, 5142b, 5154b, 5176b, 5179b, 5180b, 5214b, 5223b, 5232b, 5240b, 5245b, 5251b, 5259b, 5275b, 5294b, 5304b, 5326b, 5368b, 5372b, 5379b, 5404b, 5409b, 5413b, 5440b, 5470b, 5481b, 5511b, 5533b, 5551b, 5560b, 5639b, 5677b, 5689b, 5696b, 5719b, 5721b, 5747b, 5749b, 5761b, 5780b, 5800b, 5819b, 5835b, 5876b, 5895b, 5898b, 5958b, 5968b, 5972b Regardless of the original accentual pattern in Latin, probably the immediate source language (cf. Hofmann 1991: 99), all foreign names (with the exception of Erodes; see below) carry stress on the initial syllable, as demonstrated by consistent alliterative practice. Moreover, the initial open syllable has to be interpreted as containing a long vowel in many examples, even when their Latin counterparts begin with a short vowel. This generalisation seems hardly in doubt: a number of names indicate with varying degrees of cogency that scanning them as containing a short initial vowel is metrically infeasible, as briefly substantiated in the following. Dauid (Lat. David): Verse 3682b sunu Dauides, scanned as type D, seems more congruent with Pxx than with pxx for Dauides, because the second lift of type D is more likely to be resolved than suspended from resolution, as shown in section 3.1.1.3 above. Emaus (Lat. Emmaus): Verse 5958b uueldun im te Emaus, scanned as type C, seems most compatible with the phonological value of Pxx for Emaus: with alternative characterisations, pxx, pX, and Px, we would come out with the most unlikely consequence of scanning the verse as type A3 in the b-verse. The question that Kauffmann (1887: 352) posed, namely, whether the second or third syllable carries a secondary stress, would seem immaterial, because the second lift of type C may be realised by an unstressed medial syllable. The identification of Emaus as trisyllabic (Pxx), furthermore, may justify us in extending analogous analyses to similarly composed names ending in -Vus, such as Mattheus, which would otherwise be lacking in crucial evidence for proper scansion. Erodes (Lat. Herodes): This is the only word that is subject to variation on alliterative patterning: out of seventeen verses in which this word occurs, alliteration falls on the initial syllable E- in thirteen instances, and on the second syllable -ro- in the remaining four (71a, 548a, 716a, 728a). As regards the length of the vowels when equipped with primary stress, no conclusive evidence is available concerning the second vowel -o-: all four examples are compatible with either value. In the absence of evidence to the contrary, however, we may perhaps prefer a long vowel, given the

MOSP-Appendix 1.qxd

5/11/04

2:44 PM

Page 347

Appendix 1. Foreign names

347

pervasiveness of Sievers’s (1885) generalisation referred to below. On the other hand, there is good reason for identifying the primary-stressed initial syllable E- as long. Of crucial significance are the following three minimal verses: 685a Erodesan; 5270a cuning Erodes; 60b Erodes uuas. Were the initial vowel counted as short, these verses would contain only three metrical positions, thereby falling short of the requirement of four positions per verse. The next question to be addressed concerns the prosodic identity of the vowel in the non-primary-stressed, medial syllable. One of the verses referred to above, verse 60b Erodes uuas, has no choice but to be scanned as type E. In order to satisfy such a scansion, however, the second syllable of Erodes must have one of the following three prosodic properties: long and unstressed (X), short and stressed (s), or long and stressed (S); by contrast, the trisyllabic word with the short and unstressed, medial syllable (i.e., #Pxx#) is categorically unqualified for forming on its own the first three positions of type E (section 2.9). Furthermore, given that the second position of type E is only exceptionally suspended from resolution (section 3.1.1.5), postulating the medial vowel as short and stressed (s) would prove to be less convincing. Of the remaining two candidates, PSx on the one hand, and PXx on the other, the occasional participation of the second syllable in alliteration as noted above might motivate us in postulating some stress as a reflex of primary stress even when alliteration is realised by the initial syllable, as Hofmann (1991: 102) remarks. When alliteration involves the second syllable, however, it is metrically immaterial whether or not the initial syllable retains some degree of stress. Furthermore, there seems no particular reason for treating the second syllable of Erodes as analogous in morpholexical terms to the second element of a native compound, mancunni, for example (cf. Hofmann 1991: 102). In this light, we have to be sceptical about insisting on the prosodic structure PSx for Erodes. We may accordingly conclude that Erodes has two variant prosodic structures, PXx (or less plausibly PSx) on the one hand, and xPx (or less plausibly xpx) on the other. Eua (Lat. Eva): Two verses 1036a Adaman endi Euan and 3595a Adam endi Evan, which may be identified as type A1 rather than a rare variant of type E, show that Eua is Px. Iacob (Lat. Iacobus): Verse 1265b Krist endi Iacob, a type A1 verse in all likelihood, shows that this name has to be scanned as Px. Iesus (Lat. Iesus): The identification of Iesus as Px rather than px is supported by the following verses that could be scanned as nothing other than type B1: 326a that is Iesu Krist; 3557a and 3716a quað that thar Hiesu Crist; 3257b endi frâgode Iesu Crist. Iohannes (Lat. Iohannes): Kauffmann (1887: 352) argued for characterising Iohannes as pXx with a short initial vowel, and admitted therewith that verse 965a thar Iohannes is at variance with his scansion. Following Hofmann (1991: 98, 99), however, this verse should be regarded as crucial evidence for identifying the word as PXx: Kauffmann’s scansion would have to create a three-position verse, as noted by Kauffmann (1887: 352) himself. Furthermore, on identifying Iohannes as PXx, verses like 236b Iohannes namon are scanned naturally as normal variants of type E, while the alternative value pXx would lead to either a three-position verse or a variant of subtype A1s that is unattested elsewhere (section 2.1.8). Iudas (Lat. Iudas): Verses 4478a Thô geng imu thar Iudas forð and 4720b That uuissa Iudas uuel (type B1) testify that Iudas contains a long stressed syllable P,

MOSP-Appendix 1.qxd

348

5/11/04

2:44 PM

Page 348

Appendix 1. Foreign names

rather than a short one. This identification is corroborated by the following minimally composed type A1 verses, 4629a and 5160a Iudas gangan: the alternative assumption of a short stressed syllable would result in an unmetrical verse with three metrical positions. Further support for identifying the initial vowel involved as long is provided by verse 4810b Uuîsde im Iudas: if Iudas began with a short stressed syllable, the verse in question would constitute an exceptional variant of type E with the second position associated with a word-final syllable (section 2.9). As for the length of the second syllable, verse 5145a Iudases hugi clearly shows that it should be scanned as long (-X-): the otherwise resulting configuration Pxx#px is categorically excluded as unmetrical (section 2.9). The proposed scansion of Iudas as PX rather than Px is also compatible with the other type E verses, 4615a Iudase an hand and 5429a Iudases an grund: as pointed out in section 2.4.2.3 above, the configuration Pxx#xP is unattested in the corpus, whereas PXx#xP is common. Peter: Verse 3069a sancte Peter requires us to identify Peter as Px, rather than px, which would result in too short a verse consisting of only three positions. Rumu (Lat. Roma): The identification of Rumu as Px rather than px is supported by verses 4142a rinkos fan Rumu and 5253a the rîkeo fan Rumu, both of which should be scanned as type A1, rather than as an aberrant variant of type E. Satanas: While reading Satanas as pxx for verses like 2586a Satanas selbo, Kauffmann (1887: 352) was, without justification, inclined to scan verse 4624a uuarð imu Satanas as type C (with Satanas identified as Pxx), rather than type A3, and thus posited two variant readings pxx on the one hand and Pxx on the other. By contrast, Hofmann (1991: 104) consistently scans the word as Pxx, as with other comparable names. In the light of analogous treatments elsewhere, Hofmann’s exclusive postulation of the long initial vowel P rather than the short one seems most plausible. Of related interest in this connection is verse 1114b Satanas thanan. On Kauffmann’s assumption of pxx, the whole verse would constitute the configuration pxx#px. Yet such a sequence would have to be rejected as unmetrical: to scan it as subtype A1s, presumably the only possible scansion, must be ruled out, because the second position of this type has to be filled by a stressed syllable, as demonstrated in section 2.1.8 above. While supporting the postulation of the long initial vowel, the same verse calls into question another aspect of Hofmann’s reading Pxx, namely the length of the medial vowel. Following Hofmann’s interpretation, this verse would result in Pxx#px, an otherwise unattested example of type E. This configuration violates the generalisation that the first drop of type E has to be occupied by a long syllable if it is medial and without lexical stress, as discussed in section 2.9 above. Given this strict metrical regularity, it would seem more appropriate to scan Satanas as PXx with the long medial vowel, in parallel to Iudas (PX) as proposed above: the configuration PXx#px constitutes a frequently occurring variant of type E. Correspondingly, verse 1108a Satanasan forsuuêp would be analysed as PXxx#xP, a marked variant of type E with the second drop associated with three syllables, to be sure, but similar configurations are attested several times nonetheless, such as 1668a uuârlîco undaruuitan. Simon: Numerous examples of Simon Petrus (3054a, 3196a, 3201a, 3210a, 3306a, 4508a, 4598a, 3093b, 3108b, 4866b, and 5898b) justify us in identifying Simon as Px; otherwise, we would obtain verses containing three metrical positions or ones consisting of four positions with the unresolved first lift, extremely implausible verse forms on either scansion.

MOSP-Appendix 1.qxd

5/11/04

2:44 PM

Page 349

Appendix 1. Foreign names

349

On the strength of the foregoing examinations, it seems warranted to follow Hofmann (1991) and extend to Old Saxon metre the generalisation that Sievers (1885: 492) formulated for Old English metre: “Als allgemeines gezetz gilt, dass die betonten silben der fremdnamen als lang gelten.” The Heliand metre, however, offers a single notable exception to Sievers’s generalisation. At stake is the word Iudeo, which should be identified as pxx (with a short initial vowel), rather than Pxx, in the Heliand, as will be shown in the following discussion. By contrast, the original vowel of this word was long and unaccented in Greek (’Ioydaîos) and Latin (Iudaeus). With stress shift in Proto-Germanic, this initial vowel would have remained long as most naturally expected drawing on comparable examples adduced above. As it turns out, the prediction is corroborated in Old English metre: Iudeas should be read as PXx, as shown by type E verses such as Andreas (Krapp 1932) 560a Iudea cynn. Inapposite as evidence in this regard are the verses that Hofmann (1982: 298) adduced as support for scanning Iudea as PXx: Christ (Krapp and Dobbie 1936) 637a þone Iudeas; Elene (Krapp 1932) 216a to Iudeum. These verses scan as type C, and thus are compatible with the prosodic structure Pxx with the short medial syllable. In any event, Old English metrical practice fully conforms to Sievers’s generalisation that, whether original or derived, stressed initial vowels of Latin loan words are treated as long in Old English. Things are rather different in Old Saxon, however. Old Saxon Iudeo is characterised by a notable feature of having two variant forms in the genitive plural, Iudeo on the one hand and Iudeono on the other (Sehrt 1966: 731–2). The existence of doublets, however, is unique to the genitive plural: all other instances of Iudeo in the corpus are weak forms, Iudeon, except for a single occurrence of the nominative singular Iudeo. Table A1.1 shows the distribution of these two genitive plural variants, in combination with the two major nominal forms that they modify, P and Px. As should be clear from the representation below, the longer variant Iudeono has to be characterised as unmarked, given the by far wider contexts of its occurrence. This characterisation seems in keeping with the appearance of the strong form restricted to the genitive plural, as observed above. The marked status of Iudeo#Px is further indicated by the small selection of lexical items that realise Px: as pointed Table A1.1. Distribution of Iudeono and Iudeo in combination with the word forms P and Px context

Iudeono

Iudeo

_#P P#_ x_#P xP#_

⫺ ⫹ ⫹ ⫹

⫺ ⫺ ⫺ ⫺

_#Px Px#_ x_#Px xPx#_

⫹ ⫹ ⫹ ⫹

⫹ ⫺ ⫹ ⫺

Key: ⫹ ⫽ attested; ⫺ ⫽ unattested

MOSP-Appendix 1.qxd

350

5/11/04

2:44 PM

Page 350

Appendix 1. Foreign names

out by Kauffmann (1887: 350), the Px word in Iudeo#Px is largely limited to liudi ‘people’, while Iudeono#Px shows greater freedom in collocation including liudi itself. There is a single exception, however, to the ubiquitous occurrence of the weak form Iudeono, namely the context _#P: Iudeono is unattested in verse-initial position when immediately followed by a monosyllabic nominal form standing in verse-final position. By contrast, when preceded by one or more unstressed syllables, Iudeono is allowed to be used even before the verse-final monosyllabic word form P, as in 61b ober that Iudeono folc. In view of such a unique distributional gap of Iudeono, we may be led to assume that the logically possible configuration Iudeono#P was metrically objectionable; specifically, this configuration would not have been properly scanned as type E, and therefore ruled out as unmetrical. From this, we may in turn infer that Iudeono is prosodically characterised as pxxx (with a syllabic -e-) or Pxx (with a non-syllabic -e-), rather than Pxxx like craftigana in 2804a craftigana Crist. As discussed in section 2.9.4 above, the configuration Pxxx#P is qualified to constitute type E if only as a peripheral variant; accordingly, scannable as such, there would have been no principled reason why the collocation Iudeono#P had to be excluded. Furthermore, on the interpretation that Iudeono is Pxxx, verses like 3783a grôt folc Iudeono, 4826a grim folc Iudeono, and 4925b môdag folc Iudeono would constitute an anomalously heavy or hypermetric configuration, Px#P#Pxxx. Therefore, we can hardly avoid the conclusion that the initial vowel u of Iudeono is short. The unmetrical status of Iudeono#P is indeed rectified when one or more unstressed syllables are placed at the beginning of a verse, which would then result in a legitimate instance of type B1. Without such extra weak elements, however, Iudeono has to be placed after a monosyllabic nominal form that it modifies. Yet the monosyllabic form P would still seem to suffer a metrical disadvantage in comparison with the disyllabic one Px: in the absence of the reverse order Iudeono#P in contrast to Iudeono#Px, the monosyllabic noun P, when modified by Iudeono, would have to be invariably alliterative, and would accordingly be denied of the alternative alliteration on Iudeono, the kind of metrical license that is available to the disyllabic nominal form Px in the presence of the minimal pair Px#Iudeono and Iudeono#Px: e.g., 5259a pascha Iudeono versus 4562b Iudeono pascha. The use of Iudeono as a postnominal modifier of P, furthermore, is not without its own problems. First, whether -e- is syllabic or non-syllabic, the disyllabic ending -ono is associated with two separate positions, the penultimate and ultimate positions, respectively: 4170a uuerod Iudeono (pxxx or Pxx). Further examples are: 5057a, 5107a, 5345a, 4152b, 4850b, 4913b, 5154b, 5245b, 5326b, 5379b, 5533b, 5560b, 5639b, 5677b (Hofmann 1991: 76). Only exceptionally does the disyllabic ending -ono constitute two separate positions in their own right. Such a mode of association is confined to Iudeono, and should not be confused with the one already discussed in section 2.12.4 above, whereby the first constituent of a disyllabic ending is brought together with the immediately preceding syllable to constitute a common position, as in 2804a craftiana Crist, itself an unusual practice all the same. Well-aware of the irregularity of these verses, Hofmann (1991: 76) scans them as a special variant of type D, Da1.3a-. Further, he adduces three apparently parallel instances, all involving the same word selboro: 877a iro selboro; 880a euuar selboro; 884a iuuar selbaro. These verses, however, should be scanned as type A3, as discussed in section 2.3 above. The association of a disyllabic ending with two different positions on their own would

MOSP-Appendix 1.qxd

5/11/04

2:44 PM

Page 351

Appendix 1. Foreign names

351

disrupt its morphological integrity; this ought to be the very reason why such metrical practice was avoided elsewhere. The use of Iudeono after P#, however, would result in the violation of this well-motivated constraint. Second, under the view that Iudeono is pxxx, the configuration P#Iudeono would constitute a variant of type D, with the second lift resolved. While not totally absent in the corpus, such a configuration is extremely rare, as observed in section 3.1.4 above. Being preceded by a strong position and not followed by a stressed syllable, the position involved is hardly expected to receive resolution: resolution on the second lift is most likely when it is followed by a prominent syllable. Other things being equal, the use of P#Iudeono would thus seem scarcely in keeping with the generalisation on implementation of resolution. Yet there is a further dimension to be taken into account here, namely the relatively high frequency with which the first lift is resolved immediately before Iudeono. More specifically, we count fifteen examples of the configuration (x . . .)px#Iudeono (4170a uuerod Iudeono, 5057a, 5345a, 4152b, 4913b, 5136b, 5232b, 5245b, 5379b, 5551b, 5560b, 5639b, 5677b, 5689b, 5761b) as against ten instances of the configuration (x . . .)P#Iudeono (5107a folc Iudeono, 3685b, 3858b, 3896b, 4476b, 4850b, 5154b, 5180b, 5326b, 5533b). We can thus hardly fail to recognise a certain preference for the resolved first lift in force here. By the mechanism detailed in sections 3.1.1.2 and 3.1.1.6 above, the second of the two consecutively placed strong positions is more likely to be resolved when the preceding position itself is resolved. In the light of this syntagmatically motivated favourable condition for implementation of resolution, the occurrence of Iudeono as a realisation of the last three positions of type D would have been rendered less peculiar. Despite being less than fully satisfactory, the fact that the configuration P#Iudeono was chosen would suggest that it was the only viable sequence that was amenable to scansion; this would then imply that the remaining candidates, Iudeo#P and P#Iudeo, with the strong variant used as a prenominal or postnominal modifier, were metrically no more manageable than Iudeono#P. Following the reasoning still further would provide us an important insight into the prosodic structure of Iudeo. If this variant were identified as PXx, as would appear most plausible in view of the value of its Old English cognate Iudea as mentioned earlier, both sequences Iudeo#P and P#Iudeo would prove to be fully metrical: type E (Iudeo#P) on the one hand and type D (P#Iudeo) on the other; the resulting configuration P#PXx (PXx ⫽ Iudeo) would then count as more acceptable than the actually used P#Iudeono with its inherent metrical difficulties. The complete absence of Iudeo#P and P#Iudeo thus calls into question the identification of Iudeo as PXx in Old Saxon, pointing therewith to a short initial vowel. The plausibility of postulating a short initial vowel for Iudeo is reinforced by verses that end in this word, such as 2321a gramharde Iudeon, 2664a sô grimme under Iudeon, and 3928a gêlmôde Iudeon (further examples include 3952a, 4181a, 4700a, 4939a, 4953a, 5487a, 5540a, 5557a, 3871b). If one identified Iudeo as PXx (or Pxx), these verses would constitute type D*, rather than type A1. The resultant type D* verses like 2321a and 3928a, however, would seem aberrant as instances of this type, because the realisation of the first drop by the second element of a true rather than a quasi compound is rare in the extreme, as observed in section 2.8 above. Of related significance is that there are three instances of variation between Iudeo and Iudeono: 4946a, 696b, 2728b. In all these, the M and C manuscripts use the shorter

MOSP-Appendix 1.qxd

352

5/11/04

2:44 PM

Page 352

Appendix 1. Foreign names

and longer variants, respectively. Of particular interest, in none of these examples does Iudeo appear in verse-final position after P#. This may be viewed as further testimony for postulating a metrical restriction against the use of P#Iudeo, rather than characterising the non-existence of this collocation simply as an accidental gap. Taken altogether, then, there seems good reason for assuming that both Iudeo#P and P#Iudeo were metrically objectionable, entailing thereby that Iudeo would not have been Pxx or PXx. It would be a short step to concluding therefore that Iudeo was pxx or Px (if -e- were regarded as non-syllabic). On this interpretation, the configuration Iudeo#P, consisting of three positions, would have turned out to be too short as a legitimate verse; by the same token, the alternative sequence P#Iudeo would have been found unscannable, as it would have contained only three metrical positions with resolution on the second position, or, it would have constituted a variant of type D with the primary-stressed, second lift suspended from resolution, a configuration which is unattested elsewhere, as pointed out in section 3.1.1.3 above. On account of the resulting unmetricality, the configurations Iudeo#P and P#Iudeo would have been avoided in favour of the minimally problematic collocation P#Iudeono.

MOSP-Appendix 2.qxd

5/11/04

2:45 PM

Page 353

Appendix 2. The metre of the Old Saxon Genesis The primary objective of this appendix is to see whether and to what extent the Old Saxon Genesis (Behaghel and Taeger 1984; Doane 1991) shares the major innovative features that were introduced to the Heliand metre, as summarised in Chapter 5 of this book. To the extent that the Genesis falls short of the Heliand in gradation and harmonisation in its metrical organisation, we can hardly fail to recognise the Heliand poet’s admirable competence and ingenuity in designing and implementing metrical innovations. Since the corpus of the Old Saxon Genesis, about one-twentieth of the Heliand in size, is highly limited, however, we cannot be definitive about whatever tendencies observable in the extant data. Of central concern are the following metrical properties: identification of metrical types and their major realisation variants along with their characteristic distribution patterns (section 1); the status of anacrusis (section 2); implementation and suspension of resolution (section 3); alliterative patterns of heavy verses (section 4); CValliteration (section 5). 1. Metrical types and variants 1.1. Type A1 The absence of the configurations PSx#Px and PS#xP, as well as of the configuration PS#Px with single alliteration (either in the a-verse or the b-verse), as shown in Table A2.1, would appear to justify us in regarding the configuration PS#Px as an independent metrical type, type A2a, as in the traditional metre. The rare occurrence of this configuration even when provided with double alliteration, however, would make this characterisation less definitive. To be on the safer side, we should leave it undecided whether this configuration constitutes a metrical type on its own (as in Beowulf) or a variant of type A1 (as in the Heliand). Apart from such an indeterminate role of stress in constituting metrical types and governing their distribution, the number of unstressed syllables and the location of the boundary of the preceding lifted word seem to be responsible for the gradation pattern of the first drop of type A1 in manners analogous to the traditional metre and the Heliand. This should be hardly surprising, given that the graded patterning seems to be deep-rooted in metrical tradition, observed as it is in the traditional metre as well as in the Heliand. A further point of interest concerns verse 93a uualdanda be is faruuurohtiun. While there is no example of the configuration PXx#Px in the Genesis, this verse constitutes a longer variant of the configuration, containing as it does additional unstressed syllables after -Xx#. The attestation of such a longer variant would seem to suggest that the configuration PXx#Px itself would have been counted as metrical in the Genesis as well.

MOSP-Appendix 2.qxd

5/11/04

354

2:45 PM

Page 354

Appendix 2. The metre of the Genesis

Table A2.1. List of major variants of type A1 in the Old Saxon Genesis variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

Psx#Px Pxxx#Px Px#Px Pxx#Px

0 (0%) 1 (100%) 17 (24%) 1 (12.5%)

0 (0%) 0 (0%) 12 (17%) 0 (0%)

0 (0%) 0 (0%) 41 (59%) 7 (87.5%)

PX#Px PS#Px Pxx#Px P#xPx

1 (33%) 2 (100%) 4 (50%) 8 (57%)

Px#xPx P#xxPx PXx#Px PSx#Px PX#xPx P#xxx . . . Px PS#xPx

31 (63%) 10 (56%) 0 (0%) 0 (0%) 1 (50%) 9 (69%) 0 (0%)

0 (0%) 0 (0%) 1 (12.5%) 1 (7%) 2 (4%) 4 (22%) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%)

2 (67%) 0 (0%) 3 (37.5%) 5 (36%) 16 (33%) 4 (22%) 0 (0%) 0 (0%) 1 (50%) 4 (31%) 0 (0%)

Like the Heliand, the Genesis contains an example of the configuration Psx#px, verse 35b haramuuerek mikil, scannable as subtype A1s and type E alike (see below), alongside two instances of PS#px (subtype A1s), 30b legarbedd uuaran and 82b inuuidd mikil. 1.2. Types A2 and A3 Unlike the indeterminate configuration PS#Px noted above, the configuration Px . . . #PS should more plausibly be identified as an independent metrical type by virtue of a larger number of its attestations, five examples, all in the a-verse with double alliteration (90a, 131a, 222a, 320a, 323a). In addition, there are seven verses that are apparently open to two alternative scansions, type A2 on the one hand and type D* on the other: 13a, 135a, 161a, 164a, 190a, 248a, 280a. In contrast to the Heliand, there is no example of type A3 in which the final drop is realised by an independent word: all type A3 verses attested in the corpus (thirty instances altogether: 7a, 11a, 14a, 25a, 43a, 65a, 68a, 86a, 99a, 134a, 137a, 151a, 176a, 199a, 206a, 217a, 226a, 229a, 234a, 239a, 240a, 251a, 260a, 270a, 275a, 293a, 302a, 305a, 327a, 335a) end in a single unstressed syllable, the prototypical variant of the type. 1.3. Type B1 The type B1 verses attested in the Genesis are distinguished from those in the Heliand by their conformity to the traditional avoidance from associating the second drop with a stressed syllable: in contrast to the Heliand, the Genesis offers no instance of type B1 in which the second drop is associated with a lexical-stressed syllable, as in the

MOSP-Appendix 2.qxd

5/11/04

2:45 PM

Page 355

Appendix 2. The metre of the Genesis

355

configurations x . . . PS#P and x . . . Psx#P; nor is the configuration x . . . PS#xP attested that might alternatively be scanned as type E with anacrusis. Yet there are four examples of the configuration x . . . PXx#P, which is structurally ambiguous, scannable as type B1 and type E with anacrusis alike (section 2.4.2.3, Chapter 2): 23b uuit hebbiat unk giduan mahtigna god; 58b sô ik is nu mag drûbundian hugi, quad he; 154b habdun im sô filu fîunda barn; 256b uuas thar fîundo gimang. Since the same number of the comparable configuration is attested in Beowulf (Suzuki 1996a: 93), we have to conclude that the Genesis contains a far higher proportion of such verses, thereby coming nearer to the Heliand in this regard. The extreme rarity of the configuration PXx#P (type E; see section 1.7 below), however, would seem to add greater plausibility to scanning the configuration x . . . PXx#P as type B1 than to leaving it as purely ambiguous, type B1 as well as type E with anacrusis, as in the Heliand metre. Since type B1 in the Genesis is thus characterised by the minimal variability as regards the realisation of the second drop, the work is incapable of showing anything like a gradation pattern in this respect. The relative homogeneity of the composition of type B1 is further highlighted by the absence of verses that may be scanned as type B3 (⫻ ⫻ ⫻ /). 1.4. Type C As shown in Table A2.2, the predominance of the b-verse for the variants with the primary-stressed second lift (i.e., x . . . P#px and x . . . P#Px) is clearly ascertained, a characteristic of the traditional metre as well as of the Heliand metre. For the remaining variants, the small numbers of their attestations would make it next to impossible to say anything definite, yet the overall preference for the b-verse may be recognisable as a consistent tendency. This is a feature unknown to the metres of Beowulf and the Heliand, in which the configurations x . . . PXx and x . . . Pxx are significantly underrepresented in the b-verse (section 2.6.2, Chapter 2). Nonetheless, the distribution pattern in the Genesis, far from neatly graded as it is, seems in keeping with those in Beowulf and the Heliand insofar as the configurations with the primary-stressed second lift are distinctly encoded with maximal preference for the Table A2.2. Distribution of major variants of type C in the Old Saxon Genesis variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

x . . . P#px x . . . P#Px

1 6% 3 7%

2 3

12% 7%

14 82% 37 86%

x . . . P#P/px

4 7%

5

8%

51 85%

x . . . Psx x . . . PSx

0 0% 0 0%

2 100% 2 25%

0 6

x . . . PS/sx

0 0%

4

40%

6 60%

x . . . PXx x . . . Pxx

0 0% 0 0%

3 1

37.5% 50%

5 62.5% 1 50%

x . . . PX/xx

0 0%

4

40%

6 60%

0% 75%

MOSP-Appendix 2.qxd

356

5/11/04

2:45 PM

Page 356

Appendix 2. The metre of the Genesis

b-verse. We find no instance of type C verse that ends in a lexical-stressed syllable, -S#, in contrast to the Heliand, in which such a configuration was made possible as an occasional variant (section 2.6.3, Chapter 2). 1.5. Type D The infrequent occurrence of relevant variants would hardly allow us to draw any definite conclusion. Yet the appreciable preference for the b-verse shared by the configurations P#PX/xx and P#PS/sx may cast doubt on the legitimacy of categorically separating them as independent metrical types, the analysis that was well-founded for the traditional metre. The lack of demonstrable distinction in distribution patterning between these two configurations, however, differentiates the Genesis also from the Heliand. We may recall that the Heliand shows a fine-grained gradation rather than a lack of distinction, whereby the two configurations at issue are placed toward the opposite ends of the scale (section 2.7, Chapter 2). In the absence of evidence showing categorical differentiation, then, it would seem plausible to conclude that the traditional tripartite types D1, D2a, and D2b were grouped together as members of a single metrical type (type D) in the Genesis without demonstrable patterned variation in distribution emerging among them. Finally, we should note that the configuration P#Px . . . P, when accompanied with single alliteration, would seem ambiguous, as it may be apparently scannable as subtype D2b or type E. We count nine such verses: 31b Thuo sprak im god selbo tuo, 104b, 118b, 145b, 165b, 225b, 248b, 285b, 320b. The nearly exclusive dominance of the alliterative pattern aax for comparable heavy a-verse in this poem (see section 4 below) will not justify a resort to syntactic constituency for disambiguation such as we did in the Heliand. On the other hand, the demonstrably distinct alliterative pattern of subtype D2b (Table A2.3) and type E (Table A2.6) may motivate the purely metre-based solution to this issue that we proposed for Beowulf drawing on Bliss (1967: §80; for details, see Suzuki 1996a: 113–17). Specifically, given the exclusive use of double alliteration for subtype D2b (P#Px . . . S) in contrast to the lack of such a maximal preference for the a-verse with double alliteration on the part of type E (PSx(x)#(x)P), we may surmise that subtype D2b in general is metrically incompatible with single alliteration, and therefore scan the nine verses under consideration as type E. In view of the extremely limited corpus size, however, we have to admit that this interpretation does not command full empirical credibility. 1.6. Type D* In view of the reversal in verse distribution pattern of the configurations PPX/xx and Px . . . PX/xx as substantiated in Tables A2.3 and A2.4, a significant feature that sharply distinguishes the Heliand from Beowulf, we may infer that in the Genesis, too, the expanded configurations are treated as close to being independent. However, the proportion of subtypes D*1, D*2a, D*2b to subtypes D1, D2a, D2b, respectively, in the Genesis (Table A2.5) seems to differ markedly from that in the Heliand, in which subtypes D*2a and D*2b predominate over their shorter counterparts. On this score, the whole distribution in the Genesis is rather similar to the one in Beowulf (Suzuki 1996a: 104) by virtue of the higher frequency of the shorter types. Yet we have to recognise that all the longer subtypes are used with higher incidence in the Genesis than in Beowulf. Accordingly, in this respect the Genesis may be characterised

MOSP-Appendix 2.qxd

5/11/04

2:45 PM

Page 357

Appendix 2. The metre of the Genesis

357

Table A2.3. Distribution of major variants of type D in the Old Saxon Genesis variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

PSxx PSXx

0 0

0% 0%

0 0% 2 50%

0 2

0% 50%

PSX/xx

0

0%

2 50%

2

50%

P#Pxx P#PXx

3 0

27% 0%

2 18% 0 0%

6 55% 1 100%

P#PX/xx

3

25%

2 17%

7

P#P#px P#P#Px

0 0

0% 0%

0 0

0% 0%

1 100% 1 0%

P#P#P/px

0

0%

0

0%

2 100%

P#Px . . . P

6 100%

0

0%

0

0%

P#Psx P#PSx

1 100% 2 22%

0 0

0% 0%

0 7

0% 78%

P#PS/sx

3

30%

0

0%

7

70%

P#Px . . . S

2 100%

0

0%

0

0%

58%

as standing between the two extremes that are represented by the Heliand on the one hand and Beowulf on the other. A further notable feature bears on the characterisation of these expanded configurations, namely the composition of the extra drop. To be sure, there is no example in which this position is realised by a prefix, but it may be occupied by language materials other than a word-final unstressed syllable. The position may be filled by multiple syllables, as in verses 80a Thô geng im thanan mid grimmo hugi, 173a ik libbio bi thînum lêhene, and 294a firrian ina fon them fîundum. Of particular interest in this regard would appear verse 135a libbendian an is lîchaman, one of the seven examples that are scannable as type A2 (or A2b) or type D*, as observed in section 1.2 above in this appendix. If scanned as type D*, the extra drop in question would consist of a sequence of four syllables, the first of which is a long medial one -X-. This is absolutely disallowed in Beowulf, but allowed in the Heliand (section 2.8, Chapter 2). Regardless of the exact status of this ambiguous example, the Genesis should be regarded as closer to the Heliand in the treatment of the extra drop of the expanded configurations, in view of the occurrence of multiple syllables in several unambiguous type D* instances. In overall terms, then, we may be led to conclude that the expanded configuration in the Genesis should be identified as independent, that is, type D*, rather than reducible to the shorter counterpart as in the traditional metre. Despite the scarcity of data, the variants of type D* would seem to be differentiated in their distribution in a way that is partly comparable to, but less stratified than, the one in the Heliand: the distinction involved depends largely on whether or not the

MOSP-Appendix 2.qxd

5/11/04

358

2:45 PM

Page 358

Appendix 2. The metre of the Genesis

Table A2.4. Distribution of major variants of type D* (subtypes D*1, D*2a, and D*2b) in the Old Saxon Genesis variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

PxSxx

0

0%

0

0%

0

Px . . . Pxx Px . . . PXx Px . . . PX/xx

2 67% 2 100% 4 80%

0 0 0

0% 0% 0%

1 33% 0 0% 1 20%

Px . . . Ppx Px . . . PPx Px . . . PP/px

0 0 0

1 50% 0 0% 1 50%

1 50% 0 0% 1 50%

Px . . . Psx Px . . . PSx Px . . . PS/sx

1 100% 7 100% 8 100%

0 0 0

0% 0% 0%

0 0 0

0% 0% 0%

Px . . . Px . . . P Px . . . PxS

6 1

0 0

0% 0%

2 0

25% 0%

0% 0% 0%

75% 100%

0%

Table A2.5. Proportion of subtypes D*1, D*2a, D*2b to subtypes D1, D2a, D2b, respectively, in the Old Saxon Genesis subtype

number

percentage

D1 D*1 total

16 5 21

76% 24% 100%

D2a D*2a total

12 10 22

55% 45% 100%

D2b D*2b total

12 9 21

57% 43% 100%

penultimate drop is occupied by a stressed syllable. The differentiation seems thus insensitive to other parameters, that is, differences in syllable length and nature of stress. 1.7. Type E We seem to have a contradictory situation here which appears neither reducible to the one in the Heliand or to the one in Beowulf. On the one hand, the configuration PSx(x)#(x)P, which occurs maximally in the Genesis as in Beowulf and the Heliand, is distributed analogously to the Heliand and distinctly from Beowulf by showing preference for the a-verse with double alliteration (Table A2.6). On the other hand, the Genesis is comparable to Beowulf in that the configuration P#Px . . . P is concentrated

MOSP-Appendix 2.qxd

5/11/04

2:45 PM

Page 359

Appendix 2. The metre of the Genesis

359

Table A2.6. Distribution of major variants of type E in the Old Saxon Genesis variant

a-verse with double alliteration

a-verse with single alliteration

b-verse

PX/x#xP PXx(x)#(x)P PSx(x)#(x)P P#Px . . . P PS#xP

0 0% 0 0% 10 45% 1 10% 1 100%

0 0% 0 0% 4 18% 0 0% 0 0%

0 0% 1 100% 8 36% 9 90% 0 0%

in the b-verse. We may recall that all nine instances of this apparently ambiguous configuration is scanned here as type E on the grounds clarified in section 1.5 above. As observed in section 1.3 above in this appendix, the rare occurrence of the configuration PXx(x)#(x)P would appear peculiar, given that it constitutes a typical variant of type E next to the configuration PSx(x)#(x)P in the Heliand as well as in the traditional metre (section 2.9, Chapter 2). One might be led to surmise then that the second drop of type E in the Genesis is largely restricted to lexical-stressed syllables for association and thereby delimited as a heavy drop in a narrower form than in the Heliand and even in the traditional metre. In the Heliand, by contrast, the position at issue was reconfigured as closer to the normal drop and therewith subjected to comparable gradation patterning (section 2.9). 1.8. Type E* In the Genesis, there is no definite instance of the configurations PxPxP or PxSxP that could be identified unequivocally as nothing other than type E*. The only example that might be scanned as such is verse 244b ac he fell im after te bedu. However, this may yield to the alternative scansion as type D* (subtype D*2b), which is attested independently: 315b logna all bifeng. 1.9. Metrical types: a summary On the basis of the foregoing examination, we may postulate the following inventory of metrical types of normal verses for the metre of the Genesis. The status of the configuration / \ / ⫻ remains unclear, as represented in the parentheses, and correspondingly the designation of the related configuration / ⫻ / \ has to be provisional. (1) Inventory of metrical types in the Genesis A1: / ⫻ / ⫻ (A2a: / \ / ⫻) A2(b): / ⫻ / \ A3: ⫻ ⫻ / ⫻ B: ⫻ / ⫻ / C: ⫻ / / ⫻ D: / / ⫻ ⫻ D*: / ⫻ / ⫻ ⫻ E: / \ ⫻ /

MOSP-Appendix 2.qxd

5/11/04

360

2:45 PM

Page 360

Appendix 2. The metre of the Genesis

2. Anacrusis Anacrusis may be characterised as an optional full metrical position in the Genesis as in the Heliand, and in distinction from Beowulf, on the following grounds (Tables A2.7 and A2.8). First, anacrusis may occur in the b-verse as well as in the a-verse: of sixty-four normal verses with anacrusis, forty-one appear in the a-verse and the remaining twenty-three in the b-verse. As in the Heliand, anacrusis is thus no longer subject to the traditional restriction on verse distribution in the Genesis. Second, anacrusis also applies to the configurations that show marked preference for the b-verse, such as Px#Px (type A1; e.g., 79a thînes brôdor uurâca) and P#Pxx (type D; e.g., 105b geuuitt lînodun). Conversely, a variant with maximal preference for the a-verse with double alliteration may fail to be expanded with anacrusis, such as the configuration Px . . . PS/sx. As in the Heliand, anacrusis has thus lost the traditional privilege of attending on verses with strong predilection to double alliteration. Third, anacrusis may affect type E in the Genesis as well, a novel feature that sharply distinguishes between the Heliand and the traditional metre: e.g., 25b that ik is io bad hebanrîkean god. (The scansion of this crucial verse, however, is somewhat contestable, as noted in the following paragraph.) Fourth, anacrusis involves syllable sequences longer than disyllabic, and accordingly may concern independent words as well; the maximal number of syllables involved in anacrusis is seven, observed in two verses: 66b Nu ik ni uuelda mîna triuuua haldan; 233b huuat uuilis thu is thanna, frô mîn, duoan. Fifth, the overall distribution of the varying number of syllables for realising anacrusis in the Genesis is highly similar to that in the Heliand, as indicated in Tables A2.9 and A2.10. While it is hardly in doubt that anacrusis in the Genesis constitutes an independent metrical position as substantiated above, it is not fully identical with the practice in the Heliand in every detail. A most striking difference between the two poems seems Table A2.7. Distribution of types A1, A2, D, D*, and E with or without anacrusis in the Old Saxon Genesis*

A1 aA1 A2 aA2

a-verse with double alliteration

a-verse with single alliteration

b-verse

total

101 24

19 9% 5 13%

81 10

40% 26%

201 100% 39 100%

0% 0%

0 0

0% 0%

2 100% 3 100%

50% 62%

2 100% 3 100%

0 0

D aD

12 2

40% 20%

4 13% 0 0%

14 8

47% 80%

30 100% 10 100%

D* aD*

13 5

87% 71%

0 0% 1 14%

2 1

13% 14%

15 100% 7 100%

E aE

11 0

44% 0%

4 16% 0 0%

10 40% 4 100%

25 100% 4 100%

*Verses 164a and 280a are not included in the above list, as they are scannable as type A2(a) with anacrusis or as type D* with anacrusis, as noted in section 1.2 above in this appendix

MOSP-Appendix 2.qxd

5/11/04

2:45 PM

Page 361

Appendix 2. The metre of the Genesis

361

Table A2.8. Distribution of major variants of types A1, A2, D, D*, and E, with and without anacrusis in the Old Saxon Genesis with anacrusis

without anacrusis

total

Type A1 PxSx Px#Px Pxx#Px P#xPx PS#Px Px#xPx P#xxPx P#xxxPx P#xxxxPx

0 12 1 3 0 8 5 2 4

5 100% 61 84% 7 87.5% 11 79% 2 100% 43 84% 13 72% 4 67% 3 43%

5 73 8 14 2 51 18 6 7

100% 100% 100% 100% 100% 100% 100% 100% 100%

Type A2 Px . . . PS

3

60%

40%

5

100%

Type D PSxx PPxx PPS/sx PPPx PPx . . . S PPx . . . P

0 6 0 1 1 2

0% 50% 0% 50% 50% 20%

4 100% 6 50% 10 100% 1 50% 1 50% 8 80%

4 12 10 2 2 10

100% 100% 100% 100% 100% 100%

Type D* Px . . . Pxx Px . . . PS/sx Px . . . PPx Px . . . PxS Px . . . Px . . . P

2 1 2 0 3

40% 12.5% 100% 0% 37.5%

5 8 2 1 8

100% 100% 100% 100% 100%

Type E PXx#P PSx(x)#(x)P

0 1

0% 5%

0% 16% 12.5% 21% 0% 16% 28% 33% 57%

2

3 60% 7 87.5% 0 0% 1 100% 5 62.5% 1 100% 21 95%

1 100% 22 100%

the extremely low frequency with which anacrusis occurs on type E in the Genesis: as far as the prototypical variant of type E (i.e., PSx(x)#(x)P) is concerned, there is only a single example of anacrustic verse in the corpus, as given in the preceding paragraph. This feature seems rather comparable to the situation in Beowulf, in which anacrusis is categorically excluded from type E. In this light, one might go so far as to claim that the verse in question should be characterised as type B1, analogous to the configuration x . . . PXx#P. This analysis, however, would seem rather gratuitous, largely because there are two other instances that should be scanned as type E with anacrusis, 31b Thuo sprak im god selbo tuo and 165b endi goda selbun hnêg. In any

MOSP-Appendix 2.qxd

5/11/04

362

2:45 PM

Page 362

Appendix 2. The metre of the Genesis

Table A2.9. The varying number of anacrustic syllables in the normal verse in the Old Saxon Genesis number of verses 1 2 3 4 5 6 7

40 12 6 3 3 0 2

total

66

percentage 61% 18% 9% 4.5% 4.5% 0% 3% 100%

Table A2.10. The varying number of anacrustic syllables in the normal verse number of verses 1 2 3 4 5 6 7 8 9 10 11 total

921 268 149 80 64 39 34 10 3 2 0 1570

percentage 59% 17% 9% 5% 4% 2% 2% 1% less than 1% less than 1% 0% 100%

event, the infrequent occurrence of type E with anacrusis as indicated in Table A2.7 may warrant us in inferring that anacrusis was not as fully integrated in the Genesis as in the Heliand, and consequently not divorced completely from traditional practice in this respect. 3. Resolution Of central importance as regards resolution is to see whether the three innovations introduced to the Heliand found their way to the Genesis as well. While the scarcity of relevant examples may as well caution us in making a sweeping generalisation, the available data seems consistent with the three outstanding features that we find characteristic of the Heliand and in distinction from Beowulf. First, while the unresolved sequence PS#px is attested (30b, 82b, [35b]; on the ambiguous status of the last verse, see the following paragraph), its minimally distinct, resolved counterpart (PS#pxx) is not. Second, resolution may be suspended from the second lift of type D if it is realised

MOSP-Appendix 2.qxd

5/11/04

2:45 PM

Page 363

Appendix 2. The metre of the Genesis

363

by a secondary-stressed syllable (Psxx: 9a hebankuningas; 274a hebankuninga), whereas it has to be implemented if it is occupied by a primary-stressed one (P#pxxx: 246a is geld gereuuedi); the minimally distinct configurations Psxxx and P#pxx, by contrast, are unattested. Third, resolution is implemented on the first drop of type E, as in 55a firinuuerek gifremid, 221a godforohta gumon, 300b hebankuningas bodon. On the other hand, there are two examples which would appear to indicate that resolution is suspended from the same position, namely verses 35b haramuuerek mikil and 287b Uhtfugal sang. On the face of it, then, the Genesis might strike us as similar to Beowulf, since the first drop of type E is amenable to both operation and suspension of resolution in it. On closer inspection, however, similarity to the Heliand emerges as more significant. The authenticity of verse 287b seems somewhat in doubt because of its anomalous alliteration attested in the manuscript, Uhtfugal sang; in this light, we should probably follow Holthausen (1895: 55) in amending the verse as sang uhtfugal, a typical type D verse. This then leaves us with another problematic verse, verse 35b haramuuerek mikil, which may be scanned as subtype A1s. Thus, we have no definite example of the configuration Psx#P, which scans unequivocally as type E with the first drop suspended from resolution. Therefore, we can hardly avoid concluding that resolved configuration PsxxP is favoured at the expense of the unresolved counterpart PsxP, that is, that resolution has to be implemented on the first drop of type E. In this way, the Genesis seems to follow closely the Heliand in complying the new generalisations on suspension and implementation of resolution. 4. Alliterative patterns of heavy verses Of the seven heavy a-verses attested in the Genesis, all being of the configuration PPx . . . P, six are type D, realising the alliterative pattern aax (18a ferid ford an gimang; 49a suêt sundar ligit; 102a and 139a hêr hebanes uuard; 125a uueros uuîf undor tuisk; 178a helan holdan man), and the remaining one is type E with the alliterative pattern axa (115a godas huldi gumun). The proportion of type D to type E in the Genesis thus resembles the one in Beowulf with the proportion of thirty-three to four (Suzuki 1996a: 96, 113), in contrast to that in the Heliand, in which the difference between the two types is far less outstanding with the score of forty-one to twenty-four (sections 2.7, 2.9, Chapter 2, and 3.2.2, Chapter 3). Furthermore, all nine instances of the same configuration PPx . . . P accompanied with single alliteration are b-verses (31b, 104b, 118b, 145b, 165b, 225b, 248b, 285b, 320b), the distribution that is closely parallel to that in Beowulf, in which there is a single example of the same configuration in the a-verse as against 132 occurrences in the b-verse (Suzuki 1996a: 96, 113). By contrast, the Heliand shows a much less polarised distribution of the configuration PPx . . . P with single alliteration, namely sixteen a-verses and thirty-five b-verses (sections 2.7, 2.9, Chapter 2, and 3.2.2, Chapter 3). We find no other structurally ambiguous verses in the Genesis, such as the configurations PPPx and PPpx. The Genesis thus seems to follow the canonical alliterative pattern aax for heavy verses even more closely than Beowulf by disallowing categorically the pattern axx, although there is a single instance of the non-canonical pattern axa. Such a strict conformity to traditional practice may count as an indication that the Genesis poet was not as innovative as the Heliand poet.

MOSP-Appendix 2.qxd

5/11/04

364

2:45 PM

Page 364

Appendix 2. The metre of the Genesis

The Genesis does not offer any example in which a primary-stressed syllable, particularly the one of a class 1 content word, is used as part of the first drop of types A1 and D*, and of the final drop of type A3, which may be regarded as outstanding properties of the Heliand. 5. CV-alliteration Out of a total of 334 lines, there are 308 lines in which consonant (rather than vowel) alliteration takes place. Of these, we count 120 lines (c. 39 per cent) in which CValliteration is realised at least in two lifts (Table A2.11). This ratio is exactly the same as the one obtained in the Heliand (section 3.2.3, Chapter 3). Furthermore, out of a total of twenty-six vowel alliterations, twelve (c. 46 per cent) involve the identical vowels for alliteration at least in two lifts (Table A2.12). This figure seems rather higher than the one in the Heliand, in which the figures for the identical CV- and V- are Table A2.11. Distribution of alliterative patterns involving the identical CV- in the Old Saxon Genesis a-verse (double alliteration) first lift second lift

b-verse first lift

1 1 1 0

1 0 1 1

1 1 0 1

a-verse (single alliteration)

b-verse

1

1

total of lines involving the identical CV-

number of lines 8 39 35 23 number of lines 15 308

Key: 1 ⫽ involving identical vowel; 0 ⫽ involving non-identical vowel

Table A2.12. Distribution of vowel alliteration involving the identical vowels in the Old Saxon Genesis a-verse (double alliteration) first lift second lift

b-verse first lift

1 1 1 0

1 0 1 1

1 1 0 1

a-verse (single alliteration)

b-verse

1

1

total of lines involving the identical vowels

number of lines 0 0 0 3 number of lines 9 12

Key: 1 ⫽ involving identical vowel; 0 ⫽ involving non-identical vowel

MOSP-Appendix 2.qxd

5/11/04

2:45 PM

Page 365

Appendix 2. The metre of the Genesis

365

indistinguishable. It remains unclear, however, how we should interpret such a higher ratio obtained in the Genesis. In any event, we may be safe in concluding that the Genesis favours CV-alliteration as does the Heliand. 6. Concluding remarks As demonstrated above, the Genesis shares a number of innovative features with the Heliand. Of greater significance, however, it falls short of the consistency, systematicity, and elaborateness with which the Heliand metre is reorganised and reconfigured: the Genesis fails to show fully graded distribution patterns for major realisation variants of metrical types, notably types C, D, D*, and E; anacrusis is not as fully integrated as an independent metrical position as in the Heliand; and the alliterative pattern of heavy verses remains hardly diversified, as it is virtually limited to the traditionally canonical aax. Restated in more positive terms, however, deviating from conventional practice to a lesser extent in a number of respects, the Genesis poet, apparently lacking the Heliand poet’s innovative spirit and aesthetic sensibilities, tried to remain faithful to tradition.

MOSP-References.qxd

5/11/04

2:52 PM

Page 366

References Behaghel, Otto (ed.). 1984. Heliand und Genesis. 9th edn revised by Burkhard Taeger. Tübingen: Niemeyer Berr, Samuel. 1971. An Etymological Glossary to the Old Saxon Heliand. Berne: Lang Berron, Gottfried. 1940. Der Heliand als Kunstwerk: Vier Studien. Würzburg: Triltsch Bischoff, Bernhard. 1979. ‘Die Straubinger Fragmente einer Heliand-Handschrift’, Beiträge zur Geschichte der deutschen Sprache und Literatur 101, 171–80 Bliss, A. J. 1967. The Metre of Beowulf. Rev. edn Oxford: Blackwell Boor, Helmut de. 1979. Geschichte der deutschen Literatur. Band I. Von Karl dem Großen bis zum Beginn der höfischen Dichtung 770–1170. 9th edn revised by Herbert Kolb. Munich: Beck Bostock, J. Knight. 1976. A Handbook on Old High German Literature. 2nd edn Oxford: Clarendon Brunner, Karl. 1965. Altenglische Grammatik. 3rd edn Tübingen: Niemeyer Cable, Thomas. 1974. The Meter and Melody of Beowulf. Urbana: University of Illinois Press Campbell, A. 1977. Old English Grammar. Oxford: Clarendon Cathey, James E. 1999. ‘Interpretatio Christiana Saxonica: Redefinition for Reeducation’, Interdigitations: Essays for Irmengard Rauch, ed. Gerald F. Carr, Wayne Harbert, and Lihua Zhang, 163–72. New York: Lang Cathey, James E. (ed.) 2002. Hêliand: Text and Commentary. Morgantown: West Virginia University Press Doane, A. N. 1991. The Saxon Genesis: An Edition of the West Saxon Genesis B and the Old Saxon Vatican Genesis. Madison: University of Wisconsin Press Duncan, Edwin. 1994. ‘Metrical and Alliterative Relationships in Old English and Old Saxon Verse’, Studies in Philology 91, 1–12 Fulk, R. D. 1992. A History of Old English Meter. Philadelphia: University of Pennsylvania Press Gallée, Johan Hendrik. 1993. Altsächsische Grammatik. 3rd edn Tübingen: Niemeyer Gantert, Klaus. 1998. Akkommodation und eingeschriebener Kommentar: Untersuchungen zur Übertragungsstrategie des Helianddichters. Tübingen: Narr Genzmer, Felix (tr.). 1955. Heliand und die Bruckstücke der Genesis. Stuttgart: Reclam Green, D. H. 1998. Language and History in the Early Germanic World. Cambridge: Cambridge University Press Haubrichs, Wolfgang. 1985. ‘Altsächsische Literatur’, Europäisches Frühmittelalter, ed. Klaus von See (Neues Handbuch der Literaturwissenschaft 6), 217–36. Wiesbaden: AULA Heusler, Andreas. 1920. ‘Heliand, Liedstil und Epenstil’, Zeitschrift für deutsches Altertum und deutsche Literatur 57, 1–48 Heusler, Andreas. 1943. Die altgermanische Dichtung. 2nd edn Potsdam: Athenaion Heusler, Andreas. 1956. Deutsche Versgeschichte I. Berlin: de Gruyter Hinderschiedt, Ingeborg. 1979. Zur Heliandmetrik: Das Verhältnis von Rhythmus und Satzgewicht im Altsächsischen. Amsterdam: Benjamins

MOSP-References.qxd

5/11/04

2:52 PM

Page 367

References

367

Hirt, Herman. 1891. ‘Zur Metrik des altsächsischen und althochdeutschen Allitterationsverses’, Germania 36, 139–79, 279–307 Hofmann, Dietrich. 1982. ‘Der Name der Juden in den altgermanischen Sprachen’, German Life and Letters 35, 296–314; in Gert Kreutzer, Alastair Walker, and Ommo Wilts (eds), 416–34 Hofmann, Dietrich. 1986. ‘Die altsächsische Bibelepik zwischen Gedächtniskultur und Schriftkultur’, Settimane di studio del Centro italiano di sull’alto medioevo 32, 453–83; in Gert Kreutzer, Alastair Walker, and Ommo Wilts (eds), 528–58 Hofmann, Dietrich. 1991. Die Versstrukturen der altsächsischen Stabreimgedichte Heliand und Genesis. I. Textband, II. Verslisten. Heidelberg: Winter Holthausen, Ferdinand. 1895. ‘Zur altsächsischen Genesis’, Zeitschrift für deutsches Altertum und deutsche Literatur 39, 52–6 Holthausen, Ferdinand. 1921. Altsächsisches Elementarbuch. 2nd edn Heidelberg: Winter Howell, Robert B. 1991. Old English Breaking and its Germanic Analogues. Tübingen: Niemeyer Hutcheson, B. R. 1995. Old English Poetic Metre. Cambridge: Brewer Jones, Charles. 1989. A History of English Phonology. London: Longman Kartschoke, Dieter. 1975. Bibeldichtung: Studien zur Geschichte der epischen Bibelparaphrase von Juvencus bis Otfrid von Weißenburg. Munich: Fink Kauffmann, Friedrich. 1887. ‘Die Rhythmik des Heliand’, Beiträge zur Geschichte der deutschen Sprache und Literatur 12, 283–355 Kendall, Calvin B. 1991. The Metrical Grammar of ‘Beowulf’. Cambridge: Cambridge University Press Klaeber, Frederick (ed.). 1950. Beowulf and the Fight at Finnsburg. 3rd edn Lexington, Mass.: Heath Krahe, Hans and Wolfgang Meid. 1969a. Germanische Sprachwissenschaft I. Einleitung und Lautlehre. Berlin: de Gruyter Krahe, Hans and Wolfgang Meid. 1969b. Germanische Sprachwissenschaft II. Formenlehre. Berlin: de Gruyter Krapp, George Philip (ed.). 1932. The Vercelli Book. New York: Columbia University Press Krapp, George Philip, and Elliott van Kirk Dobbie (eds). 1936. The Exeter Book. New York: Columbia University Press Kreutzer, Gert, Alastair Walker, and Ommo Wilts (eds). 1988. Dietrich Hofmann. Gesammelte Schriften I: Studien zur nordischen und germanischen Philologie. Hamburg: Buske Kuhn, Hans. 1929. Das Füllwort of-um im Altwestnordischen: Eine Untersuchung zur Geschichte der germanischen Präfixe; Ein Beitrag zur altgermanischen Metrik. Göttingen: Vandenhoeck Kuhn, Hans. 1933. ‘Zur Wortstellung und -betonung im Altgermanischen’, Beiträge zur Geschichte der deutschen Sprache und Literatur 57, 1–109 Lehmann, Winfred P. 1953. The Alliteration of Old Saxon Poetry. Oslo: Aschehoug Lehmann, Winfred P. 1956. The Development of Germanic Verse Form. Austin: University of Texas Press Luick, Karl. 1964 [1914–21]. Historische Grammatik der englischen Sprache. Vol. I, part I. Stuttgart: Tauchnitz Lutz, Angelika. 1991. Phonotaktisch gesteuerte Konsonantenveränderungen in der Geschichte des Englischen. Tübingen: Niemeyer Martin, Ernst. 1896. ‘Zum Heliand’, Zeitschrift für deutsches Altertum und deutsche Literatur 40, 126–7 Martin, Ernst. 1907. Der Versbau des Heliand und der altsächsischen Genesis. Strassburg: Trübner

MOSP-References.qxd

368

5/11/04

2:52 PM

Page 368

References

Mayer, Chr. Aug. 1904. ‘Die vocalische Allitteration im Heliand’, Zeitschrift für deutsches Altertum und deutsche Literatur 47, 413–20 Mitchell, Bruce. 1985. Old English Syntax. Vol. II. Subordination, Independent Elements, and Element Order. Oxford: Clarendon Momma, H. 1997. The Composition of Old English Poetry. Cambridge: Cambridge University Press Murphy, G. Ronald. 1989. The Saxon Savior: The Germanic Transformation of the Gospel in the Ninth-century Heliand. New York: Oxford University Press Murphy, G. Ronald (tr.). 1992. The Heliand: The Saxon Gospel. New York: Oxford University Press Murray, Robert W. and Theo Vennemann. 1983. ‘Sound Change and Syllable Structure in Germanic Phonology’, Language 59, 514–28 Neuner, Erich. 1920. ‘Über ein- und dreihebige Halbverse in der altenglischen alliterierenden Poesie’, Doctoral dissertation, Friedrich-Wilhelms-Universität zu Berlin Petersen, William L. 1994. Tatian’s Diatessaron: Its Creation, Dissemination, Significance, and History in Scholarship. Leiden: Brill Priebsch, Robert. 1925. The Heliand Manuscript, Cotton Caligula A. VII in the British Museum: A Study. Oxford: Clarendon Prokosch, Eduard. 1939. A Comparative Germanic Grammar. Baltimore: Linguistic Society of America Rathofer, Johannes. 1962. Der Heliand: Theologischer Sinn als tektonische Form; Vorbereitung und Grundlegung der Interpretation. Cologne: Böhlau Rauch, Irmengard. 1992. The Old Saxon Language: Grammar, Epic Narrative, Linguistic Interference. New York: Lang Rieger, Max. 1876. ‘Die alt- und angelsächsische Verskunst’, Zeitschrift für deutsche Philologie 7, 1–64 Russom, Geoffrey. 1998. Beowulf and Old Germanic Metre. Cambridge: Cambridge University Press Scott, Mariana (tr.). 1966. The Heliand. Chapel Hill: University of North Carolina Press Sehrt, Edward H. 1966. Vollständiges Wörterbuch zum Heliand und zur altsächsischen Genesis. 2nd edn Göttingen: Vandenhoeck Sievers, Eduard. 1876. ‘Zum Heliand’, Zeitschrift für deutsches Altertum und deutsche Literatur 19, 1–76 Sievers, Eduard (ed.). 1878. Heliand. Halle: Waisenhaus Sievers, Eduard. 1885. ‘Zur Rhythmik des germanischen Alliterationsverses’, Beiträge zur Geschichte der deutschen Sprache und Literatur 10, 209–314, 451–545 Sievers, Eduard. 1893. Altgermanische Metrik. Halle: Niemeyer Sievers, Eduard (ed.). 1935. Heliand: Titelauflage vermehrt um das Prager Fragment des Heliand und die Vaticanischen Fragmente von Heliand und Genesis. Halle: Waisenhaus Sowinski, Bernhard. 1985. Darstellungsstil und Sprachstil im Heliand. Cologne: Böhlau Stapel, Wilhelm (tr.). 1953. Der Heliand. Munich: Hanser Stübiger, Martin. 1953. ‘Zum Verbum im Stabreimvers des Heliand’, Niederdeutsches Jahrbuch 76, 1–15 Suzuki, Seiichi. 1982. ‘Phonetic Values of Old English Vocalic Digraphs’, Linguistics 20, 323–38 Suzuki, Seiichi. 1988. ‘The Indo-European Basis of Germanic Alliterative Verse’, Lingua 75, 1–24 Suzuki, Seiichi. 1989. ‘On Determining the Sonority Value of /w/ relative to /r/ and /l/ in Early West Germanic’, Folia Linguistica Historica 10, 21–34

MOSP-References.qxd

5/11/04

2:52 PM

Page 369

References

369

Suzuki, Seiichi. 1992. ‘The Germanic Hypermetric Line in Indo-European Perspective’, Perspectives on Indo-European Language, Culture and Religion. Studies in Honor of Edgar C. Polomé. Vol. 2, 480–501. McLean: Institute for the Study of Man Suzuki, Seiichi. 1994. ‘Breaking, Ambisyllabicity, and the Sonority Hierarchy in Old English’, Diachronica 11, 65–93 Suzuki, Seiichi. 1995a. ‘Anacrusis in the Meter of Beowulf’, Studies in Philology 92, 141–63. Suzuki, Seiichi. 1995b. ‘Resolution and Mora Counting in Old English’, American Journal of Germanic Linguistics and Literatures 7, 1–28 Suzuki, Seiichi. 1995c. ‘The Decline of the Foot as a Supersyllabic Mora-counting Unit in Early Germanic’, Transactions of the Philological Society 93, 227–72 Suzuki, Seiichi. 1996a. The Metrical Organization of Beowulf: Prototype and Isomorphism. Berlin: de Gruyter Suzuki, Seiichi. 1996b. ‘On the Syllable Weight of -VC# in Old English: A Metrical Perspective’, English Historical Linguistics 1994. Papers from the 8th International Conference on English Historical Linguistics, ed. Derek Britton, 39–55. Amsterdam: Benjamins Suzuki, Seiichi. 1999. ‘Anacrusis in the Meter of Heliand’, Interdigitations: Essays for Irmengard Rauch, ed. Gerald F. Carr, Wayne Harbert, and Lihua Zhang, 189–99. New York: Lang Suzuki, Seiichi. 2000. ‘The Metrical Reorganization of Type E in the Heliand’, American Journal of Germanic Linguistics and Literatures 12, 281–90 Suzuki, Seiichi. 2001. ‘The Metrical Organization of the Heliand: Gradation and Harmonization’, Interdisciplinary Journal for Germanic Linguistics and Semiotic Analysis 6, 11–39 Taeger, Burkhard. 1970. Zahlensymbolik bei Hraban, bei Hincmar—und im ‘Heliand’? Studien zur Zahlensymbolik im Frühmittelalter. Munich: Beck Taeger, Burkhard (ed.). 1985. Der Heliand: Ausgewählte Abbildungen zu Überlieferung. Litterae. Göppinger Beiträge zur Textgeschichte 103. Göppingen: Kümmerle van Coetsem, Frans. 1996. Towards a Typology of Lexical Accent: ‘Stress Accent’ and ‘Pitch Accent’ in a Renewed Perspective. Heidelberg: Winter Vennemann, Theo. 1988. Preference Laws for Syllable Structure and the Explanation of Sound Change: With Special Reference to German, Germanic, Italian, and Latin. Berlin: de Gruyter von See, Klaus. 1967. Germanische Verskunst. Stuttgart: Metzler Weringha, J. J. fon (⫽ Weringh, J. J. van) 1965. Heliand and Diatessaron. Assen: van Gorcum Zanni, Roland. 1980. Heliand, Genesis und das Altenglische. Berlin: de Gruyter

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 370

Index to the scansion of the Heliand* 1a 1b 2a 2b 3a 3b 4a 4b 5a 5b 6a 6b 7a 7b 8a 8b 9a 9b 10a 10b 11a 11b 12a 12b 13a 13b 14a 14b 15a 15b 16a 16b 17a 17b 18a 18b 19a 19b 20a 20b 21a 21b *

A1 B1 – C A1 B1 C A1 A1 B1 E* A1 D* A1 E A1 D* B1 A3 C A1 E A1 B1 F A1 C C D* B1 A1 B1 C A1 F D F C A1 B1 A1 E

pxx#Px xxxxP#xP – xxxxxP#px Px#xxPx xxPx#P xxPSx Pxx#xpxx xPx#xxxPx xxxxPxx#px Px#Px#px Px#Px Px#P#px xxxxxPx#Px pXSx#xP xxxxPS#Px Px#pxx#P xxxxxPx#xP xxxxpxx xxxP#px Px#xpxx PXx#P P#xPx xxxxpxx#xP xxxeuangelium Px#Px xP#Px xxxxxxP#px Px#pxX#P xxxxpxx#xP pxx#Px#Px xxxxPx#xP xxP#px xpxx#Px Matheus#xxMarcus xxxxP#Pxx Lucas#xxIohannes xxxpx#Px Pxx#xxxPx xxxPX#P xpxx#xxxPx PXx#P

22a 22b 23a 23b 24a 24b 25a 25b 26a 26b 27a 27b 28a 28b 29a 29b 30a 30b 31a 31b 32a 32b 33a 33b 34a 34b 35a 35b 36a 36b 37a 37b 38a 38b 39a 39b 40a 40b 41a 41b 42a 42b

A1 B1 B1 C A3 A1 A1 C A1 B1 A1 B1 A1 B1 or E A1 B1 A1 B1 D A1 B1 A1 A1 B1 A3 A1 A1 B1 A1 E* A1 A1 D* B1 E B1 B1 A1 A1 C D* B1

Px#xPx xxpXx#px xxxPS#P xxxP#px xxxxxPx Pxxx#Px PS#xPx xxxxxxxxpx#Px xP#xxxxpXx xxPX#P Px#Px xxPx#P pXS#Px xxPXx#P P#xxPx xxxxPx#px Px#xxPx xxPx#P pXSsx pxSx xxxxPx#P Px#Px Px#xxPx xxPx#P xxxPx Px#xpxx xPx#xxxPx xxPx#xP xPx#xxxpXx PXSx#px xpx#xPx Px#Px P#xxxxxPXx xxxPx#P PX#xP xxxxxxpX#xP xxxP#xP xPx#Px px#xxPx xxxxxxpx#Px xpXx#xxxPxx xxxxxPx#px

For symbols and notations, see Symbols and Abbreviations, and section 1.4 in the text.

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 371

Index to the scansion of the Heliand 43a 43b 44a 44b 45a 45b 46a 46b 47a 47b 48a 48b 49a 49b 50a 50b 51a 51b 52a 52b 53a 53b 54a 54b 55a 55b 56a 56b 57a 57b 58a 58b 59a 59b 60a 60b 61a 61b 62a 62b 63a 63b 64a 64b 65a 65b 66a 66b 67a 67b 68a 68b 69a 69b 70a 70b

A1 B1 C A1 A1 C A1 A1 E* A1 A3 A1 D* B1 A1 A1 B1 A1 E* B1 or E B1 B1 F A1 D* A1 C A1 F A1 A1 C B1 B1 D* F F F A1 B1 F A1 A1 B1 F A1 A1 B1 F A1 C A1 F A1 B1 C

Px#xPx xxxpx#xxP xxxPsx Px#Px PX#xPx xxxxpXSx Px#Px P#xxxxP#x pxx#Px#xpx xxxPx#xxxPx xxxxPx PXSx px#xxP#px xxPx#xP PXxx#Px Pxx#Px xxxPXS pxx#xPx Px#Px#xpx xPXx#P xPxx#P xxxxPx#P Romano-Sx#xpx Px#Px Px#xpxSx Px#xPx xxxxxPxx Px#xpxx xxxRumu-S Px#xPx PxSx xxxxpxsx xPx#xP xxPx#xP Px#pxSx Erodes#P xHierusalem xxxIudeono#P xpx#xpXx xxxxPx#px xRumu-S Px#Px Px#xxxxPx xxxxxPx#xP pxx#Israheles pxxxSx px#xxxPx xxxxxxxPxx#P xRumu-S Px#Px xxxxxxPxx PxSx pxx#Israheles PXSx xxxPx#px xxxxP#Px

71a 71b 72a 72b 73a 73b 74a 74b 75a 75b 76a 76b 77a 77b 78a 78b 79a 79b 80a 80b 81a 81b 82a 82b 83a 83b 84a 84b 85a 85b 86a 86b 87a 87b 88a 88b 89a 89b 90a 90b 91a 91b 92a 92b 93a 93b 94a 94b 95a 95b 96a 96b 97a 97b 98a 98b

F E F B1 C B1 A3 F F A1 F B1 A3 D A1 B1 B1 B1 C A1 A1 C C D A1 B1 C A1 A1 B1 B1 A1 B1 C F B1 C A1 A3 E A1 D D A1 B1 A1 C C A1 C F B1 F F A1 B1

371 Erodes#xPx xxPSx#P Iudeo#Px xxxxxpxx#P xxP#px xxpXx#px xxxPx Levias#Px Iacobas#Px Pxx#Px Zacharias#xxPx xxxPx#P xxxxPx px#Pxx pXx#xxxPx xxxP#xP xxxxPx#px xxxxPxS xxxpXSx pxx#Px Px#xxxxPx xxxP#px xxxxPxx pXsXx Pxx#xxPx xxxPx#P xxPSx Px#xPx xpx#xPx xxxxPx#px xxPxS Px#xPx xxxxPx#P xxxxxP#px xxHierusalem xxxxxPx#xP xxxPSx Px#xpxx xxxxxxPx PXx#P Px#xPx pXsXx px#PXsx P#xxPx xxxxPx#px Px#Px xxxP#px xxxP#Px Px#P#xPx xxxxxP#px Zacharias#xpx xxxxPx#px xxHierusalem Iudeo#Px pxx#xxPx xxPX#P

MOSP-Index 1.qxd

5/11/04

372 99a 99b 100a 100b 101a 101b 102a 102b 103a 103b 104a 104b 105a 105b 106a 106b 107a 107b 108a 108b 109a 109b 110a 110b 111a 111b 112a 112b 113a 113b 114a 114b 115a 115b 116a 116b 117a 117b 118a 118b 119a 119b 120a 120b 121a 121b 122a 122b 123a 123b 124a 124b 125a 125b 126a 126b

2:49 PM

Page 372

Index to the scansion of the Heliand C A1 A1 C A1 A1 B1 B1 C E C F B1 C A1 B1 A1 B1 C A1 D* A1 D A1 B1 B1 A1 D* A1 B1 A1 B1 C A1 A3 B1 A1 B1 A1 C B1 B1 F C D* B1 or E A1 C A3 B1 B1 A1 A1 A1 B1 A1

xxPSx Px#Px Px#xPx xxpXsX Px#xPx xPx#Px xxxPxx#P xxxxxxPxx#P xxxP#Px xpx#Px#P xxxxpx#Px Ebreo#Px xxxPx#P xpx#Px PXx#Px xxxxxPS#P P#xxxpxx xxxxxxPxx#P xxPsx Px#Px pxx#PSx Px#Px px#PXsx Px#Px xPx#px xxPx#P Px#xPx Px#Px#P pxx#xxpxx xxxxxxxxxPx#px xxPx#Px xxxxxPx#P xxP#px pX#xPx xxxxxxPx xxPx#P(xx) PXx#Px xxxP#xP xPX#xxxPx xxxxxxP#px xxPx#P xxPx#P Gabriel#xxPx xxxpx#Px PS#xxpxSx xxxxxxxPXx#px Px#Px xxxxxxxP#px xxxxxxPx xxP#xpx xxxxPxx#px Px#Px Px#xxxxpXx Px#Px xxxxxPx#P Px#xPx

127a 127b 128a 128b 129a 129b 130a 130b 131a 131b 132a 132b 133a 133b 134a 134b 135a 135b 136a 136b 137a 137b 138a 138b 139a 139b 140a 140b 141a 141b 142a 142b 143a 143b 144a 144b 145a 145b 146a 146b 147a 147b 148a 148b 149a 149b 150a 150b 151a 151b 152a 152b 153a 153b 154a 154b

A1 B1 A1 C A3 C D B1 A1 B1 C C B1 F A1 B1 A1 B1 B1 A1 A1 B1 C A1 F B1 A1 B1 A1 B1 A1 B1 C B1 B1 A1 A1 B1 A3 A1 A1 B1 A3 A1 B1 A1 A1 B1 B1 A1 B1 B1 A1 D B1 A1

Px#xxpXx xxxxPxsx px#xPx xxP#px xxxxxPx xxxxxP#px pXsXx xxxxPx#P px#xPx xxxxPx#xpx xpx#Px xpx#Px xxxxPx#px Iohannes#xpx Px#Px xxxxxPx#P xP#xxPx xxxPx#xP xxxxPx#pX Px#Px xPx#Px xxxxxPx#px xxPsx Px#Px Zacharias#xxpXx xxxPx#P Pxx#Px xxxxxPx#xP Px#xxPx xxxxPx#P(xx) Px#xPx xxxP#xP xxxPXx xxxxxPx#xpx xxxxxPx#P Px#Px Px#xxxpXx xxxxxP#xP xxxxxxPx xpXx#Px xPx#xxxPx xxxxxxPx#xP xxxxxxpxx xPx#xPx xxPxS Px#Px Px#xxxPx xxxxPx#P xxxPx#xpx PXSx xxxxxxPx#xpx xxxxxPx#P P#xxxPx P#PSx xxxP#xpx P#xPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 373

Index to the scansion of the Heliand 155a 155b 156a 156b 157a 157b 158a 158b 159a 159b 160a 160b 161a 161b 162a 162b 163a 163b 164a 164b 165a 165b 166a 166b 167a 167b 168a 168b 169a 169b 170a 170b 171a 171b 172a 172b 173a 173b 174a 174b 175a 175b 176a 176b 177a 177b 178a 178b 179a 179b 180a 180b 181a 181b 182a 182b

C D* A2 B1 A1 A1 B1 B1 E A1 B1 A1 A3 B1 C B1 A1 C A3 C A1 C B1 A1 E A1 A1 C B1 C A1 B1 A1 B1 D* C A1 B1 A1 D A1 A1 A3 E C A1 A1 B1 A3 E A1 C A1 A1 A1 C

xxxPSx PXSxx P#xxpXS xxxxpxx#P Px#xxxxpXx xxxPx#Px xxxxPx#px xxxxxPx#xpx xxxpXsXx#px P#xxPx xxxxPx#P Px#Px xxxxxxxPx xxxxxPx#P xpxSx xxxPX#P Px#xPx xxP#Px xxxxxxPx xxxxxxxP#px xPx#xxPx xxxpx#Px xxxxPxx#px P#xPx PS#xpx Px#Px Px#xxxxpXx xxxxP#px xxxxxPx#xP xxxP#px Px#Px xxxxxPx#P xPx#xPx xxxxPx#xP Px#xpxSx xP#px Px#xPx xxPx#px Px#xxPx px#Px#P xpx#xxPx xxPxx#Px xxxxxPx PSx#P xxP#px Px#Px Px#Px xxxxxPx#xpx xxxxxPx PXx#P Px#pxx xxP#px P#xxpxx Px#Px Px#pxx xxP#px

183a 183b 184a 184b 185a 185b 186a 186b 187a 187b 188a 188b 189a 189b 190a 190b 191a 191b 192a 192b 193a 193b 194a 194b 195a 195b 196a 196b 197a 197b 198a 198b 199a 199b 200a 200b 201a 201b 202a 202b 203a 203b 204a 204b 205a 205b 206a 206b 207a 207b 208a 208b 209a 209b 210a 210b

C A1 A1 C A1 B1 A1 C A1 A1 C E C B1 A1 C A1 B1 A1 C D* C A1 B1 B1 A1 A1 D D B1 D* F B1 A1 C A1 A1 B1 A1 B1 or E A1 B1 B1 A1 A1 C A1 B1 A3 A1 B1 C A1 C C C

373 xxxPSx Px#Px Px#xPx xxxxxxxP#px xPx#xxPx xxxxxxPx#P Px#xpxx xxxxPXx Px#Px xPx#xPx xxxxxPXx PSx#P xpx#Px xxxxxxxPx#P xPx#xPx xxxxxxPXx P#xPx xxxxPx#P xPx#xPx xxxxxxP#px xPx#xP#px xxP#Px px#xxxPx xxxPxS xxPS#px pxx#Px P#xPx P#Px#P xP#Pxxsx xxPx#P P#xPx#xP Iohannes#P xPx#P P#xxPx xxP#px P#xxPx Px#xxxpxx xxxxPx#P Px#xPx xPXx#P Pxx#xPx xxxxxxxPx#P xxxxPx#P Px#Px P#xxPx xxxxxxP#px Px#Px xPx#xpx xxxxxPS Px#xPx xxxxxPx#P xxpx#Px Pxx#Px xxxP#px xxxPSx xxpx#Px

MOSP-Index 1.qxd

5/11/04

374 211a 211b 212a 212b 213a 213b 214a 214b 215a 215b 216a 216b 217a 217b 218a 218b 219a 219b 220a 220b 221a 221b 222a 222b 223a 223b 224a 224b 225a 225b 226a 226b 227a 227b 228a 228b 229a 229b 230a 230b 231a 231b 232a 232b 233a 233b 234a 234b 235a 235b 236a 236b 237a 237b 238a 238b

2:49 PM

Page 374

Index to the scansion of the Heliand A1 B1 A3 B1 C A1 A1 C A1 C A1 C A1 A1 F C A1 B1 A1 B1 B1 B1 B1 D A1 A1 E A1 B1 C A3 A1 C A1 C B1 D* C C A1 A1 B1 B1 C D* A1 B1 A1 B3 C B1 F E B1 C B1

px#xxxxpXx xxxxxPx#xP xxxxPx xxxpxx#xP xxxxxxxxPXx P#xpxx Px#Px xxP#Px xPx#xPx xxxpx#Px xP#xxxPx xxxP#px(xx) Px#Px Px#Px xxxxIohannes xpx#Px Px#Px xxxxxxpx#xxP Px#xPx xxxxPx#P xxxPS#P xxxpxX#P xxP#xxP(xx) pXsxx xxPx#xxPx xxPx#xPxx PSx#px xPx#xxPx xxxxPx#P xxxxpx#Px xxxxxxPx(xx) Px#xPx xxP#px Px#xPx xxxxxxpx#Px xxxxPx#px P#xxPsx xxxxxxxP#px xxxxPsx P#xPx px#xPx xxPx#P xxxxxP#xP xxP#Px Px#PSx PxSx xxxPxx#P Px#Px xxxxP#xP xxxxpx#Px xxPx#xpx Iohannes#px PSx#xP xxxxxxxPx#xP xxPSx xxxxxPx#xP

239a 239b 240a 240b 241a 241b 242a 242b 243a 243b 244a 244b 245a 245b 246a 246b 247a 247b 248a 248b 249a 249b 250a 250b 251a 251b 252a 252b 253a 253b 254a 254b 255a 255b 256a 256b 257a 257b 258a 258b 259a 259b 260a 260b 261a 261b 262a 262b 263a 263b 264a 264b 265a 265b 266a 266b

A1 A1 D B1 A1 C A1 B1 B1 B1 C A1 D C B1 A1 B1 A1 A3 C A1 C F F D* C E F B1 C F E* F B1 D B1 F C D* C F B1 A1 C D B1 A1 B1 A1 B1 A2 B1 A1 C D*E B1

xPx#xxPx xPx#xxxPx P#Psx xxPx#P Px#pxx xxxxPsx px#xxPx xxxxxxxPx#P xxxP#xxP xxxxxPx#P xxPSx pxx#Px P#pxSx xpx#Px xxxpxX#P px#xpXx xPx#px Px#Px xxxxxxPx xPSx px#xPx xxxPsx xGalilea-S Gabriel#px Px#xpxSx xxxxpx#Px pxSx#px Maria#xxPx xxxPx#xpx xxpx#Px Ioseph#xPx Px#Px#P xDauides#Px xxxPS#P px#PSx xxxPx#px xNazareth-S xpx#Px Px#pXSx xxxxpx#Px P#xxMaria(xx) xxxxPx#P PXx#Px xxxP#px px#Px#P xxxPx#px Px#xPx xxxxPx#px xPx#xxxPx xxxxxxxxPx#px xpx#xxxpxS xxxxPxx#px Px#xPx xxxxxpx#Px xPx#pXsXx#px xxPX#xpx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 375

Index to the scansion of the Heliand 267a 267b 268a 268b 269a 269b 270a 270b 271a 271b 272a 272b 273a 273b 274a 274b 275a 275b 276a 276b 277a 277b 278a 278b 279a 279b 280a 280b 281a 281b 282a 282b 283a 283b 284a 284b 285a 285b 286a 286b 287a 287b 288a 288b 289a 289b 290a 290b 291a 291b 292a 292b 293a 293b 294a 294b

A1 B1 E* B1 A1 B1 B1 A1 A1 B1 C B1 A1 C D* A1 B1 C D* C A1 E C D E B1 A1 C B1 B1 D* A1 C C B1 B1 C A1 A1 B1 A1 C A1 B1 D A1 B1 A1 A1 B1 A1 B1 C A1 C E

Px#xPx xPx#xpx xPx#Px#xP xxxPx#P Px#Px xxxxxpx#xpx xxxPx#px pxx#PX xxxPx#pxX xxxxPx#P(xx) xxpx#Px xxxPx#xP P#xxxxpXx xxxxxP#px Px#xpxSx xxpXx#xPx xxxPx#P xpXSx px#xP#px xxxxP#Px Px#xxxxpXx PXx#P xxxxPXx pXsXx pxx#xpx xxPxx#xP xxPx#xPx xxxxxP#px xxxPx#pX xxxxPx#px Px#xPXx P#xPx xpx#Px xxxpx#Px(xx) xxxxPXsx xxxPx#px xxxPsx xxxxPx#xPx Px#xxxxxPx xxxPx#P Px#Px xxpx#Px xP#xPx xxxxxxP#xP xpx#PXx Px#Px xPx#px xxxxPx#Px xxxPx#Px xxPxx#P xP#xxxPx xxxxxPx#xP xxxxpx#Px Px#xxPx xxxxxPxx xpxSx#P

295a 295b 296a 296b 297a 297b 298a 298b 299a 299b 300a 300b 301a 301b 302a 302b 303a 303b 304a 304b 305a 305b 306a 306b 307a 307b 308a 308b 309a 309b 310a 310b 311a 311b 312a 312b 313a 313b 314a 314b 315a 315b 316a 316b 317a 317b 318a 318b 319a 319b 320a 320b 321a 321b 322a 322b

A1 F A1 C D E A1 B1 A1 C D* B1 or E D* B1 A1 C A3 B1 D* B1 A1 B1 A1 B1 B1 F A3 D* C A1 A1 B1 B1 A1 A1 B1 C F A1 B1 A1 B1 B1 A1 E B1 A1 F A1 B1 A1 B1 A1 C A1 B1

375 Px#xpxx xxpx#Iosepes xP#xPx xxxxpx#Px xpx#PSx pXSx#P xP#xxPx xxxxxxxP#xxpx xPx#xxPx xxxxP#Px xPx#xpxSx xxxPXx#xP Px#xPsx xxxxxxxPx#P px#xxxPx xxxxxxpx#Px xxxxxPx xxxxxxxPx#P Px#Pxx xxxxPx#P Px#xpxx xxxxPx#P Px#xPx xxxxxPx#P xxPx#P Ebreo#Px xxxxxxPS px#xPxx xxxxxxPsx Px#Px P#xxxPx xxxxPx#xP xxxxPx#P Px#Px px#xxxpxx xxxxPx#P xxP#px Ioseph#xxPx Px#xxPx xxxPx#P Px#xPx xxxP#xP xxxxPx#P Pxx#Px pXsXx#px xxxxxxPx#P Px#xxxPx xxx(xx)Mariun#P Px#Pxx xxxPx#P xxpx#xxxPx xxxPx#P Px#xxxxxxpXx xxxxxpxSx P#xxPx xxxxxPsx#P

MOSP-Index 1.qxd

5/11/04

376 323a 323b 324a 324b 325a 325b 326a 326b 327a 327b 328a 328b 329a 329b 330a 330b 331a 331b 332a 332b 333a 333b 334a 334b 335a 335b 336a 336b 337a 337b 338a 338b 339a 339b 340a 340b 341a 341b 342a 342b 343a 343b 344a 344b 345a 345b 346a 346b 347a 347b 348a 348b 349a 349b 350a 350b

2:49 PM

Page 376

Index to the scansion of the Heliand D* C A1 C A1 C F D E C A1 C D* B1 A1 B1 C C E B1 A1 A1 A1 B1 B1 E E B1 A1 B1 B1 A1 F A1 D* F D* B1 D* A1 D C C A1 E A1 D* B1 D* B1 A1 A1 B1 D B1 B1

xP#xxxxPxx xxxxxxpx#Px P#xxxPx xxxxxP#px Pxx#Px xpXSx xxIesu#P px#Px#P PXx#px xxxP#Px PXSx xxxxxxpx#Px Px#xxPxS xxxxPx#px xPx#xxxPx xxxxxPx#xP xxxpx#Px xxxP#px PXx#xP xxPx#px xxxxPXSx Px#Px xPxx#xxxxPx xxxxPx#P xxPx#px PXx#P PSx#px xxxpx#xpx Px#xpxx xxxxxPx#P xxxPx#P Px#Px xxxRumu-S Px#Px xxPx#xxPXS Octauianas P#xxPsx xxxxPx#xP px#xxPxx pXx#xpxx PSXx xxxxxpxsx xxxxPsx Px#xPx xxxxxxpxSx#P xxPx#Px pxx#xxPsx xxxxxPx#px Px#xxPx#xP xxxPx#P xpx#xxPx xxP#xxPx xxxxPx#pX px#Pxx xxxxPx#xP xxxpx#xxP

351a 351b 352a 352b 353a 353b 354a 354b 355a 355b 356a 356b 357a 357b 358a 358b 359a 359b 360a 360b 361a 361b 362a 362b 363a 363b 364a 364b 365a 365b 366a 366b 367a 367b 368a 368b 369a 369b 370a 370b 371a 371b 372a 372b 373a 373b 374a 374b 375a 375b 376a 376b 377a 377b 378a 378b

C A1 E C C A1 A1 B1 A1 B1 A1 C F C A1 B1 F B1 D* A1 F B1 C D F B1 F A1 D* C A1 A1 A1 B1 F C A3 D F A1 D* A1 D* B1 A1 B1 A1 B1 B1 C C E B1 A1 D* B1

xxxPxx pxx#Px PSx#px xxxP#px xxPSx pxx#xpxx xP#xPx xxxxxxPx#P pxx#xxxPx xxxxPx#xP pxx#xxPx xxxxxxxPXx Ioseph#xPx xxP#Px PX#Px xxxxPxx#P xP#xBethleem xxxPxx#P xpxx#Psx xxxxxPxx#Px Mariun#xxPx xxxPx#P xPsx pXsXx Dauides#xPx xxxxxxxPsx#P P#xxEbreon Px#Px Px#Pxsx xxxxPXx px#xxPx Px#Px Px#xxPx xxxxxxxPx#xpx Mariun#xpxx xxP#px xxxxxPx px#Px#P xpx#xBethleem Px#PX xxxpXx#PxX px#xxPx Px#xPx#P xxxpxx#P pxx#Px xxPx#px xPx#xxxxpXx xxxxxPx#P xxxPx#P xpx#Px xxxPSx xxPSx#px xxPx#P Px#Px pxxx#Psx xxxxPx#P

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 377

Index to the scansion of the Heliand 379a 379b 380a 380b 381a 381b 382a 382b 383a 383b 384a 384b 385a 385b 386a 386b 387a 387b 388a 388b 389a 389b 390a 390b 391a 391b 392a 392b 393a 393b 394a 394b 395a 395b 396a 396b 397a 397b 398a 398b 399a 399b 400a 400b 401a 401b 402a 402b 403a 403b 404a 404b 405a 405b 406a 406b

A1 A1 A1 B1 D* E A1 C A1 B1 D A1 D* C D* B1 B1 A1 C A1 A1 A1 A1 B1 A1 C A1 B1 A1 B1 A1 C E B1 A3 A1 A1 B1 C A1 D* B1 B1 D* F A1 C A1 B1 B1 F A1 A3 B1 A1 B1

xP#xxxPx Px#PX pxx#pxx xxxxxxxPx#P Px#PSx PXx#P xP#xxxPx xxxxP#px Px#Px xxxPx#xpx P#pxXx Pxx#Px P#xPxx#P xxxxpx#Px xxpx#xxPsx xxxpxx#P xxxxPx#pX Px#xPx xxpxSx Px#Px px#xPx Px#Px px#xxPx xxxpXx#xP xPx#xPx xxxP#px Px#xxPx xxxPx#P xP#xxPx xxxxPx#P xP#xxxPx xxxxPXx px#Px#px xxxPx#P xxxxPx P#xxPx Px#xxPx xxx(xx)Pxx#P xxPSx Px#Px Px#P#px xxP#xpx xxxxPx#P Px#P#px xxxDauides#P Px#xPx xxPXx Px#Px xxxpxx#px xxxxPx#px xBethlema-S Px#PX xxxxPx xxxxPx#P Px#Px xxxxPx#px

407a 407b 408a 408b 409a 409b 410a 410b 411a 411b 412a 412b 413a 413b 414a 414b 415a 415b 416a 416b 417a 417b 418a 418b 419a 419b 420a 420b 421a 421b 422a 422b 423a 423b 424a 424b 425a 425b 426a 426b 427a 427b 428a 428b 429a 429b 430a 430b 431a 431b 432a 432b 433a 433b 434a 434b

A1 B1 A1 B1 A1 B1 A1 A1s D* C D C A1s A1 D* C A1 A1 B1 E C A1 A1 A1 C A1 A1 A1 E A1 B1 A1 B1 C D* F A1 C B1 C B1 A1 A1 C C A1 B1 A1 A1 C A1 A1 B1 B1 C A1

377 xP#xxxxPx xxxpX#xxP Px#xxpxx xxxxPx#P pXx#PX xxxxxP#xP xxxPxx#xxPx PS#px Px#pxsx xpXSx px#P#px xxpx#Px PS#px Px#Px xPx#xPXx#P xxxxpXSx Px#xxPx xPx#Px xxPxx#P pxSXx#P xxPSx Px#pxx Pxx#P#x(xxx) Pxx#Px xxPXx pxx#Px xxpx#xPx pxx#Px PSxx#px xxP#xPx xPx#px xPx#xPx xxPx#P xpx#Px PS#Psx xxxxxBethleem#px Px#Px xxP#px xxPx#P xpx#Px xxxxPx#px P#xPx Px#Px xxxP#Px xxpx#Px Px#Px xxPx#P Px#Px Px#Px xxxP#px PXx#xxxPx xxPx#Px xxxPx#P xxxxpxx#P xpXSx Px#xPx

MOSP-Index 1.qxd

5/11/04

378 435a 435b 436a 436b 437a 437b 438a 438b 439a 439b 440a 440b 441a 441b 442a 442b 443a 443b 444a 444b 445a 445b 446a 446b 447a 447b 448a 448b 449a 449b 450a 450b 451a 451b 452a 452b 453a 453b 454a 454b 455a 455b 456a 456b 457a 457b 458a 458b 459a 459b 460a 460b 461a 461b 462a 462b

2:49 PM

Page 378

Index to the scansion of the Heliand A1 B1 C A1 A1 C A1 D* A1 A1 D* A1 B1 A1 B1 C B1 A1 C F A1 B1 D B1 A1 B1 A1 A1 A1 B1 B1 A1 A1 C F C A1 B1 D* B1 F B1 A1 B1 C C B1 F F B1 E C F C A1 A1

px#xPx xP#xxP xxxpxSx Px#Px xpx#xxxPx xxxxxxxxP#px Px#xxxpxx Px#PXx xPx#xPx pxxx#Px Px#pxX#P pxx#xPx xxPxx#px Px#pxx xxPx#px xxxpx#Px xxPX#xpx Px#Px xxxpx#Px Gabriel#xP Px#Px xxxPx#xP px#Pxx xxxxxP#xP Px#xxxxpXx xxxPx#px xxxxxPXSx Px#Px xP#xxxPx xP#xxP xxxpxS#px Px#Px px#xxPx xxxxxxxP#Px xxxxxHierusalem xpx#Px PXx#xxPx xxxxPx#P xxPx#Psx xxxxxxPx#xP px#xxEbreon xxxxPX#P px#xPx xxxxPx#px xxpx#Px xpx#Px xxxxxxPx#P Ioseph#xxMaria Px#xBethleem xxxP#xP PXx#P xxxP#px xHierusalem xxxxxP#Px PXx#xxPx Px#Px

463a 463b 464a 464b 465a 465b 466a 466b 467a 467b 468a 468b 469a 469b 470a 470b 471a 471b 472a 472b 473a 473b 474a 474b 475a 475b 476a 476b 477a 477b 478a 478b 479a 479b 480a 480b 481a 481b 482a 482b 483a 483b 484a 484b 485a 485b 486a 486b 487a 487b 488a 488b 489a 489b 490a 490b

F B1 A1 D B1 A1 A1 C B1 B1 or E E* F A3 E A1 B1 A1 A1 B1 C D* C A1 C C B1 or E E B1 A1 B1 A1 A1 A1 C D* B1 A1 B1 B1 A1 E B1 A1 B1 D B1 A1 B1 A1 C D* C A1 C A1 A1

Iudeo#Px xxxxxxPx#P Px#xxpxx pXsxx xxxxxPx#xpx Px#xxpxx xP#xxPx xxxxxP#px xPx#px xxxPXx#P PXSx#px Simeon#xxPx xxxxPx PXx#P Px#Px xxxxxxxP#xpx Px#xxxxpXx xxxxPx#xPx xxPx#P xpx#Px PXx#pXsX xxxxpx#Px Px#xxPx xxxxxP#px xxxP#Px xxxxPXx#P PSxx#px xxxxxxPx#xP P#xxxPx xxxxPx#xP P#xxPx P#xPx Px#xxpxx xxxxP#px PXx#pXsX xxxPx#P(xx) Px#Px xxxxpxx#P xxxxPx#P xxxPx#Px xxxpxSx#px xxxxPx#px px#xxxxpXx xxxPx#xP px#PXx xxxxPx#xP Px#Px xxxPx#P Px#Px xxP#px Px#pxSx xxxpxSx P#xxPx xxpx#Px xPx#xxxPx Px#P#x

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 379

Index to the scansion of the Heliand 491a 491b 492a 492b 493a 493b 494a 494b 495a 495b 496a 496b 497a 497b 498a 498b 499a 499b 500a 500b 501a 501b 502a 502b 503a 503b 504a 504b 505a 505b 506a 506b 507a 507b 508a 508b 509a 509b 510a 510b 511a 511b 512a 512b 513a 513b 514a 514b 515a 515b 516a 516b 517a 517b 518a 518b

F A1 A1 E* A1 A1 D* C B1 A1 A1 A1 A1 A1 A3 A1 A1 C A1 B1 A1 C A1 B1 E* C A1 F F B1 A1 B1 D* B1 A1 A1 B1 A1 E B1 B1 A1 D B1 C D* A1 B1 B1 A1 D B1 B1 A1 D* A1

pxx#Israhelas Pxxx#Px xxPx#Px Px#Px#P xPx#P#xxpxx px#xxPx Px#PSx xxxpx#Px xxxxPXS pxx#Px xxxPx#xxxPx pxx#Px xPx#xPx xxPx#xPx xxxxPx xPx#xPx Px#Px xxx(xx)px#Px P#xxxxPx xxxpxx#P Px#Px xxxxP#px P#xxpXx xPx#xxP Px#Px#P xxxxxP#Px P#xxxpxx Anna#xxPx Px#Fanueles xxxxxPxx#P xPx#xPx xxxxPx#P xPx#xxxxpXSx xxxPx#P Px#xPx pxx#Px xxxxxxxPxsx Px#xPx px#Px#px xxxxxxxxPx#xP xxxpxx#P pxx#xPx P#Pxxsx xxxpxx#xxP xxpxSx P#xxxPxx Px#xxxpXx xxxxxP#xxP xxxxxPxx#P px#xxPx px#Pxx xxxxPx#P xxPx#P P#xPx xPxx#P#px xxxpxx#Px

519a 519b 520a 520b 521a 521b 522a 522b 523a 523b 524a 524b 525a 525b 526a 526b 527a 527b 528a 528b 529a 529b 530a 530b 531a 531b 532a 532b 533a 533b 534a 534b 535a 535b 536a 536b 537a 537b 538a 538b 539a 539b 540a 540b 541a 541b 542a 542b 543a 543b 544a 544b 545a 545b 546a 546b

A1 A1s B1 or E A1 D* B1 A1 C A1 B1 B1 A1 A1 B1 D* D A1 B1 C A1 A1 B1 B1 A1 D* C F F D* C A1 D D* B1 A1 B1 E B1 A1 B1 A1 C A1 B1 A1 C B1 A1 A1 A1 B1 C A1 C B1 B1

379 xpxx#xxxPx PS#px xxxPXx#xP xPx#Px Px#pXsXx xxxPxx#P PX#Px xxxP#px xxPXx#xPx xxxPx#px xxxxPXS pxx#Px Px#Px xxxPx#px Px#PSx px#pxxx px#xxxPx xxxPS#px xpx#Px xP#xxxxPx xpx#xxpxx xxxxxxxPx#xP xxxxxPx#P P#xPx Pxxx#PxS xxxxxxP#px xHierusalem Ioseph#xxMaria Px#PXx xxxpXsX Px#xxPx px#Pxx pxxx#Psx xxxPx#xP xPx#xxxxpXx xxxxPx#P pXsXx#px xxxxxxPx#P P#xPx xxxxxxxpXx#px xPx#xPx xxxxxPsx Px#xPx xxxxxPx#P xPx#xxxPx xxxpx#Px xxPx#px Px#Px P#xxxPx Px#Px xPx#P xxxP#px xxxxxPx#Px xxxxxP#px xPx#px xxxPx#P

MOSP-Index 1.qxd

5/11/04

380 547a 547b 548a 548b 549a 549b 550a 550b 551a 551b 552a 552b 553a 553b 554a 554b 555a 555b 556a 556b 557a 557b 558a 558b 559a 559b 560a 560b 561a 561b 562a 562b 563a 563b 564a 564b 565a 565b 566a 566b 567a 567b 568a 568b 569a 569b 570a 570b 571a 571b 572a 572b 573a 573b 574a 574b

2:49 PM

Page 380

Index to the scansion of the Heliand A1 B1 F A1 C E A1 B1 A1 C A1 B1 C A1 A2 B1 A1 B1 A1 B1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1 A3A1 A3 D E B1 D* A1 D* C A1 B1 E B1 A1 C A1 B1 A2 A1 or E D* B1 A1 C B3 A1 A3 A1

px#xxPx xxxpx#xpx xxErodesan#P Px#Px xxpx#Px PSx#pX PXx#xxPx xxxxPx#P xPx#xxxPx xpXSx pxx#xxPx xxxPxx#P xxxPXx Px#xPx px#xxxPS xxxxxPx#P xpx#xxxpxx xxxxxPx#px xpx#xPx xxxxxPx#P Px#xPx#Px xxxxxxxpxxxSx Px#xPx#Px xxxxxxpxx#xPx Px#xPx#Px xxxxxxxPx#Px xPx#xxPx#Px xxxxxPx#Px xxxPx#Px xxxxxxxPx#pxx xxxxxxPx px#PXx PSx#px xxxPx#px(xxx) Px#PXx Px#xPx xPx#PSx xxxxxxxxP#px xPx#xxxxPx xxxxpXx#P PSx#px xxPx#xpx Px#xPx xpXsXx Px#Px xxxxxPx#P P#xxpxS P#xP#x or P#P#xP xxPxx#Px#px xxxxxPx#P Pxx#xPx xxxxxP#px xxxxxpx Px#Px xxxxxPx P#xPx

575a 575b 576a 576b 577a 577b 578a 578b 579a 579b 580a 580b 581a 581b 582a 582b 583a 583b 584a 584b 585a 585b 586a 586b 587a 587b 588a 588b 589a 589b 590a 590b 591a 591b 592a 592b 593a 593b 594a 594b 595a 595b 596a 596b 597a 597b 598a 598b 599a 599b 600a 600b 601a 601b 602a 602b

E C A1 C C E D* C B1 A1 A1 B1 D* B1 A1 C A1 B1 A1 C A3 A1 B1 A1 B1 C B1 A1 C A1 E B1 B1 A1 A1 C A1 B1 D* D A1 B1 C C A1 B1 D* A1 D A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 B1

PXx#P xxxxxP#px xPx#xPx xxpx#Px xpx#Px pxSx#xP xPx#Px#P xxP#Px xxxPx#P Px#Px PxSx xxxPx#P Px#PSx xxPx#P xP#xxxxpXx xxxxxxxxxxxPsX Px#xxPx xxxPXS xPx#xPx xxxxxxxP#px xxxxxxpXx Px#Px xxPXsx Px#xxpxx xxxxxPx#px xxxPXx xxxPXS Px#xPx xxxxPxx P#xxPx pXSx#P xxxxxxPx#P xxxPx#xxpx Px#pxx xxxP#xxxPx xxxxpx#Px P#P#xxxPx xxPx#P xP#xxxxPxx P#Pxx xpx#Px#Px xxpxx#P xxxxPxx xxpx#Px Px#xxxxpXx xxxxxPx#P px#xP#px xpX#xxPx xpx#P#xxP xxxxxPx#Px Px#xpxx#Px xxxxxPx#Px Pxx#Px#Px xxxxPx#xpxx Px#xxPx#Px xxxxxxxxPx#px

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 381

Index to the scansion of the Heliand 603a 603b 604a 604b 605a 605b 606a 606b 607a 607b 608a 608b 609a 609b 610a 610b 611a 611b 612a 612b 613a 613b 614a 614b 615a 615b 616a 616b 617a 617b 618a 618b 619a 619b 620a 620b 621a 621b 622a 622b 623a 623b 624a 624b 625a 625b 626a 626b 627a 627b 628a 628b 629a 629b 630a 630b

A1A1 A3A1 B1C A3A1 A1A1 A3A1 F A1 A1 C A1 B1 C A1 E A1 A1 B1 F A1 C A1 A3 E A1 B1 C E A1 B1 C A1 E B1 D B1 F B1 E C B1 C E A1 F A1 D C D* B1 F C D* A1 B1 A1s

px#xxPx#Px xxxxxxxPxx#Px xxxxPx#xpx#Px xxxxxxPx#Px xxpX#xxxxPXSx xxxxxxxxxxxxxxPx#xPx xxErodesa Px#Px P#xPx xxxxpx#Px px#xPx xxxPx#P xxxpXSx Px#Px Pxxx#pX Px#Px Pxxx#xxxxPx xxPx#P xxxxHierusalem Pxx#Px xxxPXx Pxx#Px xxxxxPx PSx#P Px#xPx xxxxxPx#P xxPSx PSx#P pX#xxPx xP#xpx xpXSx Px#Px pxsxx#P xxxxxP#xpx xpx#PSx xxxxPx#px xxxxxBethleem#xpx#Px xxxxxPx#xpx PSx#xpx xxPsx xxPx#px xpx#Px pxSx#P P#xPx xxxxBethleem Px#Px P#Px#P xxP#px Px#Psx xPx#P Iudeono#Psx xxxpx#px Px#xxPXS pxx#Px xxxxxP#xxP PS#pX

631a 631b 632a 632b 633a 633b 634a 634b 635a 635b 636a 636b 637a 637b 638a 638b 639a 639b 640a 640b 641a 641b 642a 642b 643a 643b 644a 644b 645a 645b 646a 646b 647a 647b 648a 648b 649a 649b 650a 650b 651a 651b 652a 652b 653a 653b 654a 654b 655a 655b 656a 656b 657a 657b 658a 658b

E A1 C A1 A1 B1 B1 A1 E A1 A1 C D* C C A1 B1 E C F A1 B1 A3 C A1 B1 A1 C A1 B1 A1 B1 A1 B1 D* B1 A1 B1 A1 C D* C A1 B1 C B1 B1 C C B1 A1 A1 D* B1 A1 A1

381 xxPsxx#P xPx#Px xxxpxSx Px#Px Px#xpxx xxxPxx#xxP xxxPXsx PX#xPx xxpXSx#px Px#Px Px#xpxx xxxxxxxpx#PX xPx#xxPSx xxxxxxxP#px xxxxxPXx P#xxPx xxxPx#px xxxpX#Px#xP xxPSx Px#Iudeono xPx#Px xxxxxPx#P xxxxxPx xxxxpX#Px Px#xxPx xxxxxxxxxPx#P px#xxPx xxxxxxpx#Px Px#Px xxPX#P Px#xxPx xxxxPx#P xPx#xxxPx xxxPx#P xPx#P#px xxxxxPx#P Px#xxPx xxxxPx#P Px#xpXx xxxxP#px Px#xPsx xxxP#px Px#Px xxxxxxxPx#P xxxxP#Px xxxPx#xP xxxPx#px xxpx#Px xxxxxPSx xxxxPx#P xxxPx#pxx xxxxPx#Px xPx#xxPx#px xxxxPx#px xP#xxxxpXx xpx#xxPx

MOSP-Index 1.qxd

5/11/04

382 659a 659b 660a 660b 661a 661b 662a 662b 663a 663b 664a 664b 665a 665b 666a 666b 667a 667b 668a 668b 669a 669b 670a 670b 671a 671b 672a 672b 673a 673b 674a 674b 675a 675b 676a 676b 677a 677b 678a 678b 679a 679b 680a 680b 681a 681b 682a 682b 683a 683b 684a 684b 685a 685b 686a 686b

2:49 PM

Page 382

Index to the scansion of the Heliand D* A1 A3 E D A1 A1 B1 A1 B1 A1 B1 A1 B1 A1 A1 B1 A1 D* C C D E A1 D A1 D* C A1 C A2 C B1 A1 A1 B1 A1 B1 A1 D D* C E A1 A1 B1 A1 B1 B1 A1 A1 B1 F A1 E B1

Pxx#pXSx xpxx#xPx xxxxPx PSxx#P P#Px#px P#xpxx Px#xPx xPx#xxP P#xxxPx xxPxx#P pxx#xPx xxxxxP#xP xPx#xPx xxxxPx#px Px#xxxPx xxPx#xPx xxxpxS#px Px#Px PXx#pXsX xxxxP#Px xxxpx#Px px#PXx PSxx#P P#xPx xpx#PX#P xPx#Px xxPx#xPsx xxxxxpXSx Px#Px xxxxpx#Px P#xxPS xpx#Px xxPx#xP xP#xxpxx Px#xxxPx xxxxxPx#P pxx#xPx xxxxxxpXx#P Px#xPx PSxx px#xPsx xxpx#Px PXx#xP px#xPx xP#xxPx xxxPx#P PX#Px xxxxxxxxPx#xpx xxxxxPx#P Px#Px pxx#xxPx xxxxPx#P Erodesan P#xPx PXx#pX xxPx#px

687a 687b 688a 688b 689a 689b 690a 690b 691a 691b 692a 692b 693a 693b 694a 694b 695a 695b 696a 696b 697a 697b 698a 698b 699a 699b 700a 700b 701a 701b 702a 702b 703a 703b 704a 704b 705a 705b 706a 706b 707a 707b 708a 708b 709a 709b 710a 710b 711a 711b 712a 712b 713a 713b 714a 714b

A1 B1 A1 A1 E C A1 D* D* B1 A1 B1 D C D* B1 A1 C D* F B1 D E A1 A1 C E F A1 E D C A1s A1 A1 F A1 B1 C A1 A1 C A1 B1 C A1 A1 B1 F B1 A1 C A1 C B1 C

Px#xxxxpXx xxxxxPx#P Px#xxpxx Px#xPx PXx#P xxxP#px Px#xxxPx Px#pxSx Px#pXsX xxxxxPx#P xPx#xPx xxxxxxxxxPx#px xxP#Px#xP xxxxxP#px Px#PXx xxxxPx#px Px#xPx xxxP#Px xPx#px#Px xxxxxIudeo#pXx xxxPx#xP px#PXx PSxx#P Px#PX xPx#xxxxxPx xxxxxxPXx px#Px#px Iosepe#xPx Px#xxPx PXx#xP px#Pxx xxP#px PS#pX Px#Px Px#xPx xxxxxxAegypteo P#xPx xxxxxPx#px xxpx#Px xxxxxPx#Px px#xxxxpxx xxxP#px Px#Px xxxPxx#P xxxxPsx Px#Px Px#Px xxxPx#xP Ioseph#xxPsx xxxpx#xP P#xPx xxxxxxP#px xpx#xxxPx xxxP#Px xxPx#P xxxP#px

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 383

Index to the scansion of the Heliand 715a 715b 716a 716b 717a 717b 718a 718b 719a 719b 720a 720b 721a 721b 722a 722b 723a 723b 724a 724b 725a 725b 726a 726b 727a 727b 728a 728b 729a 729b 730a 730b 731a 731b 732a 732b 733a 733b 734a 734b 735a 735b 736a 736b 737a 737b 738a 738b 739a 739b 740a 740b 741a 741b 742a 742b

A1 B1 F B1 D* A1 A1 B1 C A1 A1 B1 D C A1 B1 A1 B1 A1 B1 D* B1 A3 A1 D* B1 F B1 A1 B1 C A1 F C E A1 A1 B1 D* B1 E* A1 E A1 A1 C A3 B1 B1 A1 A1 C A1 A1 D* A1

PXx#xPx xxP#xxP Erodes#xpX xxxxPx#P xPx#xPx#P Px#xPx Px#xxxPx xxxxxPx#P xxxxxxPXx P#xPx Px#xxPx xxxxxPx#px P#PXx xxxxxP#px pxx#xPx xxxPx#P P#xxxxPx xxxxxxpxx#P Px#xPx xxxPx#P P#xPXxsx xxxPx#P xxxxxxxxPx P#xPx P#xxxxpxsx xxxPx#xP Erodes#xxxPx xxxPx#px pX#xxPx xxxPx#xpx xxxPsx Px#xPx xxxP#xxBethleem xxxxxxpx#Px xP#Px#xpx Px#pxx xpXx#xPx xxxxxxxPX#P Px#Pxx#P xxP#xP PXSxx#xP Pxx#Px PSxx#P pxx#Px Px#pxx xxxxxpx#Px xxxxxxxPx xxxxxPx#P xxPx#P Px#xPx P#xxPx xxxxxxxP#px xpx#xxxPx Px#xPx Px#xPsx Px#Px

743a 743b 744a 744b 745a 745b 746a 746b 747a 747b 748a 748b 749a 749b 750a 750b 751a 751b 752a 752b 753a 753b 754a 754b 755a 755b 756a 756b 757a 757b 758a 758b 759a 759b 760a 760b 761a 761b 762a 762b 763a 763b 764a 764b 765a 765b 766a 766b 767a 767b 768a 768b 769a 769b 770a 770b

D* A1 E A1 E F A1 B1 A1 B1 A1 A1 F B1 E A1 A1 D B1 C A2 C A1 B1 A1 B1 F A1 F B1 A1 C A1s A1 D* B1 A1 B1 F B1 D* B1 D* F D* D F F A1 C F A1 F E D C

383 pxx#pXS#px Px#pxx pxSx#P xPx#Px PSxx#P px#xxBethleem Px#PX xxxxxPx#xP px#xPx xxxxxxPxx#P Px#xxxxpXx Px#pxx Px#xBethleem xxxxxP#xpx PSx#P Px#Px Px#xxxPx xpx#Pxx xPxx#px xxpx#PX xP#xxPS xxPXx P#Px#xPx xxxxxPx#P xpxx#xxxPx xxxPx#px xAegypteo#P Px#xPx px#xIosepe xxxPx#P xPxx#Px xxpx#Px PS#px P#xPx Px#pxXx xxpxS#px pxx#xPx xxP#xP Erodes#xxpX xxxxPx#P Px#Px#P xxxxxPx#xP Px#xPxS xxArchelaus Px#pxsx PsXxx xxxxxHierusalem Iudeono#Px pxx#xPx xxP#px xxEgypti pxx#Px xxxxIosepe px#Px#P px#Pxx xxxxxP#px

MOSP-Index 1.qxd

5/11/04

384 771a 771b 772a 772b 773a 773b 774a 774b 775a 775b 776a 776b 777a 777b 778a 778b 779a 779b 780a 780b 781a 781b 782a 782b 783a 783b 784a 784b 785a 785b 786a 786b 787a 787b 788a 788b 789a 789b 790a 790b 791a 791b 792a 792b 793a 793b 794a 794b 795a 795b 796a 796b 797a 797b 798a 798b

2:49 PM

Page 384

Index to the scansion of the Heliand A1 B1 F B1 A1 C A1 B1 D A1 F A1 A1 C A1 B1 A1 B1 F F D* D F B1 or E A1 B1 C B1 A1 B1 A1 C A1 C F F C A1 A1 C F F A1s F A1 C D C A1 B1 D* B1 A1 B1 D* B1

Px#xPx xxxxP#xpx(xx) Erodes#xpX xxxxPx#P Px#xpxx xxxxpx#Px xP#xxxxPx xxpX#xpx P#pXSx P#xPx Iosep#px#Px pxxx#xxPx xpx#xxxPx xxpx#Px Px#xxPx xxxPx#xpx PXx#Px xxxxxxxPx#xP xxxxxxxGalilea-S Ioseph#xxMaria Px#PXx pXsXx xxxxNazareth-S xxPXx#P P#xxxpxx xxPx#P xxxxP#px xxPx#P PXSx xxxxxPx#xP xpx#xxxxPx xxPsx pxx#Px xxxP#px xxxxHierusalem Iuðeo#Px xxPsx Px#Px Px#xPx xxxxxxP#Px xxHierusalem Iudeono#xPx PS#px xMaria#P P#xxPx xxxxpx#Px px#Px#P xxxP#Px Px#xxpxx xxxxxPx#xP xPx#xxxPSx xxxxxxPx#px px#xxxPx xxxxxPx#xP Px#P#px xxxxPx#P

799a 799b 800a 800b 801a 801b 802a 802b 803a 803b 804a 804b 805a 805b 806a 806b 807a 807b 808a 808b 809a 809b 810a 810b 811a 811b 812a 812b 813a 813b 814a 814b 815a 815b 816a 816b 817a 817b 818a 818b 819a 819b 820a 820b 821a 821b 822a 822b 823a 823b 824a 824b 825a 825b 826a 826b

A1 B1 A1 B1 D* E A1 B1 F A1 A1 B1 A1 D C F C B1 A1 B1 B1 C D* C A1 B1 A3 D* D B1 B1 A1 B1 D A1 A1 B1 C A1 B1 A1 C A1 B1 A3 D* A1 B1 D A1 C B1 A1 C A3 B1

xPx#xPx xxxxxxxxpxx#P Px#xxxPx xP#xxP P#xPx#px pXSx#P Px#Px xxxxxxPx#xP xMariun#P P#xPx Px#xxxxPx xxxPxx#P xP#xxxPx px#Pxx xpx#Px xxxxxxxHierusalem xxpx#Px xxxxPx#P xxPx#Px xxPx#P xxPx#px xpx#Px Px#xxPxx xxP#Px Px#xxxPx xxxxpX#xP xxxxPx Px#P#px P#pxSx xxxxxxxPx#P xxPx#P Px#Px xxPxx#P pxSSx Pxx#Px xPxx#Px xxxxPX#P xxxpx#Px xxPx#xPx xxxxPx#P Px#xxxxPx xxxxpx#Px Px#xxxpxx xxxxxPx#P xxxxxxxPx Px#PXx xpx#xxxPx xxxxPXS px#PSx Px#Px xxxxPSx xxxxxxP#xpx Px#Px xxP#px(xx) xxxxPx xxxPx#P

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 385

Index to the scansion of the Heliand 827a 827b 828a 828b 829a 829b 830a 830b 831a 831b 832a 832b 833a 833b 834a 834b 835a 835b 836a 836b 837a 837b 838a 838b 839a 839b 840a 840b 841a 841b 842a 842b 843a 843b 844a 844b 845a 845b 846a 846b 847a 847b 848a 848b 849a 849b 850a 850b 851a 851b 852a 852b 853a 853b 854a 854b

A1 C B1 A1 A1 B1 A1 F A1 C A1 D* F F A3 D A1 C A1 B1 B1 B1 D A1 C A1 B1 or E C A1 C A1 B1 A1 A1 A1 A1 A1 B1 B1 A1 A1 D* A2 B1 B1 A1 B1 C A1 B1 A1 C B1 A1 A1 A1

px#xPx xxP#px xPx#px xP#xxPx xpx#xxPx xxxxxP#xP xPx#xxPx Maria#P#xP xP#xxxPx xxxxxxxxxP#px Pxx#Px xpx#xxP#px xHierusalem Ioseph#xxMaria xxxxxPx px#Pxx xxxPx#Px xxxxxpx#Px px#xPx xxxxPx#P xPx#px xxxxxPx#P px#Px#P pxXSx xxPSx Px#Px xxxxxPXx#xP xP#px Px#Px xxxxpx#Px xP#xxxxpXx xxxxxPx#P xPx#xxxxPx Px#Px xxxxPx#Px Px#Px Px#xxPx xxPx#P xxxxPXS Px#Px Px#xxxPx Px#P#px P#xxPS xxxxxxPx#P xxPx#px xxxxxxxPx#Px xxPx#xP xxxxxP#Px px#xxxxPx xxxxxPx#P Pxx#Px xxxxxP#px xxxxxxxxPXS Px#Px Px#xPx xxxxxxxPx#Px

855a 855b 856a 856b 857a 857b 858a 858b 859a 859b 860a 860b 861a 861b 862a 862b 863a 863b 864a 864b 865a 865b 866a 866b 867a 867b 868a 868b 869a 869b 870a 870b 871a 871b 872a 872b 873a 873b 874a 874b 875a 875b 876a 876b 877a 877b 878a 878b 879a 879b 880a 880b 881a 881b 882a 882b

D* A1 A1 C C A1 A3 A1 F C D* B1 C E D B1 A1 B1 C A1 E* F B1 C B1 A1 C A1 C A1 C A1 D* B1 A3 A1 A3 F A1 B1 C A1 C A1s A3 A1 A3 E A1 C A3 A1 A1A1 A3A1 A1 B1

385 Px#Px#px Px#xxpxx Px#xxxPx xxxxxP#px xxxxPSx xPx#xPx xxxxxPx Px#Px xxxIohannes xxpXSx xPx#xxxxPXx xxxpxx#xP xxxxxPxx pxSx#px px#Pxx xxPx#xP Px#xPx xxxPx#px xxxPXx P#xpxx PS#Px#px xxJohanne#xP xxPx#px xxxP#px xxxxPXS Px#Px xxxPSx Px#Px xxxpXSx pxx#Px xxPsx Px#xPx pxx#PsX xxxPx#px xxxxxxPx Px#Px xxxxPx xxJordan#P px#xPx xxxpxx#xxP xxxPsx xPx#Px xxxPXx pXS#px xxPxx Px#Px xxxxPx(xx) pXSx#P xxxPx#Px xxxxxxxPsx xxPxx Px#Px Px#xxxxxPx#pxx xxxxPx#Px Px#xxxxPx xxxpxx#P

MOSP-Index 1.qxd

5/11/04

386 883a 883b 884a 884b 885a 885b 886a 886b 887a 887b 888a 888b 889a 889b 890a 890b 891a 891b 892a 892b 893a 893b 894a 894b 895a 895b 896a 896b 897a 897b 898a 898b 899a 899b 900a 900b 901a 901b 902a 902b 903a 903b 904a 904b 905a 905b 906a 906b 907a 907b 908a 908b 909a 909b 910a 910b

2:49 PM

Page 386

Index to the scansion of the Heliand D* B1 A3 A1 B1 A1 A1 C A1 B1 A3 A1 B1 B1 B1 B1 B1 A1 A1 B1 A1 B1 A3 B1 D* B1 C C A1 A3A1 B1 A3A1 A1A1 B3A1 A1A1 B3A1 ED A3A1 D* A3A1 D*E A1 B1 A1 D B1 A1 B1 A1 B1 B1 A1 A1 B1 F F

xPx#PSx xxxxPx#xpx xxPxx Px#xPx xxxxPxS Px#Px Pxx#xPx xxxxxP#px Px#xPx xxxxxPx#P xxxxPx xpx#xPx xxxPx#P xxxPxx#px xxxPxx#P xxPx#xxP xxxxxPx#xpx PxSx Pxx#pxx xxxxxPx#P Px#xxxxpXx xxPx#px xxxxPx xxxxPx#px xPx#P#px xxxxPsx#px xxxpX#px xxxxxxP#Px Px#xxPx xxxxxxxPx#Px xPx#px xxxxxPx#xPx px#xP#xPx xxxxpx#xxPx P#xxpxxx#Px xxxxP#xPx Px#xpx#PXx xxxxPx#xPx Px#pXsXx xxxxxPx#Px P#xxpxSxx#px Px#pxx xxPx#P Px#Px px#PSx xxPX#P Px#Px xxxxxxPx#xP Pxx#Px xxxxPx#P xxxxpxx#P pxx#xpxx Px#xxxPx xxxxxPx#px xHierusalem Iudeo#Px

911a 911b 912a 912b 913a 913b 914a 914b 915a 915b 916a 916b 917a 917b 918a 918b 919a 919b 920a 920b 921a 921b 922a 922b 923a 923b 924a 924b 925a 925b 926a 926b 927a 927b 928a 928b 929a 929b 930a 930b 931a 931b 932a 932b 933a 933b 934a 934b 935a 935b 936a 936b 937a 937b 938a 938b

A1 C B1 A1 D C F B1 D C D C A1 A1 C A1 A1 C F C A1 C B1 B1 B1 C D* B1 A1 B1 B1 E A1 C A1 C E C F D C A1 A1 C A1 B1 A1 B1 D* B1 A1 B1 A1 B1 A1 B1

px#xxxPx xxxxxxxxxxP#px xxPx#P(xxx) Px#Px px#PSx xxxxxxxpX#px Iohannes#xxpXx xxxPx#P xpx#PSx xxx(xx)xP#px P#PX#P xxxxxxP#Px Px#Pxx xpxx#Px xxxxPXx Px#Px px#xxxPx xxxxxxP#px xxxxElias xxxPsx P#xxxxxpxx xxPsx xxxxPXS xxxxxPx#P xxPx#px xxP#Px Pxx#Psxx xxxxxpxx#xP Px#xPx xxxpx#xP xxxPXS P#Px#px Px#xPx xxxxPXx px#xxxxxPx xxxxpxsxx Pxx#xP xxxP#px Iohannes#xPx P#PSx xxpxsx Px#Px Px#Px xxxP#px xpx#xxxPx xxxxxPx#xP Px#Px xxxxxxxPx#px pxx#xxxxPXx xxxxPx#xP Pxx#Pxx xxxxPx#xP xPx#xxxPx xxxpxx#P px#xxxxxpXx xxxPx#xP

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 387

Index to the scansion of the Heliand 939a 939b 940a 940b 941a 941b 942a 942b 943a 943b 944a 944b 945a 945b 946a 946b 947a 947b 948a 948b 949a 949b 950a 950b 951a 951b 952a 952b 953a 953b 954a 954b 955a 955b 956a 956b 957a 957b 958a 958b 959a 959b 960a 960b 961a 961b 962a 962b 963a 963b 964a 964b 965a 965b 966a 966b

A3 C D* A1 B1 B1 A1 B1 A1 B1 A1 B1 C B1 D* C E A1 A2 C C C A1 D F F F A1 A1 A1 B1A1 B1 A1 C A1 A1 D C B1 A1 A1 B1 F D D* A1 A1 E A3 A1 A1 B1 F F B1 A1

xxxxxxxPx xxxxP#Px xxPxx#Pxx xPx#xPx xxxxxxpxx#xP xxpx#xpx Px#xxPx xxPx#xP Px#xxxxpXx xxxxPx#px Px#xxxPx xxPx#P xxxxpx#Px xxPxxS xPx#Pxx#P xxxxxP#px PSx#P Px#Px Px#pXS xxxxxpx#Px xxxP#px xpx#Px Px#xPx px#Pxx xxBethania P#Israheles Px#xxIohannese pXx#xPx Px#xPx xxxxxPx#xPx xPx#xpx#xpxx xxxxxPx#P Pxx#Px xxxxxxP#px Px#Px pXSx#Px(xx) px#pxx#xP xxpx#Px xxxxxPX#px Px#xPx Px#xPx xxxP#xP xxxGalilea#xP px#Px#P PS#Pxx#px Px#Px P#xxxxPx PXx#P xxxxxxPx Px#Px Px#xxpXx xxxxPx#P xIohannes xIordana#Px xxPx#P Px#pxx

967a 967b 968a 968b 969a 969b 970a 970b 971a 971b 972a 972b 973a 973b 974a 974b 975a 975b 976a 976b 977a 977b 978a 978b 979a 979b 980a 980b 981a 981b 982a 982b 983a 983b 984a 984b 985a 985b 986a 986b 987a 987b 988a 988b 989a 989b 990a 990b 991a 991b 992a 992b 993a 993b 994a 994b

D* B1 A1 C B1 B1 C F A3 A1 A1 B1 D* E D* B1 A3 B1 C E C F E* A1s A1 B1 D* A1 A1 B1 C D A1 A1s D B1 D* B1 C A1 C A1 A1 CA1 D*A1 B3A1 A1A1 A3A1 A1A1 A3A1 A1A1 D A1A1 A3A1 F C

387 Px#PSx xxxxxPx#xP Px#Px xxxxpx#Px xxxPx#xP xxxxxxxPx#P xxP#px Iohannes#xPx xxxxxxxxPx Px#P#x Psxx#Px xxxxxPxx#P xxxxxxpXx#PxX P#Px#xP PX#PSx xxxxxxxPx#xP xxxxxxPx(xx) xxxPx#xpx xxPXx PSx#P xpx#Px Iohannes#P Px#Px#P PS#px px#xpxx xxxPX#P Px#pXsX Px#Px xxxxpx#xPx xxxxxpx#xP xP#Px P#P#xP px#xxPx pxS#px P#Px#P xxxxP#xP xxpx#xpxx#px xxxxPxx#P xxpxSx pxx#xPx xxxxPSx Px#Px PSxx#Px xxxxxxxxPxx#Px pxx#xxxxPXx#Px xxxxP#xpxx P#xxPx#pxx xxxxxPX#Px Px#xxxpXx#Px xxxxxxpxx#Px Px#xPx#Px xxxxpx#Pxx P#xxxxpxxx#Px xxxxxxxxxPx#PX xxxIohannes#P xxxP#Px

MOSP-Index 1.qxd

5/11/04

388 995a 995b 996a 996b 997a 997b 998a 998b 999a 999b 1000a 1000b 1001a 1001b 1002a 1002b 1003a 1003b 1004a 1004b 1005a 1005b 1006a 1006b 1007a 1007b 1008a 1008b 1009a 1009b 1010a 1010b 1011a 1011b 1012a 1012b 1013a 1013b 1014a 1014b 1015a 1015b 1016a 1016b 1017a 1017b 1018a 1018b 1019a 1019b 1020a 1020b 1021a 1021b 1022a 1022b

2:49 PM

Page 388

Index to the scansion of the Heliand A1 B1 A1 C A1 E D C A1 C A1 B1 A1 C B1 C B1 C A1 C D* B1 B1 A1 D* B1 B1 A1 A1 B1 D A1 D* B1 A1 B1 C B1 A1 B1 C C A1 B1 C B1 A1 C A1 B1 F A1 A1 A1 A1 C

xpx#xxxPx xxxxxP#xxP Px#Px xxxPXx Px#Px xx(xx)pXsXx#px P#pxSX xxxxxPSx px#xxxxpXx xxxxxP#px Pxx#Px xxxPx#P px#xpxx xxxxxxxPSx xxPxx#P xpXSx xxxPXS xxxP#px px#xPx xxxxxP#px PS#Pxx#px xPx#P xxxPxx#P xxPx#pxx Px#PSx xxxP#xpx xxxPx#P Px#xpxx px#xxPx xxPx#P px#Px#P pxx#Px px#xPXx xxxxxxPS#px px#xxxxpXx xxxPx#xP xxxPXx xxPx#P Px#xPx xxxxPxx#P xxP#px xpx#Px Px#xPx xxPx#px xxpx#Px xxxPX#P Px#xPx xxxpx#Px pxx#xpxx xxPx#P xxxxxIohannes#P pxx#xpxx pxx#xPx Px#Px Px#Px xxpXSx

1023a 1023b 1024a 1024b 1025a 1025b 1026a 1026b 1027a 1027b 1028a 1028b 1029a 1029b 1030a 1030b 1031a 1031b 1032a 1032b 1033a 1033b 1034a 1034b 1035a 1035b 1036a 1036b 1037a 1037b 1038a 1038b 1039a 1039b 1040a 1040b 1041a 1041b 1042a 1042b 1043a 1043b 1044a 1044b 1045a 1045b 1046a 1046b 1047a 1047b 1048a 1048b 1049a 1049b 1050a 1050b

C A1 A2 B1 C A1 C E C A1 A1 B1 A1 B1 A3 A1 F B1 A2 C A1 C C A1 A1 E F C A1 B1 C A1 A1 C A1 B1 D B1 or E C F B1 C A1 C B1 D F C E B1 A1 B1 E C C E

xxPXx pxx#xxPx Px#PS xxxPx#xP xxxPXx Px#xPx xxxPXx PXx#px xxxxxxPSx Px#Px Px#Px xxxPx#xP Px#xxPx xxxxPx#xP xxxxxxPx Pxx#Px Px#Satanasan xxxPx#px P#xPS xxxxPsx Px#Px xxxxpX#PX xxPSx PXSx xP#xPx xxxPSx#P Adaman#xxEuan xPSx xPx#xpxx xPx#P xxxxPSx Px#Px pxx#Px xxxxP#Px PX#Px xxxxxxpxx#xpx P#pXSx xxxxxPXx#px xpx#Px xxSatanase xxP#xxpx xxxpXSx Px#Px xxxPXx xxPx#px px#Pxx xxAdaman xPsx PXx#xP xxxxPxx#P xP#xxxPx xxxxxPx#P PXx#P xxxxxpx#Px xxxPsx PXx#P

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 389

Index to the scansion of the Heliand 1051a 1051b 1052a 1052b 1053a 1053b 1054a 1054b 1055a 1055b 1056a 1056b 1057a 1057b 1058a 1058b 1059a 1059b 1060a 1060b 1061a 1061b 1062a 1062b 1063a 1063b 1064a 1064b 1065a 1065b 1066a 1066b 1067a 1067b 1068a 1068b 1069a 1069b 1070a 1070b 1071a 1071b 1072a 1072b 1073a 1073b 1074a 1074b 1075a 1075b 1076a 1076b 1077a 1077b 1078a 1078b

A1 C A1 B1 C A1 A1 B1 B1 A1 A1 A1 D* C E A1 D* B1 C A1 B1 B1 D* C D* B1 D* C A3 C B1 A1 A1 B1 B1 A1 A1 B1 A1 C A3 A1 B1 B1 A1 B1 C A1 A3 B1 A1 A1 A1 A1 A1 B1

Px#xxPx xxpXSx Px#xPx xxxPx#P xPXx Px#Px Px#Px xxxpx#xxP xxxxxPx#P Px#Px PSx#PX Px#Px Px#xxpXSx xxxxP#PS xxxPSx#P Px#Px Px#pxx#P xxxxxxPx#P xxxxxxxxPXx Px#Px xxxpxx#px xPX#xxP Px#Psx xxxxP#PS Px#PXx xxxxxPx#P Px#xxxPSx xxxpx#px(xx) xxxxxxxPx xxxP#px xxxPx#P P#xxxPx xPx#xxPx xxxxPxx#P xxxPxS(xx) PSx#Px Px#Px xxxxxPx#px px#xxxxpXx xxxxxP#Px xxxxxPx xxxPxx#Px xxPx#px xxpxx#P Px#xpxx xxPx#px xxPXx Px#xPx xxxxPx xxPx#P PSx#PX Px#Px Pxx#xPx xpxS#pxx Px#Px xxxxP#xP

1079a 1079b 1080a 1080b 1081a 1081b 1082a 1082b 1083a 1083b 1084a 1084b 1085a 1085b 1086a 1086b 1087a 1087b 1088a 1088b 1089a 1089b 1090a 1090b 1091a 1091b 1092a 1092b 1093a 1093b 1094a 1094b 1095a 1095b 1096a 1096b 1097a 1097b 1098a 1098b 1099a 1099b 1100a 1100b 1101a 1101b 1102a 1102b 1103a 1103b 1104a 1104b 1105a 1105b 1106a 1106b

C A1 A1 C F C D* A1 A1 E D* C D* A1 B1 C C E B1 A1 A1 B1 A3 A1 B1 B1 B1 B1 B1 A1 A1 B1 B1 C B1A1 C D A1 E C A1 B1 A1 B1 A3 C D* B1 A1 B1 A1 B1 D* B1 B1 A1

389 xxxxxP#px Px#Px P#xxxPx xxPsx xxxxxHierusalem xxpx#Px Px#pXSx P#xPx xxxxPx#PX xxPSx#P xpx#xP#px xxxpx#px(xx) P#xxPx#px xpx#xxxPx xPx#xpx xxpx#px xPxx pxSx#px xxxxpx#xP Px#Px Px#xxxxxPx xxPx#xP xxxPx xpx#xPx xPx#P xxxxPxx#P xxxPx#P xxxxPx#xpx(xx) xxxPx#xP Px#Px Px#xxPx xxxxxxpxx#xP xxxxxxxPx#P xxPsx xPx#xxxxP#xPx xxxxpxSx xP#pXsx PXSx pXsx#px xxpXSx xxP#xxPx xxPx#xpx pxxx#pxx xxxxxxPX#xpx xxxxxxxPS xpx#Px Px#pxSx xxxPx#xP Px#xxxPx xxxxPx#px px#xxxPx xxxxxPx#P Px#xPsx xxxxxxPx#P xxxxxPx#P Pxx#Px

MOSP-Index 1.qxd

5/11/04

390 1107a 1107b 1108a 1108b 1109a 1109b 1110a 1110b 1111a 1111b 1112a 1112b 1113a 1113b 1114a 1114b 1115a 1115b 1116a 1116b 1117a 1117b 1118a 1118b 1119a 1119b 1120a 1120b 1121a 1121b 1122a 1122b 1123a 1123b 1124a 1124b 1125a 1125b 1126a 1126b 1127a 1127b 1128a 1128b 1129a 1129b 1130a 1130b 1131a 1131b 1132a 1132b 1133a 1133b 1134a 1134b

2:49 PM

Page 390

Index to the scansion of the Heliand D* B1 F B1 B1 C D*E A1 C A1 A1 B1 A1 C D* F D* C C A1 E B1 B1 A1 D* C A1 D C D* A1 B1 C A1 A1 B1 E B1 D* B1 F F B1 A1 D* A1 F B1 C B1 B1 A1 E A1 D* D*

Px#xPxx#P xxxxxxPx#xP Satanasan#xP xxP#xxP xxxPx#P xxxpx#Px P#xxpxSxx#px xxxPx#Px xxPSx Px#pxx pxx#xxxPx xxxPx#xP Px#xpxx xxxxxPsx Px#PXS Satanas#px PX#xxPsx xxP#px xxpxSx pxx#xPx px#Pxx#px xxxxPXS xxPXsx Px#Px Px#PSx xxxPsx Px#xxPx pXsXx xxxxPSx Px#P#px Px#Px xxxxPxx#P xxxP#px Px#Px pxx#xPx xxxxPx#P PSx#P xxxxxxPx#xP Px#pXSx xxPx#P P#xxxIordanes#px xxxIohannes#xP xpxS#px Px#Px Pxxx#pXsX xxxpxx#Px Iohannes#xPxx xxxxPx#xP xxxP#px xxPx#P xxxxPx#pX Px#Px PSx#P Px#Px pXx#PxX P#xP#xP

1135a 1135b 1136a 1136b 1137a 1137b 1138a 1138b 1139a 1139b 1140a 1140b 1141a 1141b 1142a 1142b 1143a 1143b 1144a 1144b 1145a 1145b 1146a 1146b 1147a 1147b 1148a 1148b 1149a 1149b 1150a 1150b 1151a 1151b 1152a 1152b 1153a 1153b 1154a 1154b 1155a 1155b 1156a 1156b 1157a 1157b 1158a 1158b 1159a 1159b 1160a 1160b 1161a 1161b 1162a 1162b

F D A1 B1 E A1 A1 A1 A3 A1 B1 A1 D* B1 B1 A1 A1 A1 D* B1 B1 D A1 B1 A1 A1 A1 B1 D* A1 E B1 F F A1 B1 F C A1 C C A1 A1 B1 B1 A1 C C E F A1 B1 A1 C A3 A1

xGalileo#P px#Px#P P#xxxPx xxxPx#P PSx#xpx xxPx#xPx P#xxxPx pXx#PX xxxxxxPxx Px#Px xxxxxxPS#px Px#Px Px#xxpXSx xxxxxPx#P xxPx#P P#xxPx xPx#xxPx pXSx xxxxxPx#xxPXx#P xxxxPx#P xxxPx#px px#Pxx Px#xxxPx xxxpxx#px xxPx#xPx xxxxPx#Px xPx#xxPx xxxxPx#P pxx#xPxx Pxx#Px PSx#px xxxxxxpxx#px xxxxIordan xxxGalileo#P xxP#xPx xxPx#P Andreas#xxPetrus xxpxSx Px#xxPx xxxP#px xxPSx Px#pxx Pxx#xxxPx xxxpxS#px xxPx#px Px#Px xxxxPxx xxxxxpx#Px px#Px#xpx xxxxxIordanes#Px Px#Px xxxxxpxx#P px#xxxPx xxxpXSx xxxPx Px#Px

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 391

Index to the scansion of the Heliand 1163a 1163b 1164a 1164b 1165a 1165b 1166a 1166b 1167a 1167b 1168a 1168b 1169a 1169b 1170a 1170b 1171a 1171b 1172a 1172b 1173a 1173b 1174a 1174b 1175a 1175b 1176a 1176b 1177a 1177b 1178a 1178b 1179a 1179b 1180a 1180b 1181a 1181b 1182a 1182b 1183a 1183b 1184a 1184b 1185a 1185b 1186a 1186b 1187a 1187b 1188a 1188b 1189a 1189b 1190a 1190b

D B1 A1 C A1 D* F C A1 B1 C A1 A1 A1 A1 B1 B1 A1 A1 B1 A1 B1 A1 C F B1 D* A1 A1 A1 D* B1 A1 B1 D* B1 F A1 E B1 A1 B1 B1 A1 A1 C D* B1 or E A1 B1 C B1 A1 B1 or E B1 or E B1

px#P#px xxPS#px Px#xxPx xxxxP#px Px#Px xPx#P#px Andreas#xxPetrus xxxxxxxpx#Px xPx#xxpxx xxPx#px xxxxpx#Px Px#Px px#xxxPx xxPXSx P#xPx xxPx#xpx xxPx#px Px#xPx xPx#xPx xxxxpxx#px Px#Px xxxxxxxPx#P Px#xxPx xxxpx#Px Iacobus#xxIohannes xxxPx#P Px#xxxPsx xxxxPx#Px Px#xxPx Px#Px xPx#PSx xxxxPx#P xpx#xxPx xxxPx#P Px#P#px xxxxxxP#xP Iacobus#xxIohannes Px#Px PSx#P xxxxPx#P xPx#xxxxpXx xxxxpxx#px xxPx#px Px#xPx Px#xxPx xxxxxxpx#Px Px#xxPXsx xxxxxxPXx#P PXx#xPx xxxPxx#P xxPXx xxxxxPx#xP px#xxxxpXx xxxxxPXx#px xxPXx#P xxxxxxPx#xP

1191a 1191b 1192a 1192b 1193a 1193b 1194a 1194b 1195a 1195b 1196a 1196b 1197a 1197b 1198a 1198b 1199a 1199b 1200a 1200b 1201a 1201b 1202a 1202b 1203a 1203b 1204a 1204b 1205a 1205b 1206a 1206b 1207a 1207b 1208a 1208b 1209a 1209b 1210a 1210b 1211a 1211b 1212a 1212b 1213a 1213b 1214a 1214b 1215a 1215b 1216a 1216b 1217a 1217b 1218a 1218b

D* D* E F C A1 A3 A1 A1 A1 D C A1 C A1 B1 D* A1 D* C A1 B1 E B1 B1 C A1 B1 B1 B1 A1 B1 A1 B1 B1 B1 A1 B1 B1 A1 A1 B1 C A1 A1 A1 D* A1 A1 C B1 A1 B1 B1 A1 B1

391 P#xxxxPsx pXx#Pxx PSxx#P Mattheus#xxPx xxPSx pxxx#Px xxxxxPx Px#xPx P#xxPx Px#xxxPx pXSXx xxP#px P#xxpxx xxpx#pxx Px#Px xxxxxxPxx#P P#xxpXx#P P#xPx Pxx#PXsx xxxPSx Px#xxxxpXx xxPxx#P Psxx#P xxxxxxPx#P xxxxPx#xP xxP#px Pxx#xPx xxPx#xP xxxPS#P xxPx#xpx Px#xPx xxPx#px xPx#xxxxpXx xxxxPx#P xxxPx#xpx xxPx#P pxX#Px xxxPx#P xxxPXS Px#Px Px#xxxPx xxxxxxxxPx#P xxxPSx xxxPx#xPx xxPx#xxPx Px#xxPx Px#xxxPSx Px#pxx xpxx#Px xxxxxPXx xpxx#P PX#xPx xxPS#px xxxxxPx#P xxxpx#xpxx xxxPx#P

MOSP-Index 1.qxd

5/11/04

392 1219a 1219b 1220a 1220b 1221a 1221b 1222a 1222b 1223a 1223b 1224a 1224b 1225a 1225b 1226a 1226b 1227a 1227b 1228a 1228b 1229a 1229b 1230a 1230b 1231a 1231b 1232a 1232b 1233a 1233b 1234a 1234b 1235a 1235b 1236a 1236b 1237a 1237b 1238a 1238b 1239a 1239b 1240a 1240b 1241a 1241b 1242a 1242b 1243a 1243b 1244a 1244b 1245a 1245b 1246a 1246b

2:49 PM

Page 392

Index to the scansion of the Heliand A1 B1 A1s A1 A3 A1 A1 B1 or E E* B1 A3 A1 A1 B1 C A1 A1 F D* B1 C A1 A1 B1 A1 B1 A1 A1 A1 B1 B1 C A1 A1 B1 A1 A1 B1 A1 B1 B1 C D* A1 C B1 E B1 A1 B1 B1 B1 A1 B1 B1 A1

Px#xxxxPx xxxxPx#P pXS#px pxxx#Px xxxxxxxPxx xPx#xPx px#xxxPx xxxxxxPXx#P Pxx#Px#px xxPx#P xxxxxxxpxx px#xxPx pxx#xxxPx xxxxxpx#xP xxxpxSx Px#Px Px#Px xxxxxxxIudeono#Px Px#Psx xxxxpxx#xP xxxxPxx Px#xxPx Px#Px xxxPx#px Px#Px xxPX#P xPx#xPx xxxPx#xPx xPx#xxxPx xxxxxxxxPx#P xxxPx#px xxpx#Px xpxx#xxxPx xxxxxxxPx#Px xxxPx#P Px#Px Px#xxPx xxxxxxPx#xP Px#xPx xxxPx#px xxxPx#xP xxxxPsx Px#xxPsx P#xPx xpx#Px xxPx#xP PSx#px xxPS#xP xPxx#Px xxxxxxPx#P xxxpx#xpx xxxxPx#P Px#xPx xxxxxxxPx#px xxxPx#px px#xPx

1247a 1247b 1248a 1248b 1249a 1249b 1250a 1250b 1251a 1251b 1252a 1252b 1253a 1253b 1254a 1254b 1255a 1255b 1256a 1256b 1257a 1257b 1258a 1258b 1259a 1259b 1260a 1260b 1261a 1261b 1262a 1262b 1263a 1263b 1264a 1264b 1265a 1265b 1266a 1266b 1267a 1267b 1268a 1268b 1269a 1269b 1270a 1270b 1271a 1271b 1272a 1272b 1273a 1273b 1274a 1274b

A1 A1 A1 B1 C A1 A1 B1 A1 E A1 B1 A1 A1 C A1 E* A1 F A1 A1 B1 F C A1 B1 A1 C D* B1 A1 F F F B1 B1 or E A1 F D* B1 B1 or E A1 E A1 A1 F F C A1A1 F E B1 A1 B1 D* C

Px#Px xP#xxPx pxx#xPx xxxxPx#P xxxP#Px Px#PX Px#xPx xxxPx#xP pxx#xPx PSx#P Pxx#pxx xxxxPxx#P xxxpx#xpxx Px#Px xxxPSx Px#Px Pxx#xP#xpx xxxxxxPx#Px Andreas#xxPetrus PX#Px xPx#Px xxPx#xP Iacobus#xxIohannes xxxpx#Px Px#xxxxxPx xxxxxPx#px Px#xxPx xxxP#px Px#xPxx xxpxx#px Pxx#Px Mattheus#xxThomas Iudasa#Px xxIacob#Px xPx#px xxxxxPXx#P Px#pxx P#xxIacob Px#pxXx xxxxxpxx#P xPXx#P pxx#xPx PSx#P xxxxxxpXx#Px Px#xxxPx Simon#xxPx xxBartholomeus xxxP#Px px#xxPx#Px xxPhilippus#xP PSx#P xxxxpxx#px Px#xxxPx xxPX#P pxxx#Psx xxxxPSx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 393

Index to the scansion of the Heliand 1275a 1275b 1276a 1276b 1277a 1277b 1278a 1278b 1279a 1279b 1280a 1280b 1281a 1281b 1282a 1282b 1283a 1283b 1284a 1284b 1285a 1285b 1286a 1286b 1287a 1287b 1288a 1288b 1289a 1289b 1290a 1290b 1291a 1291b 1292a 1292b 1293a 1293b 1294a 1294b 1295a 1295b 1296a 1296b 1297a 1297b 1298a 1298b 1299a 1299b 1300a 1300b 1301a 1301b 1302a 1302b

B1 A1 A1 B1 A1 B1 C A1 B1 or E A1 A1 B1 A1 B1 D* A1 A1 B1 A1 A1 D* A1 A1 A1 E D A3 A1s A1 C C A1 D* A1 B1 A1 A1 B1 A1 E D A1 A1 B1 D C A3 A1 D A1 A1 CA1 A1A1 A3A1 D* CA1

xPx#xP Px#Px Px#xxPx xxxPx#px xPSx#Px xxxPx#P xxxP#px Px#Px xxxxxPXx#P Px#Px pxx#xPx xxxPx#xP PX#xxxpxx xxPx#P px#xxxxpx#px Px#Px px#xPx xxxxPx#P Px#xxpxx xxxxPx#Px Px#PX#P Px#Px xxPx#xPx xxxxPx#Px pXS#xxpx px#Px#P xxxxPx PS#px Px#xPx xxP#px xxxpXSx Px#Px P#xxxxPxx xxP#xxPx xxP#xxpx Px#Px Px#xxPx xxxxP#xP Px#xPx PXx#px px#PS#P xxxPx#Px Px#Px xxxxxxPx#px P#pxSx xpx#Px xxxxxPxx PXSx px#PXx pxx#Px Px#xxxPx xxxPxx#Px P#xxxxPXSx xxxxxPx#Px Px#xPSx xxxPxx#Px

1303a 1303b 1304a 1304b 1305a 1305b 1306a 1306b 1307a 1307b 1308a 1308b 1309a 1309b 1310a 1310b 1311a 1311b 1312a 1312b 1313a 1313b 1314a 1314b 1315a 1315b 1316a 1316b 1317a 1317b 1318a 1318b 1319a 1319b 1320a 1320b 1321a 1321b 1322a 1322b 1323a 1323b 1324a 1324b 1325a 1325b 1326a 1326b 1327a 1327b 1328a 1328b 1329a 1329b 1330a 1330b

B1 C E CA1 E A3A1 A1A1 CA1 A1A1 A3A1 A1A1 A3A1 A1A1 D* A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 CA1 A1A1 A3A1 A1A1 A3A1 A1A1 CA1 A1A1 A3A1 A1A1 DD* A3A1 A3A1 A1A1 CA1 A1A1 A3A1 D* B1 E B1 D* B1 D B1 B1 A1 A1 B1 E B1 B1 A1 A1 B1

393 xxPXS xpXSx PS#xpx xxxPxx#Px PSx#P xxxxPx#Px xPx#xPx#Px xxxPxx#Px xxPx#xxPx#Px xxxxPx#xPx Px#xxxPx#Px Pxx#xxxxxpxx#xPx Px#xxPx#xPx xxxxxxxxPx#Pxx xPx#xxxPx#Px xxxxxxxpxx#xPx xPx#xxPx#xPx xxxxPx#xPx P#xxxPx#Px Pxx#xxxxPx#Px px#xpxx#Px xxxxPxx#Px Px#Px#Px Pxx#xxxxxxxpxx#Px xPx#xxPx#xPx xxxxxpxx#PX px#xPx#Px xxxPxx#Px xxpxsx#xxxxxPx#Px xxxxxxxxPx#xPx px#xxxPxx#Px xxxxxpx#Pxx#xPxx xxxxxxPx#Px xxxxPx#Px Px#xPx#Px xxxPxx#Px xPx#xPx#Px xxxxxxPxx#Px px#xxPsx xxxxpxx#P px#P#xpx xxPS#P Px#xPXsx xxxPx#xpx px#Psx xxxxPX#P xxPx#P Px#xPx Px#xPx xxxPx#xP pXSx#xpx xxxPx#px xxxxxPXsx Px#Px px#xxPx xxxxxpX#xpx

MOSP-Index 1.qxd

5/11/04

394 1331a 1331b 1332a 1332b 1333a 1333b 1334a 1334b 1335a 1335b 1336a 1336b 1337a 1337b 1338a 1338b 1339a 1339b 1340a 1340b 1341a 1341b 1342a 1342b 1343a 1343b 1344a 1344b 1345a 1345b 1346a 1346b 1347a 1347b 1348a 1348b 1349a 1349b 1350a 1350b 1351a 1351b 1352a 1352b 1353a 1353b 1354a 1354b 1355a 1355b 1356a 1356b 1357a 1357b 1358a 1358b

2:49 PM

Page 394

Index to the scansion of the Heliand E B1 B1 B1 A1 A1 D A1 D A1 C C A1 C A1 B1 A1 A1 D* B1 A1 B1 A1 C A1 C E A1 A2 B1 B1 A1 A1 A1 D* C D* B1 A1 B1 C A1 D* B1 A1 B1 A1 C A1 C D B1 E C B1 A1

PSxxsx xxxxxPx#P xP#xP xxxxxPx#P xPx#xxxxpXx xxxxxxxPx#Px P#pxSx pXx#PX px#Px#P Pxx#Px xxxxxPxx(xx) xxpx#Px Px#xxxxPx xxP#px Px#xxPx xxPx#px xPx#xxxxpXx xxPx#xPx Px#xpXSx xxPXsx Px#xxPx xxPx#px Px#xxxPx xxxxxxpx#Px P#xPx xxxP#Px xpx#Px#px Px#xpxx px#xxpXS xxxxPx#xpx xxxP#xpx Px#pxx Px#xxxxpXx P#xxPx pxx#Pxx#P xxxP#Px Px#pXsx xxxxxxPx#P xPx#xxPx xxxpxxx#P xxxxPSx pxx#pxx xPx#xPSx xxxxPx#P Px#xxxPx xxxxxPx#P PxSx xxxxxP#px Pxx#xPx xxxP#px px#PXS xxxxxPx#px PXx#xpx xxxxP#px xxxxxpXx#xpx Px#xPx

1359a 1359b 1360a 1360b 1361a 1361b 1362a 1362b 1363a 1363b 1364a 1364b 1365a 1365b 1366a 1366b 1367a 1367b 1368a 1368b 1369a 1369b 1370a 1370b 1371a 1371b 1372a 1372b 1373a 1373b 1374a 1374b 1375a 1375b 1376a 1376b 1377a 1377b 1378a 1378b 1379a 1379b 1380a 1380b 1381a 1381b 1382a 1382b 1383a 1383b 1384a 1384b 1385a 1385b 1386a 1386b

D* C A1 B1 D* A1 A1 B1 C A1 D* B1 A1 B1 or E A1 A1 A3 A1s A1 B1 A1 B1 A3 B1 A1 B1 B1 A1 A1 C A1 C B1 C A1 C A1 B1 A1 B1 B1 A1 A1 B1 D* C D* A1 B1 A1 D* A1 A1 B1 A1 A1

P#xxxPsx xxxxxxP#px Px#xPx xxxxxxPx#P Px#PSx xPx#Px Px#Px xxxxxpXx#xP xxP#px Pxxx#Px Px#xxpxSx xxxpxx#P P#xPx xxxxxPXx#xP Pxx#xPx xxxxxxPxx#Px xxxxxxxPx PS#px Px#xxxxPx xxxxxPx#px xPx#xPx xxPx#P xxxxxPx xxxPx#px Px#xPx xxxPx#xP xxpxx#P Px#Px px#xPx xxxxxxpx#P#x Px#Px xxxxxxxP#px xxxxxxPx#px xpXSx px#xxPx xxxxP#px xPx#xxPx xxxxPX#P Px#Px xxxxxPx#P xxPx#P PXSx Px#xPx xxPx#xpx xP#xxPSx xxxxP#px Px#xPx#P Px#Px xPx#px pxx#Px px#xxxxpx#px Px#Px px#xPx xxxxPx#P Px#xxpxx xxxxxPx#Px

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 395

Index to the scansion of the Heliand 1387a 1387b 1388a 1388b 1389a 1389b 1390a 1390b 1391a 1391b 1392a 1392b 1393a 1393b 1394a 1394b 1395a 1395b 1396a 1396b 1397a 1397b 1398a 1398b 1399a 1399b 1400a 1400b 1401a 1401b 1402a 1402b 1403a 1403b 1404a 1404b 1405a 1405b 1406a 1406b 1407a 1407b 1408a 1408b 1409a 1409b 1410a 1410b 1411a 1411b 1412a 1412b 1413a 1413b 1414a 1414b

C A1 D* B1 A3 A1 A1 B1 C A1 A1 C A2 C C C B1 B1 D C E B1 B1 A1 A1 A1 D* A1 A1 C A1 B1 B1 E A1 B1 B1 A1 A1 B1 C C A1 B1 A1 B1 C A1 E B1 D* B1 C A1 A3 C

xxP#px PxSx xP#xpXSx xxxxpxx#P xxxxPx xPx#Px Px#Px xxxxxpXx#xP xxP#px Px#Px px#xpxx xxP#px px#xxPS xxxxxP#px xpx#Px xxxxxxpx#Px xxxxP#xP xxPx#P P#Psx xpx#Px pxS#xP xxxxxP#xP xxxxPXS Px#Px xxPx#xPx P#xxxPx Px#xxP#px Px#Px Px#Px xxxxxxxPsx xxP#xxxxPx xxxPX#P xPx#px pxXx#px px#xxxxPx xxxxxxPx#xP xxxxP#xxpx Px#Px xPx#xPx xxxPx#P xpx#Px xxxpx#px Px#xPx xxPx#P pxx#xPx xxxxxxxxPx#P xxxxPSx Px#Px pXSx#xpx xxxPx#px Px#xxP#px xxxxxPx#xpx xxxxPsx Px#xPx xxxxPx xxxPsx

1415a 1415b 1416a 1416b 1417a 1417b 1418a 1418b 1419a 1419b 1420a 1420b 1421a 1421b 1422a 1422b 1423a 1423b 1424a 1424b 1425a 1425b 1426a 1426b 1427a 1427b 1428a 1428b 1429a 1429b 1430a 1430b 1431a 1431b 1432a 1432b 1433a 1433b 1434a 1434b 1435a 1435b 1436a 1436b 1437a 1437b 1438a 1438b 1439a 1439b 1440a 1440b 1441a 1441b 1442a 1442b

B1 A1 B1 A1 A1 B1 B1 D A3 A1 A1 C B1 A1 A1 C A1 B1 E B1 A1 C A3 A1 A1 B1 E B1 D*E B1 A1 A1 B1 B1 B1 C E B1 A3 A1 A1 B1 A1 B1 A1 B1 A1 C A1 B1 D* A1 D* B1 A1 B1

395 xxPx#P Px#xPx xxxxPx#P Px#Px xxxpxx#Px xxxxPx#P xxxpxx#xpx px#Pxx xxxxxPx Px#xPx xPx#xxxPx xxxxxxxpX#Px xxxxPx#P Px#Px Px#xxxxxPx xxxxpxsxx P#xxPx xxxxPx#P PSx#xpx xxPx#xpx px#xxPx xxxpx#Px xxxxPx P#xPx xPx#xxxxPx xxxxPx#P PSx#xpx xxxxxxpX#xP xxPx#xxpxsxx#P xxxPx#P Px#xxPx PxSx xxxPx#xpx xxP#xpx xxPx#P xxxxP#px PSx#P xxxxxxxxpXx#xP xxPxx Px#xPx Px#xPx xxxPx#P P#xPx xxxxxxPx#P Px#xPx xxxxxxPXsx P#xxPxx xxPsx Px#xxPx xxxxxPx#P Px#P#px Px#xPx P#xPsx xxxxxxxxxxPx#xP Px#xxxxxPx xxxxxPx#P

MOSP-Index 1.qxd

5/11/04

396 1443a 1443b 1444a 1444b 1445a 1445b 1446a 1446b 1447a 1447b 1448a 1448b 1449a 1449b 1450a 1450b 1451a 1451b 1452a 1452b 1453a 1453b 1454a 1454b 1455a 1455b 1456a 1456b 1457a 1457b 1458a 1458b 1459a 1459b 1460a 1460b 1461a 1461b 1462a 1462b 1463a 1463b 1464a 1464b 1465a 1465b 1466a 1466b 1467a 1467b 1468a 1468b 1469a 1469b 1470a 1470b

2:49 PM

Page 396

Index to the scansion of the Heliand A3 B1 D* B1 C A1 D B1 A1 C C E A1 B1 E C D* B1 A1 B1 A1 B1 A1 B1 A1 B1 C B1 B1 A1 A1 B1 A1 B1 D* C A3 E B1 B1 A1 B1 A1 B1 B1 C E B1 C E* B1 A1s A1 C A1 A1

xxxxxPx xxxxpxx#px P#xxxPSx xxPx#P xxxPsx Px#xPx P#PXx xxxxP#xpx Px#Px xxpx#Px xxxPXx PSx#P Px#xxPx xxxPx#P pxXx#P xxxpx#Px Px#xPx#xpx xxxxPX#px xxPx#xxPx xxxPx#px Px#xxPx xxxxxxPx#P PSx#xxxxPx xxxxPX#px Px#xxxxPx xxxxxxxPx#P xpx#px xxPx#px xxxPx#px Px#Px P#xxxxPx xxPS#P Px#xpxxx xxP#xP xPx#xxxPXx xxxxxpx#Px xxxxPx pXsXx#px xPx#P xxxxxpxx#P xPx#xxxxpXx xxxxxxPx#P Px#xpxx xxxxxxxPxx#px xxPx#P xpx#Px PXx#xpx xxxxPx#P xxPXx xxxxxPXSx#P xxPx#P PS#px xxxxPx#xPx xxxPSX xPx#xPx xxxxPx#Px

1471a 1471b 1472a 1472b 1473a 1473b 1474a 1474b 1475a 1475b 1476a 1476b 1477a 1477b 1478a 1478b 1479a 1479b 1480a 1480b 1481a 1481b 1482a 1482b 1483a 1483b 1484a 1484b 1485a 1485b 1486a 1486b 1487a 1487b 1488a 1488b 1489a 1489b 1490a 1490b 1491a 1491b 1492a 1492b 1493a 1493b 1494a 1494b 1495a 1495b 1496a 1496b 1497a 1497b 1498a 1498b

E A1 D A1 E F A3 A1 A1s B1 A3 A1 C A1 A1 B1 B1 A1 B1 C A1 B1 B1 A1 A1 A1 C B1 D* B1 B1 A1 A1 C C C B1 A1 A3 C A1 B1 A3 B1 A1 B1 E B1 B1 C D* A1 B1 A1 A1 B1

xxpx#Pxx#xpx xxxxxPx#Px pXsXx xxxxxxxPx#Px px#Px#xP xxxIudeon#P xxxxPx Px#Px PS#px xxxxPx#P xxxxxPx P#xPx xxxP#Px px#xxPx P#xPx xxxxxxPx#P xxxxPx#px Px#xPx xPx#P xxxxxxxP#px xxxPx#xxPx xxxxPx#xP xxxxxxxPx#P Px#xPx xPx#xxPx PxSx xxxxxxP#px xxxPxx#P xPx#xpx#px xPx#P xxPx#xP Px#pxx pxx#Px xxxxP#Px xxxxP#Px xxPsx xxxxPx#px P#xpxx xxxxPx xxPSx Px#xxPx xPxS xxxxPS xxxPx#xP xPx#xPx xxxxxpxx#px P#P#xpx xxxxxxxxxxxPx#xP xxxPsx#xpx xxxxxP#px Px#pxSx pxx#xxxxPx xxxxP#xpx P#xPx Px#xPx xxxxxxxxxPx#P

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 397

Index to the scansion of the Heliand 1499a 1499b 1500a 1500b 1501a 1501b 1502a 1502b 1503a 1503b 1504a 1504b 1505a 1505b 1506a 1506b 1507a 1507b 1508a 1508b 1509a 1509b 1510a 1510b 1511a 1511b 1512a 1512b 1513a 1513b 1514a 1514b 1515a 1515b 1516a 1516b 1517a 1517b 1518a 1518b 1519a 1519b 1520a 1520b 1521a 1521b 1522a 1522b 1523a 1523b 1524a 1524b 1525a 1525b 1526a 1526b

A3 A1 D B1 D A1 D B1 A1 C D* E A1 B1 A1 B1 A3 A1 A2 A1 A1 B1 A1 C D A1 A1A1 B1 A1 B1 A1 B1 D* B1 A1 B1 A3 A1 A3 A1 A1 B1 C A1 C A1 B1 A1 B1 B1 B1 A1 A1 A1 D* E

xxxxPx P#xPx P#pXSx xxPxxS P#pxSx Px#xPx px#Pxx xxxxP#xpx Px#Px xxpx#Px xPx#PSx PSx#xpx xxPx#xxPx xxxPx#xpx xPx#Px xPx#P xxxxxxPx xxxPx#Px Pxx#PS PxSx xxpxx#xxPx xxxxPx#P xxPx#xPx xxxxpxSx px#Psx xxxpxx#Px xPx#xxPx#Px xxxxxxP#xP xxP#xPx xxxxxPxx#P xPxx#Px xxxxPx#px Px#PSx xxxxPx#P xPx#xxxPx xxxxxPx#xP xxxxxxxPx Px#xPx xxxxPx Pxx#Px Px#xPx xxxxxxPx#P xxPSx Px#xPx xxxxpx#Px xPx#xPx xxP#xxP px#xxP#x xxP#xxP xxxxP#xP xxxxP#xxP P#xPx xxxxP#xpxx PxSx xP#xPSx Px#xpx

1527a 1527b 1528a 1528b 1529a 1529b 1530a 1530b 1531a 1531b 1532a 1532b 1533a 1533b 1534a 1534b 1535a 1535b 1536a 1536b 1537a 1537b 1538a 1538b 1539a 1539b 1540a 1540b 1541a 1541b 1542a 1542b 1543a 1543b 1544a 1544b 1545a 1545b 1546a 1546b 1547a 1547b 1548a 1548b 1549a 1549b 1550a 1550b 1551a 1551b 1552a 1552b 1553a 1553b 1554a 1554b

A1 B1 A3 A1 B1 A1 D* C B1 A1 A1 B1 B3 A1 C A1 A1 A1 D* A1 A1 B1 D* C A1 B1 D* C A1 B3A1 A1A1 A3A1 A1 C D* B1 B1 A1 D* B1 A1 B1 A1 B1 B1 B1 A1 C A1 B1 A3 B1 C A3A1 A3A1 CA1

397 Px#xPx xxxxxxPx#P xxxxxPx P#xPx xxxPx#xpx Px#Px Px#xxPsx xxxpx#pxx xxxxxxPx#P P#xPx xxPx#Px xxxxxPx#P xxxxpx Px#Px xxxxPSx P#xpxx Px#xxPx xxxxxxxxxpXx#xPx xxPxx#Px#xpx Px#Px px#xxxPx xxxxxxpxx#P Px#xpx#Px xxxxP#Px Px#xpxx xxPx#px Px#xPx#P xxxxPsx xxxxPxx#Px xxxxxxxxxxP#xPx xxP#xxxxPx#pXx xxxxxxxPx#Px xxpxx#xPx xxxP#Px Px#Psx xxxxxxxPx#xP xxxpxx#px Px#Px pxx#pxSx xxxpxx#px P#xPx xxxxxxxxxPx#xpx xxP#xxxxPx xxxPx#px xxxPx#xP xxPx#xP Px#xPx xxxxxxP#px xxPx#xxxxPx xxxxxxxPX#P xxxxxPx xxxPx#xxxpx xxPsx xxxxPx#Px xxxxxxxxpXx#xPx xxxxxxxPXx#Px

MOSP-Index 1.qxd

5/11/04

398 1555a 1555b 1556a 1556b 1557a 1557b 1558a 1558b 1559a 1559b 1560a 1560b 1561a 1561b 1562a 1562b 1563a 1563b 1564a 1564b 1565a 1565b 1566a 1566b 1567a 1567b 1568a 1568b 1569a 1569b 1570a 1570b 1571a 1571b 1572a 1572b 1573a 1573b 1574a 1574b 1575a 1575b 1576a 1576b 1577a 1577b 1578a 1578b 1579a 1579b 1580a 1580b 1581a 1581b 1582a 1582b

2:49 PM

Page 398

Index to the scansion of the Heliand E* A3A1 A1A1 E A2 C B1 B1 A1 B1 A1 B1 B1A1 B1 B1 A1 E C C A1 A3 C A3 A1 A3 A1 A1 B1 A1 B1 A1 B1 A1 B1 B1 A1 A3 A1 D* C C A1 A1 B1 A1 C A1 B1 A3 C B1 A1 D* A1 A1 B1

xP#xP#xP xxxxxPx#xPx xxpxXx#xxPx#Px xxxxPSx#px Px#xpx#P xxxxP#px xxPS#P xxxPx#xP pxxx#pxx xxxxxxxPx#px Px#xPx xxxxxPxx#P xPx#xxxxpx#xPx xxxpxx#xP xxxxxPxx#P P#xPx PSx#xpx xxxxP#px xxpx#Px Pxx#Px xxxxxPx xxxxxpx#Px xxxxxxxxPx Pxx#Px xxxxPx Px#Px xxpxx#xxxxPxx xxxPx#P Px#xPx xxxxxxxxpxx#xP xPx#xxpxx xxxP#xpx xPx#xxPx xxxxPxx#xP xxPxx#P P#xxPx xxxxxxxxxPx Pxx#Px Px#PSx xxxP#px xxpxSx Pxx#Px xP#xxxPx xxxxxPx#P Px#xpxx xxxxxpx#PX xPx#xPx xxxxxxxPx#P xxxxxPx xxxxxpx#Px xPx#px pxx#Px px#xxxxpx#px Px#Px px#xPx xxxxPx#P

1583a 1583b 1584a 1584b 1585a 1585b 1586a 1586b 1587a 1587b 1588a 1588b 1589a 1589b 1590a 1590b 1591a 1591b 1592a 1592b 1593a 1593b 1594a 1594b 1595a 1595b 1596a 1596b 1597a 1597b 1598a 1598b 1599a 1599b 1600a 1600b 1601a 1601b 1602a 1602b 1603a 1603b 1604a 1604b 1605a 1605b 1606a 1606b 1607a 1607b 1608a 1608b 1609a 1609b 1610a 1610b

A1 C A3 B1 B1 A1 A1 B1 A1 C A1 B1 A1 B1 B1 C A1 F D* A1 A1 B1 A1 B1 A1 B1 D D E C A1 A1 D* A1 A1 A1 A3 A1 B1 A1 ? A1 A1 C A3 B1 A3 A1 B1 A1 A1 B1 D* A1 A1 A1

Px#xxpxx xxP#px xxxxxPx xxxPxx#P xxxPx#P px#xPx Px#xPx xxxxxxxpXx#xpx Pxx#pxx xxpx#Px Px#xPx(xx) xxxxxPx#P xxPXx#xxPx xxxxxxPx#xP xxxPx#P xxxpx#Px Pxx#Px xIohannes#P PS#Pxx px#xpxx xpx#xPx xxPX#px Px#Px xxxPxx#xP xPx#xxxPx xxxxxPx#px P#Px#P px#Pxx P#P#xpx xxP#Px(xx) px#xxxPx PX#Px xxxpXx#PxXx xpx#xxxxPx px#Px pxx#Px xxxxPx pxx#Px xPx#xxpx Px#xpxx px#P Px#Px Px#xPx xxxxpX#Px xxxxPx xxPx#P xxPx pxx#Px xxpx#xxxxP Px#xPx xxPxx#Px xxxxxpxx#P pxxx#PSx xxxxxPx#Px xP#xxPx Px#Px

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 399

Index to the scansion of the Heliand 1611a 1611b 1612a 1612b 1613a 1613b 1614a 1614b 1615a 1615b 1616a 1616b 1617a 1617b 1618a 1618b 1619a 1619b 1620a 1620b 1621a 1621b 1622a 1622b 1623a 1623b 1624a 1624b 1625a 1625b 1626a 1626b 1627a 1627b 1628a 1628b 1629a 1629b 1630a 1630b 1631a 1631b 1632a 1632b 1633a 1633b 1634a 1634b 1635a 1635b 1636a 1636b 1637a 1637b 1638a 1638b

A3 B1 A3 A1 A3 C A1 B1 A1 C A3 A1 A1 B1 A1 B1 D A1s D* B1 A3 A1 D B1 E C B1 A1 D* B1 A1 B1 A1 C A1 A1 B1 A1 C A1 D* B1 A1 B1 A1 C E* B1 B1 C C A1 A1 C A2 B1

xxxxxPx xxPxx#P xxxxxPx pxx#Px xxxxPx xxxpx#Px px#xxxPx xxPX#P Px#xPx xPSx xxxxxxPx Px#xpxx xxpxx#xxxxPxx xxxxPx#P Px#xPx xxxxxPX#P px#pxSx pXS#px pxxx#PSx xxxxxxxP#xP xxxxxPx Px#xPx px#PSx xxxxxPX#P PS#xpx xxxxxP#px xxPS#P xPxx#Px Px#xPSx xxPx#P xPx#xxxxPx xxxxPx#P xpx#xxPx xxxxxP#px px#xxxxpXx xxxxxPx#Px xxPx#px Px#Px xxxxPXx Px#Px Px#xxPSx xxxxxxpxx#P xPx#xxxxPx xxPx#P PX#xxPx xxpx#Px Px#P#xpx xxxxxP#xxP xxPx#px xpXSx xxxxxxxPSx Px#Px xPSx#xxxxxPx xxxpx#Px Px#xPS xxxP#xpx

1639a 1639b 1640a 1640b 1641a 1641b 1642a 1642b 1643a 1643b 1644a 1644b 1645a 1645b 1646a 1646b 1647a 1647b 1648a 1648b 1649a 1649b 1650a 1650b 1651a 1651b 1652a 1652b 1653a 1653b 1654a 1654b 1655a 1655b 1656a 1656b 1657a 1657b 1658a 1658b 1659a 1659b 1660a 1660b 1661a 1661b 1662a 1662b 1663a 1663b 1664a 1664b 1665a 1665b 1666a 1666b

A1 B1 A3 A1 E A1 D* A1 B1 A1 A1 E A1 B1 D* B1 A3 A1 D* B1 A3 C A1 B1 C A1 D* C B1 A1 A1 C A1 B1 B1 A1 A3 A1 E B1 A1 B1 B1 A1 D* C A1 C A1 B1 A1 C A1 B1 A3 B1

399 P#xxPx xxPx#P xxxxPx Px#Px pXSx#xpx xxxxxxPx#Px xxPx#xxP#px pXx#xPx xxxxPXS PXSx xxpx#xxPx xxpXSx#xpx Px#xPx xxxxPx#xpx xPx#xPsx xxxxPx#P xxxxpxx P#xPx pxx#pxSx xxxxxxPX#xpx xxxPx xxpx#Px px#xxPx xxxxPx#px xxpXSx Px#xPx pxx#xxxPsx xxxxxxxxpx#Px xxxxxxPx#xP PxSx px#xxPx xxP#px P#xPx xxxPx#P xxxxxxxPx#pX Px#xPx xxxxxxxPx Px#Px xxxpXSx#px xxPX#px xPx#xPx xxxxxxPx#xP xxPXsx P#xPx xxPx#xPsx xxP#Px xxxPx#xxxxxxpxx xxxxpx#Px xPX#xxxxPx xxxxPx#px px#xxPx xxP#Px px#xxPx xxxPX#P xxxPx xxxPx#P

MOSP-Index 1.qxd

5/11/04

400 1667a 1667b 1668a 1668b 1669a 1669b 1670a 1670b 1671a 1671b 1672a 1672b 1673a 1673b 1674a 1674b 1675a 1675b 1676a 1676b 1677a 1677b 1678a 1678b 1679a 1679b 1680a 1680b 1681a 1681b 1682a 1682b 1683a 1683b 1684a 1684b 1685a 1685b 1686a 1686b 1687a 1687b 1688a 1688b 1689a 1689b 1690a 1690b 1691a 1691b 1692a 1692b 1693a 1693b 1694a 1694b

2:49 PM

Page 400

Index to the scansion of the Heliand A1A1 B1 E B1 D* C B1 A1 A1 C A1 B1 A1 B1 A1 B1 F C E* C A1 B1 A3 B1 A1 C B1 A1 D*A1 A3A1 A1A1 A3A1 D*A1 A3A1 A1A1 A3A1 A1A1 A1A1 A1A1 A3A1 A1⫹C A3A1 A1A1 A3A1 A1A1 A3A1 B1 A1 A1 E A1 D B1 B1 A1 B1

Px#xxPx#Px xxxxxxPx#px PSx#xxpx xxxxxxpXx#P px#xpXsx xxxxxxpx#Px xxxxPx#P px#xpxx Px#xxPx xxxxxxxpx#Px px#xxxxxPx xxPx#P pxx#xpxx xxxPx#P PSx#xPx xxxxPx#P Salomon#xpX xxxP#px PXSx#P xxxxxP#Px pxx#xPx xxxxxxPx#P xxxxxxxPx xxxxxxxxPx#xP xPx#xxxPx xxP#px xxxPx#P pxx#xpxx Px#xxPSx#Px xxxxxPx#PX P#xpxx#Px xxxxxxxpxx#Px Px#xxPxx#pxx xxxxxxxPx#xPx PX#xPx#Px xxxxxxxxxxxPx#Px xPx#xxxxxxpxx#xPx P#xxxPx#Px Px#xpxx#Px xxxxxxxPx#Px px#xxxPX#xpx#Px xxxxxxxxPx#Px Px#xPxx#Px xxxxxPx#Px px#xxxxPx#xpxx xxxxxPx#Px xxxxPx#P Px#Px xxxxPxxx#Px PSx#P Px#xPx xxP#P#px xxxxPx#P xxxxPx#P Px#xxPx xxxxPx#xpx

1695a 1695b 1696a 1696b 1697a 1697b 1698a 1698b 1699a 1699b 1700a 1700b 1701a 1701b 1702a 1702b 1703a 1703b 1704a 1704b 1705a 1705b 1706a 1706b 1707a 1707b 1708a 1708b 1709a 1709b 1710a 1710b 1711a 1711b 1712a 1712b 1713a 1713b 1714a 1714b 1715a 1715b 1716a 1716b 1717a 1717b 1718a 1718b 1719a 1719b 1720a 1720b 1721a 1721b 1722a 1722b

A1 B1 A1 A1 B1 A1 A1s B1 A1 B1 A3 B1 A1 B1 B1 A1 A1s B1 A1 B1 A1 B1 A1 B1 E C A3 B1 A1 B1 E* A1 A1 C B1 A1 A1 B1 C A1 A1 B1 E C B1 A1 A1 A1 A1 B1 A1 B1 E A1 D* C

PS#Px xxxxPx#xP pxx#xxxPx Px#xxPx xxPS#xP Pxx#Px PS#px xxxxPx#P Px#xpxxx xxxxxPxx#xP xxxxxxxPx xxPx#xpx xpx#xPx xxxxPx#P xxxPx#px Px#Px PS#px xxxxxxxP#xpx Px#Px xxxxxPx#xpx P#xxPx xxxPx#xpx xxPx#Px xxxxxxPx#px P#P#xxpx xxxxxxpx#Px xxxxxxxPx xxxxP#xpx Px#xxxxpxx xxxPx#P Px#Px#xP Px#xPx xPx#xxPx xxxxxpx#Px xxxxPXS Px#xpxxx px#xxxxpXx xxxPx#xP xxPxx Px#Px px#xxPx xxxxxPx#P pXSx#xpx xxxxxpx#Px xxxxxPx#P Pxx#xPx PSx#Px xxxxxxPx#Px pxx#xPx xxxxxPx#P Pxx#pxx xxxxPx#xpx xxpxSx#px xxPx#xPx Px#Psx xxxxpx#Px

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 401

Index to the scansion of the Heliand 1723a 1723b 1724a 1724b 1725a 1725b 1726a 1726b 1727a 1727b 1728a 1728b 1729a 1729b 1730a 1730b 1731a 1731b 1732a 1732b 1733a 1733b 1734a 1734b 1735a 1735b 1736a 1736b 1737a 1737b 1738a 1738b 1739a 1739b 1740a 1740b 1741a 1741b 1742a 1742b 1743a 1743b 1744a 1744b 1745a 1745b 1746a 1746b 1747a 1747b 1748a 1748b 1749a 1749b 1750a 1750b

A1 B1 A1 C B1 A1 D* B1 B1 A1 B1 C A1 B1 D* C A1 B1 B1 C A1 B1 A1 B1 A1 B1 B1 A1 A1 B1 A1 B1 A3 B1 B1 B1 A1 B1 E C A1 B1 A1 B1 B1 B1 E* B1 B1 A1 D* A1 A1 B1 A1 C

Px#xPx xxxPx#xP pxxx#pxx xxxxP#px xxxPx#P Px#xPx xPx#px#Px xxxPx#xP xxxPx#P Pxx#pxx xxPx#P xxPsx P#xxPx xxxPxx#P xxxPx#xxPx#P xxxxxxxxxpx#Px xPx#xPx xxxxxxxxPx#P xxxPx#px xxP#pxx xxPx#xxPx xxxxxxPx#P Pxx#Px xxxxxPx#px Px#xxxxPx xxxxxxxPx#px xxxPx#xpx Px#xPx xxPx#xxPx xxxxxxxxPx#xP pxx#pxx xxxxPx#px xxxxxxPx xxxpx#xpx xxxPS#P xxxP#xpx xxPx#xPx xxxxxxxPx#xpx Psx#px xxpx#PX pxxx#Px xxPx#xpx pxx#xPx xxxxxxPx#P xxxpxx#P xxxPx#P Px#Px#xpx xxxP#xxP xxPx#P pxx#Px Px#Px#P xxxxxxxPx#xpxxx xxPx#xxxxpXx xxxxPx#xpx xxP#xxPx xxxxPsx

1751a 1751b 1752a 1752b 1753a 1753b 1754a 1754b 1755a 1755b 1756a 1756b 1757a 1757b 1758a 1758b 1759a 1759b 1760a 1760b 1761a 1761b 1762a 1762b 1763a 1763b 1764a 1764b 1765a 1765b 1766a 1766b 1767a 1767b 1768a 1768b 1769a 1769b 1770a 1770b 1771a 1771b 1772a 1772b 1773a 1773b 1774a 1774b 1775a 1775b 1776a 1776b 1777a 1777b 1778a 1778b

D* A1 D* A1 A1 C A1 C B1 A1 D* B1 A1 C A1 B1 A1 D A1 B1 A1 B1 A1 A1 B1 B1 A1 A1 A1 B1 E* D A1 B1 A1 A1 A1 B1 B1 A1 A3 C A1 B1 D B1 E B1 E C E B1 A1 C A1 B1

401 pxxx#Psx Px#Px xPxx#Px#xpx Px#Px Px#xxPx xxxxP#px px#xxxPx xxxxxpx#PX xxxxxpxx#P PXSx Pxx#pxSx xxxxxxPx#px xPx#xxxPx xxxpx#Px xPx#xxPx xxxxxP#xxP xxxxxxPx#Px P#PSx PS#xxxPx xxxxxxPx#xpx P#xxPx xxxxxxPx#px P#xxxPx xxxxxPxx#Px xxPS#P xxxxxP#xxP Px#xPx Px#pxx Px#xPx xxxPX#P Px#Px#xpx P#pxSx pxX#Px xxxxxPx#xpx xxPx#xxPx P#xPx Px#xxxPx xxxxpxx#P xxPx#P Px#xPx xxxxPx xxpx#Px Px#xxxxPx xxxPx#P P#PXS xxxPx#P P#Px#xxP xxxpxx#px PSx#px xxxxxxP#px pXSx#px xxxPx#P Px#Px xxxxpx#Px pxx#xPx xxP#xxP

MOSP-Index 1.qxd

5/11/04

402 1779a 1779b 1780a 1780b 1781a 1781b 1782a 1782b 1783a 1783b 1784a 1784b 1785a 1785b 1786a 1786b 1787a 1787b 1788a 1788b 1789a 1789b 1790a 1790b 1791a 1791b 1792a 1792b 1793a 1793b 1794a 1794b 1795a 1795b 1796a 1796b 1797a 1797b 1798a 1798b 1799a 1799b 1800a 1800b 1801a 1801b 1802a 1802b 1803a 1803b 1804a 1804b 1805a 1805b 1806a 1806b

2:49 PM

Page 402

Index to the scansion of the Heliand A1 A1 A1 B1 A3 A1 A1 B1 D* B1 A1 C B1 A1 A3 B1 A1 C A3 C B1 C E B1 A1 C A1 A1 C C A3 B1 D* C D*A1 B1 C C B1 B1 E B1 C A1 A1s B1 A1 A1 B1 B1 A1 A1 A1 B1 A1 C

pXS#xPx Px#xxxxpXx PxSx xxxxxxPx#xpx xxxxPx Pxx#pxx P#xxxxpXx xxxxpxx#P Px#Psx xxxxxpxx#P Px#Px xxxpx#Px xxPxx#P Px#Px xxxxxxxPx xxxPx#xP pxx#xpXx xxxxpx#Px xxxxxxxPx xxxP#px xxPS#P xxP#Px PSx#P xxxPx#px PX#Px xxxxP#Px xpx#xPx xxP#xxPx xxpx#Px xxpx#Px xxxxpXx xxxxPx#px Px#pxx#P xxpx#Px P#xxxPxx#Px xxxxxxPx#P xxxxpx#Px xxxxP#px xxxPxx#px xxxxxP#xxP pXSx#xpx xxxxPxx#P xxpx#Px Px#Px PS#px xxxxPx#P xxxxpxx#Px PS#pxx xxxxPx#xpx xxxxxPx#px xPx#xxPx xxxxxxxpx#xPx Px#xxxxxPx xxPx#P Pxx#Px xxP#px

1807a 1807b 1808a 1808b 1809a 1809b 1810a 1810b 1811a 1811b 1812a 1812b 1813a 1813b 1814a 1814b 1815a 1815b 1816a 1816b 1817a 1817b 1818a 1818b 1819a 1819b 1820a 1820b 1821a 1821b 1822a 1822b 1823a 1823b 1824a 1824b 1825a 1825b 1826a 1826b 1827a 1827b 1828a 1828b 1829a 1829b 1830a 1830b 1831a 1831b 1832a 1832b 1833a 1833b 1834a 1834b

D* A1 A1 A1 A1 B1 D* A1 A1 A1 A1 B1 A1 C A1 B1 B1 B1 A1 B1 C A1 B1 B1 B1 A1 B1 B1 A1 B1 A1 B1 A1 B1 A1 B1 A1 C D* C A1s B1 A1 B1 B1 A1 A1 B1 A3 A1 A1 C A1A1 C C A1

Px#pxSx xxPsx#Px xPxx#Px xxxpxx#Px Px#Px xxP#xP xP#xpxx#P Px#xPx xxxxxPxsXx Px#Px xxxpxx#Px xxxPx#P xPx#xxxPx xxxxxpx#pxx px#xxPx xxPx#xP xxxPx#xpx xxxPx#xpx Px#Px xxxPx#P xxxPSx Px#xPx xxPxx#px xxxpxx#px xPx#px pxS#Px xxPxx#P xxPx#P Px#xxxPx xxxP#xxP xPx#xxxPx xxxxPx#P xPx#xxxPx xxxPxx#P Px#xPx xxxxxPx#xP P#xPx xxxxxxxP#px Px#Px#xP xxxxxxxpx#Px pXS#px xxxPxx#px PSx#Px xxxxxxPx#xpx xxxxpxx#P Px#xPx Px#xxPx xxxPx#P xxxPx Px#Px Px#Px xxxP#Px Px#xPxSx xxxPsx xxxPSx Px#Px

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 403

Index to the scansion of the Heliand 1835a 1835b 1836a 1836b 1837a 1837b 1838a 1838b 1839a 1839b 1840a 1840b 1841a 1841b 1842a 1842b 1843a 1843b 1844a 1844b 1845a 1845b 1846a 1846b 1847a 1847b 1848a 1848b 1849a 1849b 1850a 1850b 1851a 1851b 1852a 1852b 1853a 1853b 1854a 1854b 1855a 1855b 1856a 1856b 1857a 1857b 1858a 1858b 1859a 1859b 1860a 1860b 1861a 1861b 1862a 1862b

A1 B1 A1 B1 A1 B1 C A1 C C E B1 A3 A1 D* A1s A1 B1 A3 A1 A1 C A2 B1 B1 C A1 B1 C A1 D* E D A1s A1 B1 A1 C D* C A1 B1 A1 C A1 B1 C A1 D* B1 D* C A1 B1 A1 C

xpx#xxxxPx xxxxPx#P xpx#xPx xxxxxpxx#P xP#xxxxPx xxxPx#xP xxPXx Px#Px xxpXSx xpx#Px PSx#px xxxxP#xP xxxxPx Px#xxPx Px#PSx pXS#px Px#Px xxxPx#xP xxxxxxxPx Px#xPx Px#Px xxxx(xx)xxxxP#px xxP#xxxPS xxxxxP#xpx xxxpxx#px xxxxxxxxxxpx#Px Px#xxxxPx xxxxxxPx#P xxxpxSx Pxx#Px Px#xPx#P PSx#P px#PSx pXS#px Px#Px xxxxpxx#xP Px#xPx xxxxxxxP#px pxx#pxSx xxxxxxxxxpx#Px Px#xxxxxPx xxxxxPx#xP Px#Px xxxxxP#Px px#xpXx xxPx#px xxxP#Px xxxxxxpx#xPx P#xxxxPsx xxxPX#px Px#xPsx xxxpx#Px PSx#Px xxxPx#px Px#xxPx xxxxP#Px

1863a 1863b 1864a 1864b 1865a 1865b 1866a 1866b 1867a 1867b 1868a 1868b 1869a 1869b 1870a 1870b 1871a 1871b 1872a 1872b 1873a 1873b 1874a 1874b 1875a 1875b 1876a 1876b 1877a 1877b 1878a 1878b 1879a 1879b 1880a 1880b 1881a 1881b 1882a 1882b 1883a 1883b 1884a 1884b 1885a 1885b 1886a 1886b 1887a 1887b 1888a 1888b 1889a 1889b 1890a 1890b

A1 B1 A1 C A2 B1 B1 A1 A3 C C A1 A1 C D* C C C A1 B1 A1 B1 C A1 B1 C E B1 A1 B1 A1 C A1 B1 A1 B1 B1 A1 A1 B1 A1 B1 or E A3 A1 B1 E D* B1 A3 A1 A1 C C B1 or E B1 B1

403 xxP#xPx xxpxxx#P Px#xPx xxxxxP#px Px#xpx#P xxPxx#P xxxPx#P Px#pxx xxxxPx xpXSx xxxxPsx PxSx Px#xPx xxP#px Px#xxPSx xxxpXSx xxP#Px xxxPsx PXx#xPx xxpXS#px Px#Px xxxPx#P xxxxxPSx xP#xxPx xxxxxxxxPX#px xxpx#Px xxPSx#P xxxxP#xxP xPx#xPx xxxxxPx#P Px#xPx xxxxPSx pxx#Px xxxxxxxpxx#xpx xPx#xxxPx xxxPx#px xxxP#xpx PxSx Px#xPx xxxxpx#xxP xxxPx#Px xxxxPXx#P xxxxxxxxPx Px#xPx xxxPx#xpx PSx#px Px#Psx xxP#xP xxxPx xpx#xPx xpx#xxxPx xxxxxxxP#px xxPXx xxxxPXx#xpx xpx#xxxxP xxxP#xpx

MOSP-Index 1.qxd

5/11/04

404 1891a 1891b 1892a 1892b 1893a 1893b 1894a 1894b 1895a 1895b 1896a 1896b 1897a 1897b 1898a 1898b 1899a 1899b 1900a 1900b 1901a 1901b 1902a 1902b 1903a 1903b 1904a 1904b 1905a 1905b 1906a 1906b 1907a 1907b 1908a 1908b 1909a 1909b 1910a 1910b 1911a 1911b 1912a 1912b 1913a 1913b 1914a 1914b 1915a 1915b 1916a 1916b 1917a 1917b 1918a 1918b

2:49 PM

Page 404

Index to the scansion of the Heliand A2 B1 A1 B1 C A1 A3 B1 A1 A1 D* C D B1 C B1 C A1 B1 A1 E C A1 B1 A1 B1 D* B1 C A1 A1 B1 A1 B1 D* A1 C B1 A1 C B1 B1 A1 B1 C C D* C A3 B1 A2 D* A1 A1 A1 A1

px#xxPS xxxPxx#px xPx#Px xxxxxxPx#px xxpXsXx Px#xPx xxxxxxPx xxxPx#P xPxx#Px xxPx#xpxx xP#xPsx xxxxxxxxxpx#Px px#PXx xxxxxxxxPx#px xxxxpx#Px xxxxxPx#P xxxPsx Px#Px xxxxxPx#px Pxx#Px PSxx#xpx xxxP#px Px#xpxx xxxxxPxx#P Px#xxxxPx xxxxxxxxxxPx#P xPx#xxPXsx xxxxxxpxx#xP xxxxxxPsx Px#xPx xpx#xPx xxxxPx#xpx P#xPx xxxxPX#P Px#px#Px Px#Px xxPsx xxxxPx#xP Px#Px xxxxxPsx xxxPx#xP xxxxxxPx#px xPx#xxxPx xxxxxxxxxPx#px xpx#Px xxpx#Px Px#Px#P xpXSx xxxxxxxpxx xxxPx#xP Px#xPS pxx#P#px xxxxxxpx#xpxx xPxx#Px Px#xxPx xxPx#xPx

1919a 1919b 1920a 1920b 1921a 1921b 1922a 1922b 1923a 1923b 1924a 1924b 1925a 1925b 1926a 1926b 1927a 1927b 1928a 1928b 1929a 1929b 1930a 1930b 1931a 1931b 1932a 1932b 1933a 1933b 1934a 1934b 1935a 1935b 1936a 1936b 1937a 1937b 1938a 1938b 1939a 1939b 1940a 1940b 1941a 1941b 1942a 1942b 1943a 1943b 1944a 1944b 1945a 1945b 1946a 1946b

D* C A3 B1 D* B1 A3 D A1 B1 A1 B1 B1 A1 A1 B1 A3 C C A1 B1 C A1 C E B1 A1 C A1 B1 B1 A1 B1 B1 A1 C A1 C A1 B1 A1 D C B1 A1 C A1 B1 A1 C B1 B1 A1 C D B1

Px#PSx xxxxxP#px xxxxPx xxpxx#P Px#xxpx#Px xxPx#P xxxxPx px#pxSx P#xxPx xxxxxPx#xpx Px#xPx xxxPxx#P xxxxPx#xP PxSx P#xxPx xxxxxxPx#P xxxxxPx xxxxxP#px xxxPXx xPx#xPx xxPx#px xxxxxPSx Px#xxxxxpXx xxxpx#Px P#Px#xP xxxxxxxxPx#P P#xxxxpxx xxxxxxxxPsx Px#Px xxxxPxx#P xxxxPx#P Px#xPx xPx#px xxxxxPx#xP px#xPx xxxP#Px Px#xxPx xxxxpx#Px Px#xPx xxxxxPx#px Pxx#Px pXsXx xxxxPxx xxxPxx#P Px#xPx xxxxP#Px Px#xxPx xxxxPx#P px#xxxPx xxxpx#Px xxxxxPxx#P xxxxxxPx#P xpxSx#Px xxxxxP#Px px#Psx xxxxxxPx#P

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 405

Index to the scansion of the Heliand 1947a 1947b 1948a 1948b 1949a 1949b 1950a 1950b 1951a 1951b 1952a 1952b 1953a 1953b 1954a 1954b 1955a 1955b 1956a 1956b 1957a 1957b 1958a 1958b 1959a 1959b 1960a 1960b 1961a 1961b 1962a 1962b 1963a 1963b 1964a 1964b 1965a 1965b 1966a 1966b 1967a 1967b 1968a 1968b 1969a 1969b 1970a 1970b 1971a 1971b 1972a 1972b 1973a 1973b 1974a 1974b

A1 B1 A1 C D* B1 A3 C B1 C F B1 C A1 A1 C A2 C A3 C A3 B1 B1 B1 A1 B1 D* A1 D* B1 D A1 A1 B1 B1 A1 E A1 A1 A1 A1 A1 A3 C C A1 D* B1 A3 A1 A1 B1 D*E CA1 A1 A1

Px#xxxPx xxxxxxPx#xpx xPx#xxxxPx xxxxxpx#Px xxpxx#xxPSx xxxPx#xP xxxxxxPx xxxxpX#Px xxxPx#P xxP#px xxSodomo-S xxxPx#P xPSx Px#Px Px#xPx xxxxxpx#Px Pxx#PS xxP#Px xxxxxPx xxxxxxxP#Px xxxxxxPx xPx#px xPx#P xxxPx#P Px#xPx xxxPX#P xPx#px#Px pxx#Px Px#Psx xxxxP#xP P#PX#P xxPx#Px pxx#xpxxx xxxxxPx#xP xxxPx#px Px#pxxx PXx#xpx pxx#Px xxPxxx#Px P#xPx Px#Px xxpx#xxPx xxxxPx xxxP#px xxpx#Px P#xPx Px#pXSx xxxxxxxxPx#xP xxxxxxPx PxSx pxx#xxxxPx xxxxxpxx#xP xPx#xxpxSx#px xxxxxxxPxx#Px xxxpxx#pxx xxxxxxPx#Px

1975a 1975b 1976a 1976b 1977a 1977b 1978a 1978b 1979a 1979b 1980a 1980b 1981a 1981b 1982a 1982b 1983a 1983b 1984a 1984b 1985a 1985b 1986a 1986b 1987a 1987b 1988a 1988b 1989a 1989b 1990a 1990b 1991a 1991b 1992a 1992b 1993a 1993b 1994a 1994b 1995a 1995b 1996a 1996b 1997a 1997b 1998a 1998b 1999a 1999b 2000a 2000b 2001a 2001b 2002a 2002b

B1 A1 D* B1 C B1 B1 C C A1 A1 C D* B1 A1 C A1 C A3 E A1 C A1 C A1 A1 A1 B1 D* A1 B1 A1 E B1 D* B1 B1 A1 E A1 F B1 D* C E F C A1 A1 A1 D* D B1 B1 F B1

405 xxxxpXx#xpx Px#Px xpx#xPsx xxxPxx#P xxxxxP#Px xPx#px xxxxxpxx#px xP#px xxxpxSx Px#Px pxx#xxxPx xxxxxxxP#px Px#Psx xxxxPx#P Px#Px xxxP#px xxxxxxxPx#Px xpx#Px xxxxPx PXx#px xPx#xPx xxP#px Px#Px xxxxpx#px xxxxPx#xpxx pxsx#Px Px#xPx xxPx#P xPx#pXsXx Pxx#Px xxxpXx#P Px#xxPx PSx#px xxxxPXS Pxx#xPxx xxxxP#xP xxxxxPx#xP Px#PX xxxxxxxP#P#xxP xxxPx#Px xGalileo#P xxxxxPx#P xpx#xP#px xxxxxxP#px pxSx#px xMaria#P xxxpx#Px Px#Px Pxx#Px pxxx#Px P#xxxxxPxx px#Px#P xxPx#P xxpx#P xIudeon#xxxPsx xxxxxPx#P

MOSP-Index 1.qxd

5/11/04

406 2003a 2003b 2004a 2004b 2005a 2005b 2006a 2006b 2007a 2007b 2008a 2008b 2009a 2009b 2010a 2010b 2011a 2011b 2012a 2012b 2013a 2013b 2014a 2014b 2015a 2015b 2016a 2016b 2017a 2017b 2018a 2018b 2019a 2019b 2020a 2020b 2021a 2021b 2022a 2022b 2023a 2023b 2024a 2024b 2025a 2025b 2026a 2026b 2027a 2027b 2028a 2028b 2029a 2029b 2030a 2030b

2:49 PM

Page 406

Index to the scansion of the Heliand A3 C D* E A2 D A3 A1 D D* A1 B1 D* B1 A1 B1 B1 A1 A1 B1 A1 B1 A1 B1 A1 C A1 B1 A3 D* A1 C C A1 B1 A1 A1 B1 B1 A1 A1 C D* B1 B1 B1 B1 B1 A1 B1 A1 B1 C A1 A3 E

xxxxPx xxxxP#px Px#xpXsx PXx#P PXx#PS px#Pxx xxxxPx Px#xPx px#PSx Px#PXS Px#xPx xxPxx#P xPx#xxxpxsx xxPx#P px#xPx xxP#xxP xxPx#xP Px#xPx Px#xxPx xxxPx#P xPx#xPx xxxxPx#P Px#xxxPx xxxxpx#P Px#Px xxpx#Px Px#xPx xxxP#xP xxxxPx Px#PXx Px#Px xxxxP#px xxxpx#Px Px#xxPx xxPx#xP Px#xPx xPx#xPx xxPx#P xxxPxx#P Px#xPx xxpxx#xPx xxxxxP#px Px#P#px xxxxPx#P xxP#xxP(xx) xxxxxPx#P xxxxpxx#P xxxxxxP#xpx px#xxxxxpxx xxPx#P Px#pxx xxxPxx#xP xxxpxSx Px#Px xxxxxPx PXx#P

2031a 2031b 2032a 2032b 2033a 2033b 2034a 2034b 2035a 2035b 2036a 2036b 2037a 2037b 2038a 2038b 2039a 2039b 2040a 2040b 2041a 2041b 2042a 2042b 2043a 2043b 2044a 2044b 2045a 2045b 2046a 2046b 2047a 2047b 2048a 2048b 2049a 2049b 2050a 2050b 2051a 2051b 2052a 2052b 2053a 2053b 2054a 2054b 2055a 2055b 2056a 2056b 2057a 2057b 2058a 2058b

E A1 B1 A1 D* C B1 A1 B1 A1 A1 D* A1 B1 D* B1 A1 B1 B1 A1 C B1 A1 A1 A1 B1 A1 B1 A3 B1 C C A1 B1 A1 B1 A1 B1 B1 B1 D* A1 A1 B1 A1 C A1 B1 E B1 A1 A1 A1 B1 A3 C

PXxx#P Px#Px xxxPXS pxx#PX Px#xxPSx xxxxxxpx#Px xxxxP#xP P#xxPx xxxPxx#P Px#Px Px#xxPx Px#Px#P Psx#Px xxPx#xP Px#P#px xxxPx#px xPx#xPx xxxxxPx#xP xxxPx#P Px#pxx xpx#Px xxxxxxPx#P Pxx#Px Px#Px Px#xxPx xxxxxxPx#px Px#xxxxPx xxxxxPx#P xxxxPx xxxPx#P xxPXx xP#px P#xPx xxxxPx#P xP#xxxxPx xxxxPx#xP xxPx#xxPx xxxxxpxx#P xxxPxsx xxxxxPx#P Pxx#Px#xpx PX#Px px#xxPx xxxxxxxpxx#px xpx#xPx xxP#Px Px#Px xxxxpx#xxP PSxx#P xxxxxPx#P xxxxxPXSx Psx#Px xPx#xxxxpxx xxxxPx#P xxxxPx xPXx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 407

Index to the scansion of the Heliand 2059a 2059b 2060a 2060b 2061a 2061b 2062a 2062b 2063a 2063b 2064a 2064b 2065a 2065b 2066a 2066b 2067a 2067b 2068a 2068b 2069a 2069b 2070a 2070b 2071a 2071b 2072a 2072b 2073a 2073b 2074a 2074b 2075a 2075b 2076a 2076b 2077a 2077b 2078a 2078b 2079a 2079b 2080a 2080b 2081a 2081b 2082a 2082b 2083a 2083b 2084a 2084b 2085a 2085b 2086a 2086b

B1 A1 A1 B1 C A1 C C D* B1 A1 B1 A1 B1 A1 C A1 A1 B1 A1 A1 A1 A2 C A1 B1 F F B1 D A1 A1 F F A1 B1 A1 B1 A1 B1 C F B1 A1 D* B1 A1 C A1 B1 D C E* C A1 C

xxPXS PX#Px px#xxxPx xxxxPx#px xxxPXx Pxx#Px xxP#PS xxxxxP#px xxxPx#Psx xxxxxxxxxPx#xP Px#Px xxxxxxPx#P px#xxPx xxxxxpxx#xpx xpxx#xPx xxxpx#px xP#xxxPx xxxxPx#xPx xxxPxx#P xxxPx#Px Px#Px Pxx#xPx xP#xxPS xxxxP#px xP#xxxxpXx xxxxPx#P xxGalileo#P Iudeo#Px xxPx#xpx px#Pxx px#xPx xxxPx#PX xxxxxxGalilea Iudeo#Px Px#xPx xxxxPx#P xPx#xPx xxPx#P Px#xxxxpxx xPX#P xpx#px Iudeo#Px xxPx#P Px#Px xP#xpXSx xxPx#xP pxx#xPx xxpx#px PS#Px xxPxx#P P#Pxx xxPSx PSSx#px xP#px px#xPx xxP#Px

2087a 2087b 2088a 2088b 2089a 2089b 2090a 2090b 2091a 2091b 2092a 2092b 2093a 2093b 2094a 2094b 2095a 2095b 2096a 2096b 2097a 2097b 2098a 2098b 2099a 2099b 2100a 2100b 2101a 2101b 2102a 2102b 2103a 2103b 2104a 2104b 2105a 2105b 2106a 2106b 2107a 2107b 2108a 2108b 2109a 2109b 2110a 2110b 2111a 2111b 2112a 2112b 2113a 2113b 2114a 2114b

A1 E C B1 F A1 B1 D A1 A1 A1 B1 A1 B1 E B1 D* C B1 A1 A1 B1 A1 B1 A1 B1 D A1 A1 C A1 B1 A1 C A1 B1 A1 C A1 B1 A1 C B1 A1 A1 B1 A1 C A1 B1 B1 B1 A1 B1 D B1

407 xxP#xPx pXsXx#xP xxxxxxxPxx xxPx#P Px#xCapharnaum pXx#PX xxxPx#P px#Pxx px#xxxPx Pxx#Px Px#xPx xxxxPx#P Px#Px xxxPx#P xP#P#xpx xxxxPx#P Px#PXx xxxxxxPXx xxPx#px Px#Px Px#xxPx xxxxxPx#xP Px#xPx xxxxxxPxx#P P#xxPx xxxxxpxS#px P#Px#P Px#xPx P#xxxPx xxxP#Px Px#xxxPx xxPx#P xP#xxxpxx xxPXx Px#Px xxPx#xP(xx) Px#xPx xxxxP#px Px#xxpxx xxxxPx#P xPx#xxxPx xxxxxxxP#px xxxxpxx#P Px#xPx PX#P#x xxxxxxPx#xpx xxxxPS#Px xxxxxPsx P#xxPx xxxxxPx#xpx xxxPXS xxxPx#xP pxx#xPx xxxxxPx#P pXsXx xxxxPx#xP

MOSP-Index 1.qxd

5/11/04

408 2115a 2115b 2116a 2116b 2117a 2117b 2118a 2118b 2119a 2119b 2120a 2120b 2121a 2121b 2122a 2122b 2123a 2123b 2124a 2124b 2125a 2125b 2126a 2126b 2127a 2127b 2128a 2128b 2129a 2129b 2130a 2130b 2131a 2131b 2132a 2132b 2133a 2133b 2134a 2134b 2135a 2135b 2136a 2136b 2137a 2137b 2138a 2138b 2139a 2139b 2140a 2140b 2141a 2141b 2142a 2142b

2:49 PM

Page 408

Index to the scansion of the Heliand D* B1 B1 A1 C A1 A1 B1 A1 C E B1 D* C D* C A1 B1 A1 B1 D* F F A1 A1 B1 C A1 A1 B1 B1 A1 C D* E A1 D* C F F F A1 A1 A1 A1 B1 E F B1 A1 A1 C C A1 A3 A1

Px#pxSx xxxxPxx#P xxxxP#xP P#xxPx xxxxxxxPSx Px#Px xxpx#xxPx xxxxxxPx#px Px#xxxPx xxxxxxP#Px PSxx#px xxpxx#xP pxx#pxSx xxxxxxxPXx Px#P#px xxxxP#Px Px#xxpxx xxxPx#P P#xxxPx xxxPX#P xpx#xxPxx xxxxIudeon#Px xxIsraheles pxx#xPx xpx#xPx xxP#xpx xxxPsx xPx#Px xPx#xpxx xxxxxxPx#P xxxxxPx#P Px#Px xxxxpxSx Pxx#xxPxx PSx#px px#xPx Px#P#px xpXSx xxxxxAbrahames xxxIsaakes#xP xxxxIacobes Pxx#Px Px#Px xxPx#xpxx px#xxPx xxpXS#P P#P#xpx xxIudeono#px xxPx#px xPxx#Px xPxx#xxxxPx xxxxxpx#Px xxxxxxPXx Px#Px xxxxPx pxx#Px

2143a 2143b 2144a 2144b 2145a 2145b 2146a 2146b 2147a 2147b 2148a 2148b 2149a 2149b 2150a 2150b 2151a 2151b 2152a 2152b 2153a 2153b 2154a 2154b 2155a 2155b 2156a 2156b 2157a 2157b 2158a 2158b 2159a 2159b 2160a 2160b 2161a 2161b 2162a 2162b 2163a 2163b 2164a 2164b 2165a 2165b 2166a 2166b 2167a 2167b 2168a 2168b 2169a 2169b 2170a 2170b

C A1 C B1 D A1 D A1 A1 B1 A3 B1 A1 B1 A1 B1 E A1 A1 B1 A3 B1 C C B1 E A1 B1 C A1 A1 C A1 C A1 B1 E D* A1 C A1 B1 A1 B1 B1 A1 A1 B1 A1 B1 E D* A1 A1 E* A1

xxxxP#px Px#Px xxPSx xxPx#P P#Px#xP P#xxPx P#PSx Px#xPx Pxx#xPx xxxxxPx#xpx xxxxxPx xxxP#xpx Px#xxxxpXx xxxxxxxPx#P Px#xPx xxxxxP#xP pxSx#P P#xxxxPx xP#xxPx xxPx#xP xxxxxxPx xxxPx#px xxxxpx#Px xxxpXsXx xPXS pxSx#px P#xxxPx xxxxxxxxPx#P xxxxPXx P#xxPx PXSx xxxxxxxP#px Px#xxPx xxpx#Px P#xxPx xxP#xP PSx#P Px#Px#P P#xPx xxP#Px xPXx#Px xxxxxPx#P xPx#xxxxxPx xxxxPx#P xxxxPXS Pxx#xpxx Px#xPx xxxxxPx#P px#xxPx xxxxxxPxx#P PSx#px pxx#pxSx xxxpx#xpxx Px#xPx Px#Px#P Px#xPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 409

Index to the scansion of the Heliand 2171a 2171b 2172a 2172b 2173a 2173b 2174a 2174b 2175a 2175b 2176a 2176b 2177a 2177b 2178a 2178b 2179a 2179b 2180a 2180b 2181a 2181b 2182a 2182b 2183a 2183b 2184a 2184b 2185a 2185b 2186a 2186b 2187a 2187b 2188a 2188b 2189a 2189b 2190a 2190b 2191a 2191b 2192a 2192b 2193a 2193b 2194a 2194b 2195a 2195b 2196a 2196b 2197a 2197b 2198a 2198b

E B1 A1 D* A2 A1 D* B1 A1 B1 D* A1 F C A1 B1 E B1 E C D* A1 A1 B1 E B1 C B1 A1 B1 D B1 A3 B1 A1 A1 A1 B1 D* D E B1 E* C A1 B1 A3 A1 A1 C E A1 A1 C A1 B1

px#Px#px xxxxPxxx#px Px#xxPx Px#P#px Px#pXS pxxx#Px Px#pxsx xxPxx#P Px#Px xxxxxpxx#P xxPx#xP#px xP#xxPx xPXx#xNaim xxxxpx#Px Px#xPx xxPx#P PXx#P xxxxPx#P PXxx#P xxxxxxP#px xxPSx#Psx xPx#Px px#xxxxPx xxxxPx#px pxSx#P xPx#xxP xxxpx#Px xxPx#P pxx#xxPx xxPx#P px#Psx xxxxPx#P xxxxpxx xxxPx#xP xxxxxxPxx#Px P#xPx Px#xxPx xxxxxxP#xP Px#pxxxsx px#Pxx PSx#xP xxxxxPx#P Px#P#xpx xxxxxpx#px Px#Px xxxxxPx#P xxxpxx P#xPx px#xxxxPx xxxP#px PXx#xP xxxPx#xPx Px#xxxxPx xxxpx#px Px#Px xxxxxPx#P

2199a 2199b 2200a 2200b 2201a 2201b 2202a 2202b 2203a 2203b 2204a 2204b 2205a 2205b 2206a 2206b 2207a 2207b 2208a 2208b 2209a 2209b 2210a 2210b 2211a 2211b 2212a 2212b 2213a 2213b 2214a 2214b 2215a 2215b 2216a 2216b 2217a 2217b 2218a 2218b 2219a 2219b 2220a 2220b 2221a 2221b 2222a 2222b 2223a 2223b 2224a 2224b 2225a 2225b 2226a 2226b

B1 D A1 B1 C A1 A1 B1 A1 C A2 B1 A1 B1 E* A1 A1 B1 A1A1 A3A1 A1A1 A3A1 A1A1 CA1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 C A1 A1 C A1s B1 E B1 E B1 F B1 A3 A1 A1 B1 A1 B1 A1 C A1 B1

409 xxxxPx#xP px#Pxx Pxx#Px xxxxpxx#P xxxxpxSx P#xPx xPx#xxxPx xP#xxP xP#xxxPx xxxxxP#px xP#xxpx#P xxxxPx#P xP#xxPx xxxxxxPx#xpx PXx#P#xP px#xxxxPx xPx#xPx xxxxxxPx#xP P#xxxPx#Px xxxxPx#Px pxx#xxxPx#pxx xxxxxxPx#pxx Pxx#xxpxxxSx xxxxxxxPxx#Px xPxx#xpxx#xPx xxxxxxxPx#pxx Px#PXSx xxxxxPx#pxx xPx#xxxxpxx#xpxx xxxPX#Px Px#xxxxxpxx#Px xxxxxxPx#Px Psx#xxxpXx#Px xxxxxxPx#pxx xxxP#px pxx#xPx xP#xxpXx xxxxxxpx#Px px#P#px xxxxP#xP xPSx#P xxxxxP#xxP PS#xpx xxxP#xxP pxx#Israheles xxxPX#P xxxxxPx Pxx#Px Pxx#xxPxx xxxxPx#P xPx#xxxPx xxxxxxPx#P pxx#xPx xxxxxxP#px P#xxxPx xxxxxxPxx#px

MOSP-Index 1.qxd

5/11/04

410 2227a 2227b 2228a 2228b 2229a 2229b 2230a 2230b 2231a 2231b 2232a 2232b 2233a 2233b 2234a 2234b 2235a 2235b 2236a 2236b 2237a 2237b 2238a 2238b 2239a 2239b 2240a 2240b 2241a 2241b 2242a 2242b 2243a 2243b 2244a 2244b 2245a 2245b 2246a 2246b 2247a 2247b 2248a 2248b 2249a 2249b 2250a 2250b 2251a 2251b 2252a 2252b 2253a 2253b 2254a 2254b

2:49 PM

Page 410

Index to the scansion of the Heliand A1 C A1 B1 D* A1 A1 B1 B1 B1 E* A1 A2 C A1 F D* C E B1 C E D* D E C A2 B1 A1 B1 D A1 D* A1 B1 D A1 B1 A1 B1 A1 B1 A3 E A1 A1 A1 D D* B1 B1 A1 A3 C A1 B1

Px#xxxPx xxxxxP#px Px#xPx xxxPx#P Px#Psx Px#Px Px#xpxx xxxPx#P xxP#xxxxP xxxpxx#xpx Pxx#pxSx#px xxPx#Px xxPxx#PS xxxxxxxpx#Px xpx#px#xxPx xxxGalilea-S PX#xxPx#P xxxxpx#Px PSx#px xxxxxxpxx#P xxxpx#Px PXx#P Px#PSx px#P#Px pXSx#px xxP#Px px#xxxxpxS xxxxPx#P PX#xxpxx xxxxpxx#P P#P#Px Px#Px P#xP#xxP xP#xxPx xP#xxpx px#pxxx xpx#xxPx xxxxxPx#xP Px#Px xxPx#P pxx#xxxPx xxxxxxpxx#P xxxxxPx PXx#P Px#xxpxx xxxxxxxPXSx Px#xxxPx P#P#xP xPx#px#px xxxxPx#P xxxxpxx#xP P#xxPx xxxxxPx(xx) xxxP#px xPx#xxxPx xxP#xP

2255a 2255b 2256a 2256b 2257a 2257b 2258a 2258b 2259a 2259b 2260a 2260b 2261a 2261b 2262a 2262b 2263a 2263b 2264a 2264b 2265a 2265b 2266a 2266b 2267a 2267b 2268a 2268b 2269a 2269b 2270a 2270b 2271a 2271b 2272a 2272b 2273a 2273b 2274a 2274b 2275a 2275b 2276a 2276b 2277a 2277b 2278a 2278b 2279a 2279b 2280a 2280b 2281a 2281b 2282a 2282b

B1 A1 C B1 B1 B1 A1 C E D A1 B1 D A1 C A1 B1 A1 D* C A1 B1 D A1 A1 C A2 B1 D* B1 B1 A1 A1 C A1 B1 A1 F D* A1 A1 C A1 C A1 B1 E B1 D* A1 A1 B1 A3 C B1 B1

xxPx#px Px#Px xxpx#PS xxxxPx#P xxxxPx#xP xxxPx#P Px#xPx xxP#Px PXx#P px#Pxx px#xxPx xxxxP#xxP px#Pxx xxxxxxxPx#Px xxxxPxxx Px#Px xxxxxP#xxxP Px#Px Px#xxPSx xxxxxP#px xpx#xxxPx xpx#xxP P#PXS pxx#Px Px#xPx xxP#px Pxx#xpXS xxPx#px xpx#xxpx#Px xxPx#xP xxxxxxPx#px Px#Px Px#xxxxPx xxxxPsx Px#Px xxxxPx#xP xPx#xPx xxxSatanases Px#Pxx PXx#Px Px#xxPx xxxpxSx xP#xPx xxPXx Px#xxxxPx xxxxP#xP PXx#P xxxxPx#P P#xPx#px Pxx#Px Px#Px xxxxxP#xP xxxxPx xxPXx xxxxPX#px xxxP#xP

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 411

Index to the scansion of the Heliand 2283a 2283b 2284a 2284b 2285a 2285b 2286a 2286b 2287a 2287b 2288a 2288b 2289a 2289b 2290a 2290b 2291a 2291b 2292a 2292b 2293a 2293b 2294a 2294b 2295a 2295b 2296a 2296b 2297a 2297b 2298a 2298b 2299a 2299b 2300a 2300b 2301a 2301b 2302a 2302b 2303a 2303b 2304a 2304b 2305a 2305b 2306a 2306b 2307a 2307b 2308a 2308b 2309a 2309b 2310a 2310b

A3 B1 D* A1 D* F B1 A1 C A1 A1 C A2 C B1D B1 F D A1 B1 A3 E A1 C A1 C A1 B1 A1 A1 D* B1 A3 A1 D* B1 E C A1 B1 C A1 A1 C A1 C E A1 E C B1 A1 A1 C C B1

xxxxxxPx xxxPX#P xpx#xPxx#px px#xpxx P#P#xxPxx xxIudeon#xxP xxxpxx#P xP#xxPx xxpxSx Px#Px Px#xxPx xxxP#px Pxx#PS xxxxxP#px xPx#xxpx#Pxx xxxxxxPx#xP xxGalilæo#P px#Px#P P#xxxPx xxxPx#P xxxxxxPx PS#xP xPxx#PX xxxxpxsx Px#Px xxpx#px xPx#xxxxPx xxxxxPx#P Px#xxxPx xxxxxPx#Px Px#xxP#px xxPxx#P xxxxPx pxx#PX Px#Psx xxxxpxx#P pxSx#xpx xxxxPsx P#xPx xxxpxx#xpx xxxxxxxP#px Px#xPx xPx#xxPx xxxPxx Px#xPx xxxxxxxxpx#Px PXx#P P#xxPx pXSx#xP xxxxxP#px xxxPx#px Px#Px Px#xxPx xxxxxxxxxxP#px xxpx#Px xxxPX#P

2311a 2311b 2312a 2312b 2313a 2313b 2314a 2314b 2315a 2315b 2316a 2316b 2317a 2317b 2318a 2318b 2319a 2319b 2320a 2320b 2321a 2321b 2322a 2322b 2323a 2323b 2324a 2324b 2325a 2325b 2326a 2326b 2327a 2327b 2328a 2328b 2329a 2329b 2330a 2330b 2331a 2331b 2332a 2332b 2333a 2333b 2334a 2334b 2335a 2335b 2336a 2336b 2337a 2337b 2338a 2338b

A1 B1 A1 C D* A1 C B1 D* B1 B1 B1 D* B1 A1 B1 B1 A1 A1 B1 F C A1 B1 E C A1 C D* B1 C C A1 C A1 B1 C B1 E A1 E A1 D B1 E* C A1 C A1 B1 D B1 D C A1 A1

411 Px#xPx xxxxxPx#P Px#xxxxxPx xxxxxP#px px#xxpx#pxx xxxxxPx#Px xxxpx#Px xxPx#P pXx#PxX xxxxxxpx#xP xxPx#P xxxxxxpx#xP xxxPx#Psx xxpxxx#xpx xPx#Px xxxxPx#P xxxxxPx#P Pxx#Px Px#Px xxxxxxPx#xpx PSx#Iudeon xxpx#Px P#xxpxx xxxxxxxxPx#P PS#xpx xxxP#Px PX#xxxpXx xxxxxP#px Px#P#px xxxx(xx)xxxxPx#P xxxP#px xxxxpx#Px xPx#xPx xxxP#Px Px#xxpXx xxxPx#P xxPXx xxxxPx#xP pxx#xxP xPx#Px PXx#px xxxxxPx#xPx P#pxSx xxxxxxxPx#px xPxSx#xpx xxxP#Px Px#xxxxPx xxxxxxxP#px P#xxxPx xxxxxPx#P px#Pxx xxxxxPX#P P#pxSx xpx#Px Px#Px xxxxxxxPx#Px

MOSP-Index 1.qxd

5/11/04

412 2339a 2339b 2340a 2340b 2341a 2341b 2342a 2342b 2343a 2343b 2344a 2344b 2345a 2345b 2346a 2346b 2347a 2347b 2348a 2348b 2349a 2349b 2350a 2350b 2351a 2351b 2352a 2352b 2353a 2353b 2354a 2354b 2355a 2355b 2356a 2356b 2357a 2357b 2358a 2358b 2359a 2359b 2360a 2360b 2361a 2361b 2362a 2362b 2363a 2363b 2364a 2364b 2365a 2365b 2366a 2366b

2:49 PM

Page 412

Index to the scansion of the Heliand A1 B1 F C A1 B1 A1 C A2 B1 A3 D A1 B1 A1 C A1 B1 D* B1 A1 B1 A1 B1 A1 C E A1 B1 B1 A2 B1 D* A1 B1 A1 D* B1 A1A1 B1 E A1 A2 F D* A1 B1 B1 B1 A1 A3 A1 A1 B1 D A1

P#xxPx xxxxxPx#P Iudeo#Px xxP#Px xxPx#xPx xxxxPx#P Px#xxxPx xxP#px PS#PS xxxxPx#px xxxxxPx pXsXx Px#Px xxPx#xxP Px#xxxxxpXx xxxxxP#px xxxpx#xpxx xPx#px Px#xPx#P xxxPx#xP Px#Px xxxxxPXS Px#xPx xxxPxx#xP xPx#xxPx xxxxxPsx pxSx#xP xxPx#xpxx xxPx#px xxPx#P px#xPS xxxxxxxPX#P P#xxP#px px#xxPx xxxxxxxpXx#P Pxx#Px xPx#xxPx#P xxxPx#xP Px#xxPx#Px xxxPx#P PSx#px Px#Px pxx#PS xxxIudeono#xP Px#PSx xPx#xpxx xxxPxx#P xxxxPxx#P xxPx#P xPx#xPx xxxxxPx PXx#xPx Px#xxxxPx xxxxpxx#xP px#Pxx xxPx#xPx

2367a 2367b 2368a 2368b 2369a 2369b 2370a 2370b 2371a 2371b 2372a 2372b 2373a 2373b 2374a 2374b 2375a 2375b 2376a 2376b 2377a 2377b 2378a 2378b 2379a 2379b 2380a 2380b 2381a 2381b 2382a 2382b 2383a 2383b 2384a 2384b 2385a 2385b 2386a 2386b 2387a 2387b 2388a 2388b 2389a 2389b 2390a 2390b 2391a 2391b 2392a 2392b 2393a 2393b 2394a 2394b

B3 A1 A1 B1 A1 C D C D* A1 A1 B1 B1 A1 B1 A1 A1 C A1 C C D E B1 A3 B1 A1 A1 A1 B1 A1s A1 C B1 A1 B1 A1 C A1 B1 B1 A1 B1 A1 B1 A1 A1 B1 E A1 A3 A1 A1 B1 B1 B1

xxxxxpx pxx#Px Px#xxxxPx xxxxxxPx#xpx xpx#xxPx xxpx#px P#Pxx xxxxpx#Px xpXx#xP#px xxxxxxxxPx#xPx xxPx#xxxPx xxxxPxx#P xxxPx#P pxx#Px xxPx#P Px#Px Px#xxPx xxxpx#Px Px#Px xxP#px xxPXx pXsXx PSx#P xxxxxxxxpxx#px xxxxxxxxPx xxxPx#P xxxPx#Px xPx#Px xP#xxxxPx xxxPxx#xpx pxS#px xxPx#Px xxP#Px xxxPx#P Px#Px xxxxPx#xpx Px#xPx xpx#px px#xxxpxx xPX#P xxxPx#P Px#Px xxxPx#P(xx) xPx#Px xxxxP#xP xPx#Px PSx#xxPx xxxPx#P pXSx#P Px#xPx xxxxxPx xxPx#xPx Px#xxxPx xxxP#xpx xxxxxPx#xP xxxxP#xP

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 413

Index to the scansion of the Heliand 2395a 2395b 2396a 2396b 2397a 2397b 2398a 2398b 2399a 2399b 2400a 2400b 2401a 2401b 2402a 2402b 2403a 2403b 2404a 2404b 2405a 2405b 2406a 2406b 2407a 2407b 2408a 2408b 2409a 2409b 2410a 2410b 2411a 2411b 2412a 2412b 2413a 2413b 2414a 2414b 2415a 2415b 2416a 2416b 2417a 2417b 2418a 2418b 2419a 2419b 2420a 2420b 2421a 2421b 2422a 2422b

D* B1 D* A1 A1 B1 E B1 A1 A1 D* A1 A3 B1 C B1 D* A1 B1 A1 A1 B1 A3 C A1 C A3 B1 A1 B1 A1 B1 A1 C B1 A1 D* A1 E B1 A3 C B1 A1 A3 D A1 A1 C C A1 A1 A1 C A1 B1

xPx#pXSx xxxxPx#P Px#PSx xxPx#Px P#xPx xxP#xP PXx#xpx xxxxPx#P xxxPx#Px xPx#Px Px#Psx xxpxx#Px xxxxxPx xxxP#xP xxxxxxxpx#Px xxxxpxx#xP xPx#P#px xxPx#xPx xxxxPx#P Px#xPx Px#xPx xxxxxP#xP xxxxxPx xxpx#Px Pxx#Px xpx#px xxxxxPx xxxP#xP P#xxxxxpxx xxxxxP#xxP pxx#xxxxPx xxxxPx#P pxx#xxPx xxxxxxxxxxpx#Px xxxPx#P Px#Px xPx#xxPxx xPx#Px PSx#px xxPx#px xxxxpXx xP#px xxPS#P Px#xPx xxxxxxPx P#Pxx Px#Px P#xxxPx xxPSx xxxP#px(xx) xxpxx#xxPx Px#Px Px#xxpxx xxpxSx pxx#Px xxxxxPxx#P

2423a 2423b 2424a 2424b 2425a 2425b 2426a 2426b 2427a 2427b 2428a 2428b 2429a 2429b 2430a 2430b 2431a 2431b 2432a 2432b 2433a 2433b 2434a 2434b 2435a 2435b 2436a 2436b 2437a 2437b 2438a 2438b 2439a 2439b 2440a 2440b 2441a 2441b 2442a 2442b 2443a 2443b 2444a 2444b 2445a 2445b 2446a 2446b 2447a 2447b 2448a 2448b 2449a 2449b 2450a 2450b

C A1 B1 B1 A1 B1 B1 A1 A1 A1 D B1 B1 B1 B1 A1 A3 A1 E B1 C A1 A1 C A1 C A1 C C A1 A3 C A1 B1 A1 B1 C A1 A1 B1 A1 D D* B1 A1 A1 D C D* B1 A3 B1 B1 A1 A3 A1

413 xxxxpx#Px Px#xPx xxxPx#P xxxxxPx#P Px#xPx xxxPx#P xxxPXS Px#Px xpx#xxxPx PS#pxx P#Pxx xxxxpxx#P xxxP#xxxP xxxxxxxxPx#px xxxxxxxPx#xP Px#Px xxxxPx pxx#Px PSx#xP xxxxPx#P(xx) xxPXx Px#Pxx Px#xxPx xxxxpx#Px Pxx#Px xxxpx#px PX#xxxpXx xxpx#Px xxxpxSx pxX#xPx xxxxxxpXx xxP#px Px#Px xxxxxxPx#P Px#xxPx xxPx#xP xxxxPsx Px#xPx xP#xxxPx xxPx#P xPxx#Px pXsXx xP#xPx#P xxxxPXS Px#xxxxxpXx px#xxxpxx P#PSx xxxxP#px Px#pxSx xxPx#px xxxxxPx xxxxxxPx#P xxxxxxPx#P Px#Px xxxxxxpxx Px#Px

MOSP-Index 1.qxd

5/11/04

414 2451a 2451b 2452a 2452b 2453a 2453b 2454a 2454b 2455a 2455b 2456a 2456b 2457a 2457b 2458a 2458b 2459a 2459b 2460a 2460b 2461a 2461b 2462a 2462b 2463a 2463b 2464a 2464b 2465a 2465b 2466a 2466b 2467a 2467b 2468a 2468b 2469a 2469b 2470a 2470b 2471a 2471b 2472a 2472b 2473a 2473b 2474a 2474b 2475a 2475b 2476a 2476b 2477a 2477b 2478a 2478b

2:49 PM

Page 414

Index to the scansion of the Heliand D B1 A1 C B1 A1 A1 A1 B1 A1 E A1 A1 B1 B1 or E A1 D B1 B1 A1 C A1 B1 A1 A1 B1 A1 B1 B1 B1 A1 B1 A1 D* C B1 C C A1 B1 A3 A1 E C A1 C A1 C A1 B1 A1 C B1 B1 A1 C

px#PSx xxxxxpxx#P xxxpxx#Px xxxP#Px xxP#xP xxxPx#xPx xxxPx#Px pxS#Px xxxP#xpx pxxx#Px PXx#px xxxxxxpxx#Px Px#xPx xxxxxPx#P xxPXx#P P#xPx xpx#PSx xxxpxx#P xxxPx#P P#xxPx xxPsx Px#Px xxxxxxPx#xP Px#Px Px#xPx xxxpxx#xpx Px#xxxxxPx xxxPx#P xxxxP#xxP xxxxxxPx#P Pxx#Px xxxxxxPx#xP px#xxxxPx xxPx#xxxPx#P xxPSx xxPx#P xxP#px xxP#px Px#xxPx xxxPx#xP xxxxxxPxx P#xPx PSxx#P xxP#px Px#Px xpXsXx xPx#xxPx xxP#px xpxx#xPx xxxxPx#P xP#xPx xxxP#px xxxxxP#xpx xxpxx#P px#xxPx xxxP#px

2479a 2479b 2480a 2480b 2481a 2481b 2482a 2482b 2483a 2483b 2484a 2484b 2485a 2485b 2486a 2486b 2487a 2487b 2488a 2488b 2489a 2489b 2490a 2490b 2491a 2491b 2492a 2492b 2493a 2493b 2494a 2494b 2495a 2495b 2496a 2496b 2497a 2497b 2498a 2498b 2499a 2499b 2500a 2500b 2501a 2501b 2502a 2502b 2503a 2503b 2504a 2504b 2505a 2505b 2506a 2506b

C A1 A1 B1 A1 C A1 E A3 A1 B1 A1 B1 C A3 C C E A1 C A1 A1 A1 B1 A1 B1 A1 C A1 B1 A1 E B1 C A3 B1 A3 C C B1 A1 B1 C C A1 B1 B1 A1 A3 C A1 B1 A1 B1 A3 B1

xxxpx#Px xxxPx#Px px#xxPx xxxxxxPx#P Px#Px xxxP#px Px#pxx Px#xxpx xxxxxPx xxPx#Px xxPx#xpx Px#Px xxxPx#px xxxxP#px xxxxPx xxxxpXSx xxpxSx pXSx#xP pxx#xxPx xxxxxxP#px Pxx#Px Px#xxPx pxx#xpxxx xxPx#P xP#xxxxxPx xxxxxPxx#P PSx#Px xxxPSx pxx#xPx xxxxxPx#P Px#Px PXx#px xxxPx#P xxpXSx xxxxxxPx xxxxxxPx#P xxxxPx xxP#px xpx#Px xxxxxxxxPx#P Px#Px xxxxxxxPx#px xxpx#Px xxxxxxxP#px px#xxPx xxpxx#P xxPx#px Px#Px xxxxxPx xpxsx xPx#xxxPx xxxxxPx#px xxxxxxPx#xPx xxxPx#xpx xxxxPx xxxxpx#xP

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 415

Index to the scansion of the Heliand 2507a 2507b 2508a 2508b 2509a 2509b 2510a 2510b 2511a 2511b 2512a 2512b 2513a 2513b 2514a 2514b 2515a 2515b 2516a 2516b 2517a 2517b 2518a 2518b 2519a 2519b 2520a 2520b 2521a 2521b 2522a 2522b 2523a 2523b 2524a 2524b 2525a 2525b 2526a 2526b 2527a 2527b 2528a 2528b 2529a 2529b 2530a 2530b 2531a 2531b 2532a 2532b 2533a 2533b 2534a 2534b

A1 B1 C B1 C B1 B1 A1 B1 C B1 B1 A1 B1 A1 B1 D* A1 A1 – A1 C A1 C A1 B1 D* B1 A1 C B1 A1 A1 C A1 B1 A1 B1 A3 B1 B1 A1 D* A1 E* B1 A1 B1 C A1 B1 A1 B1 D* A1 B1

Px#xxxPx xxxxxxxxPx#P xxxpXSx xxPx#px xpx#Px xxxxPx#P xxxPx#xpx Px#xPx xxxPx#P xxpXsXx xxxPx#xpx xxxPX#px Px#xpxx xxxxPx#P Px#xxxxPx xxxxxPx#px xPSx#Psx Px#Px xPx#Px – xxxxxxxP#xPx xxP#pxx xxxP#xPx xxxpXsXx Px#xPx xxxPx#xP xPxx#xP#px xxxxPx#px Px#xPx xxxpx#Px xxxxxP#xxxP xP#xPx Px#xxxPx xxxpx#Px xPx#xPx xxxxPx#xP xP#xxPx xxP#xP xxxxPx xxxxxxxxpXx#P xxxPXsx Px#Px Px#xxxPx#P pxx#Px xPxSx#px xxxxPx#P px#xxxxpXx xxxPx#xP xxPXx px#xxPx xxxPx#px Px#xPx xxxxPx#px xPX#P#px Px#xpxx xxpXS#P

2535a 2535b 2536a 2536b 2537a 2537b 2538a 2538b 2539a 2539b 2540a 2540b 2541a 2541b 2542a 2542b 2543a 2543b 2544a 2544b 2545a 2545b 2546a 2546b 2547a 2547b 2548a 2548b 2549a 2549b 2550a 2550b 2551a 2551b 2552a 2552b 2553a 2553b 2554a 2554b 2555a 2555b 2556a 2556b 2557a 2557b 2558a 2558b 2559a 2559b 2560a 2560b 2561a 2561b 2562a 2562b

A1 B1 B1 A1 C A1 A1 C C B1 B1 A1 B1 D* D* A1 D* A1 A1 A1 B1 B1 D* A1 B1 A1 E A1 A1 A1 B1 A1 A1 B1 A1 B1 D* A1 A1 B1 B1 D* D* B1 A1 C A1 B1 A1 C A1 B1 A1 B1 A1 B1

415 Px#xpxx xxxxPx#P xxPx#xP xPx#xPx xxxP#px Px#Px xPx#xxxPx xpx#px xxxP#px xxxxxxpxx#px xxxxxpXx#xP Px#Px xxxxxpxx#P xxPx#Pxx Px#PSx Px#Px Px#xxxPsx Px#Px pxx#Px xxxxPX#Px xxPx#px xxxxxPx#xxP xPx#PXx xPx#xPx xxP#xxP xPx#Px xpxSx#xP xxPx#Px Px#xxPx Px#Px xxxxxPx#P Px#xPx PS#xxxPx xxxxxxxPx#xP Px#Px xxxxxPx#P xxP#xpxx#P xPx#xPx Px#xxPx xxxxxxxxPx#P xxxPS#P Px#Pxx PX#Px#P xxxxxxxPx#P xPx#xxxPx xxxpx#Px xPx#xPx xxxxxxxPx#P px#xPx xxxxxP#px px#xxxxPx xxxxxxPx#xpx xPx#xxxxPx(xx) xxxPx#xpx xPx#xxxPx xxxPx#xP

MOSP-Index 1.qxd

5/11/04

416 2563a 2563b 2564a 2564b 2565a 2565b 2566a 2566b 2567a 2567b 2568a 2568b 2569a 2569b 2570a 2570b 2571a 2571b 2572a 2572b 2573a 2573b 2574a 2574b 2575a 2575b 2576a 2576b 2577a 2577b 2578a 2578b 2579a 2579b 2580a 2580b 2581a 2581b 2582a 2582b 2583a 2583b 2584a 2584b 2585a 2585b 2586a 2586b 2587a 2587b 2588a 2588b 2589a 2589b 2590a 2590b

2:49 PM

Page 416

Index to the scansion of the Heliand B1 A1 A1 C A1 C B1 A1 D* B1 A1 E A1 C A1 B1 A1 C D* B1 A3 A1 D C D* C D* A1 A1 A1 A3 A1 A1 B1 A1 B1 D* B1 A1 B1 D* B1 A1 A1 A2 D* F A1 A1 B1 A1 C A1 B1 B1 B1

xxxPx#xpx Px#xPx Px#xxxxPx xxxxP#px Px#Px xxpx#px xxxxPx#P Px#Px Px#xxPxx xxxxxPx#P px#xxxxPx xxxPSx#px Px#xPx xxxxxxpx#Px Px#xxxPx xxPx#xpx P#xPx xxxP#px Px#xxPXx xxxxxxPx#P xxxxpxx Px#Px P#PSx xxP#px Px#PXx xPsx Px#Px#P Px#Px Px#xxpXx Px#PX xxxxPx Px#Px xPx#xPx xxxxxPx#P Pxx#Px xxxxxxPx#xpx Px#Px#P xx(xx)Px#px xPx#xxxPx xxxxxPxx#P xPx#PSx xxxPx#P Px#xxPx xpX#xxPx xPx#PS Px#PSx Satanas#Px xxxPx#Px xPSx#Px xxxxxPx#xpx pxx#xPx xxP#Px Px#xxxPx xxxxxPx#P xxpxx#px xxxxxPx#P

2591a 2591b 2592a 2592b 2593a 2593b 2594a 2594b 2595a 2595b 2596a 2596b 2597a 2597b 2598a 2598b 2599a 2599b 2600a 2600b 2601a 2601b 2602a 2602b 2603a 2603b 2604a 2604b 2605a 2605b 2606a 2606b 2607a 2607b 2608a 2608b 2609a 2609b 2610a 2610b 2611a 2611b 2612a 2612b 2613a 2613b 2614a 2614b 2615a 2615b 2616a 2616b 2617a 2617b 2618a 2618b

E C A1 B1 A1 B1 D* C C A1 D*A1 A3A1 EA1 B3A1 A1 B1 D* B1 A1 C D B1 A1 B1 or E C A1 A1s C C A1 A1 C D C A1 B1 A1 A1 C B1 B1 A1 A1 B1 C A1 A1A1 A3A1 A1A1 B1 B1 A1 A1 B1 B1 B1

xxPSx#px xxP#px Px#xxxpXx xxxxxPxx#xpx xPx#xxxxPx xxxxpXxsx Px#pxx#P xxpx#Px xxxxxpx#PX xxxxPx#Px pxx#xxPxx#Px xxxxPx#xPx Px#xxxP#xPx xxxxxP#xPx pxx#xxPx xxxPxx#px Pxx#pXSx xxxxxPx#P Px#xPx xxxxxPSx P#pxx#P xxxxxxPx#P Px#xxPx xPXx#P xxxxxPxx Px#Px PS#px xxxxxPsx xpXSx Pxx#Px Px#xPx xxP#px px#PSx xxxxP#Px xPx#xxPx xxxPx#px P#xxPx xxxxxxxPx#Px xxPsx xxxxxxPx#px xxxPx#P xPx#Px Px#xxPx xxxxxxxxxpXx#xP xxpXSX Pxx#Px PSX#pxx#Px xxxxxxxPx#Px px#xxxPx#Px xxxPx#xxpx xxxPx#xpx Px#Px Pxx#pxx xxxxPx#P xxxPx#xpx xxxP#xpx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 417

Index to the scansion of the Heliand 2619a 2619b 2620a 2620b 2621a 2621b 2622a 2622b 2623a 2623b 2624a 2624b 2625a 2625b 2626a 2626b 2627a 2627b 2628a 2628b 2629a 2629b 2630a 2630b 2631a 2631b 2632a 2632b 2633a 2633b 2634a 2634b 2635a 2635b 2636a 2636b 2637a 2637b 2638a 2638b 2639a 2639b 2640a 2640b 2641a 2641b 2642a 2642b 2643a 2643b 2644a 2644b 2645a 2645b 2646a 2646b

B1 E D B1 A3 D B1 A1 A3 C C E B1 A1 A1 C A3 C A1 B1 C A1 E A1 A1 B1 A1 D A1 C B1 B1 B1 E B1 A1 A3 C A3 C A1 B1 D* C C A1 B1 C A1 B1 D* A1 A1 C B1 A1

xxPx#px pXx#P P#pXSx xxPx#px xxxxxxPx px#Pxx xxxPx#P pXx#Px xxxxPx xpXSx xxpXSx pXSx#xP xxxPxx#P Pxx#Px xPx#xPx xxpXSx xxxxPx xxP#Px Px#xxxxpXx xxxxxP#xP xxxP#Px pxx#Px PS#xP xxPx#Px pxx#xxPx xxP#xpx px#xxPx px#Px#P xPx#xPx xxxxxxxxxP#px xPx#P xxPX#P xxxPx#px PXxx#P xxPXS Px#xPx xxxxxxPx xpXSx xxxxpxx xP#px Px#Px xxpxx#P xPx#pXSx xxpx#Px xxxPSx PXSx xxxxxxPx#P xpx#Px xxpx#xxPx xxPX#px Px#P#Px pxx#xpxxx xxxP#xPx xxxpXSx xxPsx#P Px#Px

2647a 2647b 2648a 2648b 2649a 2649b 2650a 2650b 2651a 2651b 2652a 2652b 2653a 2653b 2654a 2654b 2655a 2655b 2656a 2656b 2657a 2657b 2658a 2658b 2659a 2659b 2660a 2660b 2661a 2661b 2662a 2662b 2663a 2663b 2664a 2664b 2665a 2665b 2666a 2666b 2667a 2667b 2668a 2668b 2669a 2669b 2670a 2670b 2671a 2671b 2672a 2672b 2673a 2673b 2674a 2674b

A1 B1 F A1s A3 C E C C A1 E* B1 D* B1 A1 C A2 B1 A1 C A1 C D* B1 D* A1 C B1 A1 A1 A1 B1 A1 C F F A1 B1 D* B1 D* B1 A1 B1 A1 C A1 A1 C A1 A1 B1 A1 B1 C C

417 xPx#xxxPx xxxxxPx#P xxxGalilæo#P xpx#P#px xxxxxxPx xxxxxxxxP#px xPSx#xpx xxP#px xxPSx Px#Px xPXSx#xpx xxxxPx#px(xxx) xP#xPsx xxxPx#xP P#xxxxxpxx xxxxpx#Px xP#xxxpxS xxxPx#xP xP#xxxxxxpxx xxxxxxxxxP#px Px#Px xxxxP#Px xxPx#xxxPx#P xxxxxxPS#P xPx#xxxPXx Px#xPx xxPSx xxxxpXx#px xxxPxSx Px#xPx Pxx#Px xxxxxxPx#px xxPx#Px xxxxpx#Px xPx#xxIudeon xxxxxGalilæo#P xPx#xpxx xxxxPxx#P xpx#xP#px xxxxxxPsx#P xPx#PSx xxxxP#xxP Px#Px xxxxPx#P Px#xPx xxxxxpx#px xPx#xPx Px#Px xxxpx#px Px#xPx Px#Px xxxxPx#P Pxx#Px xxxxxpx#xxP xxxxxPXx xxxxpx#Px

MOSP-Index 1.qxd

5/11/04

418 2675a 2675b 2676a 2676b 2677a 2677b 2678a 2678b 2679a 2679b 2680a 2680b 2681a 2681b 2682a 2682b 2683a 2683b 2684a 2684b 2685a 2685b 2686a 2686b 2687a 2687b 2688a 2688b 2689a 2689b 2690a 2690b 2691a 2691b 2692a 2692b 2693a 2693b 2694a 2694b 2695a 2695b 2696a 2696b 2697a 2697b 2698a 2698b 2699a 2699b 2700a 2700b 2701a 2701b 2702a 2702b

2:49 PM

Page 418

Index to the scansion of the Heliand B1 C A1 B1 E C A3 E C F B1 A1 A1 B1 A2 C B1 A1 A1 B1 A1 B1 A1 A1 F B1 A1 B1 or E A1 D A3 A1 A3 A1 A1 C A1 B1 A1 A1 C E D* B1 A3 A1 A1 C F D A1 E C A1 A2 B1

xxxxPx#P xxxP#px px#xPx xxxxxxPx#px PSx#P xxxxpX#px xxxxxxPx PXxx#P xxxPSx Iudeo#Px xxPx#P Px#xPx Pxx#xPx xxxxxxPx#px P#xxxxPS xxxxxxpx#Px xxxxxxxPx#px Px#Px Px#xPx xxxxxPx#P xPx#xPx xxxxPx#px xxxPx#Px Px#xPx Iuðeono#xPx xxxxxxxxxPx#px xxPx#Px xxxxxxPXx#px P#xPx xxxxxP#Pxxx xxxxxxxPx xxxxxxxxpxx#Px xxxxxxPx Pxx#Px px#xxxxPx xxxxxxxpx#Px PS#xxxpxx xxxxxxxPx#px xPXx#Px P#xxxxxPx xxxPXx PXx#px pXx#PxX xxxxPx#xP xxxxxxPx PX#Px xpXx#xxxxpXx xxxxxP#Px Iohannes#xxPxx px#PXS Px#xPx Psxx#P xxxpx#pxx pxx#xPx P#xxPS xxxpxxx#P

2703a 2703b 2704a 2704b 2705a 2705b 2706a 2706b 2707a 2707b 2708a 2708b 2709a 2709b 2710a 2710b 2711a 2711b 2712a 2712b 2713a 2713b 2714a 2714b 2715a 2715b 2716a 2716b 2717a 2717b 2718a 2718b 2719a 2719b 2720a 2720b 2721a 2721b 2722a 2722b 2723a 2723b 2724a 2724b 2725a 2725b 2726a 2726b 2727a 2727b 2728a 2728b 2729a 2729b 2730a 2730b

A1 F B1 B1 F E A1 B1 A1 B1 D B1 B1 C A1 C F C A2 C B1 C A1 B1 A1 C A1 B1 A1 B1 A1 C A1 A1 A3 B1 A1 B1 D A1 C A1 A1 A1 A1 C A3 C C C C F A1 C D* B1

Pxx#pxx xxxxxxxxIudeono#pX xxpxsx#xP xPx#P Erodes#xxxPxx pXSx#P Pxx#xxxxxPx xxxxPx#P px#xPx xxxPx#P pX#Pxx xxxxP#xP xpX#xpx xxxxxP#Px P#xxPx xxxxxxP#px Iohannes#xPx xxxpx#Px PXx#pXS xxxxpx#pxx xPx#P xxP#Px Px#xxxxPx xxxPx#px xPx#xxPx xxxxP#Px xP#xxxxxxPx xxxxxxxPx#P xPx#xxPx xxxPx#px xPx#xxxPx xxxxxxpXsX Pxx#xPx xxPx#Px xxxxPx xxxxxxPx#px Px#xPx xxxxPx#P PSxxx Px#Px xxxxxxxxPXx PXSx pxSx#xPx xxPx#xxPx xxpxx#xPx xxxxxxP#Px xxxxxPx xxpx#Px xxxxxPsx xxpx#Px xxxxxxPsx Iudeo#pXx Px#pxx xxxP#Px Px#Psx xxxPx#P

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 419

Index to the scansion of the Heliand 2731a 2731b 2732a 2732b 2733a 2733b 2734a 2734b 2735a 2735b 2736a 2736b 2737a 2737b 2738a 2738b 2739a 2739b 2740a 2740b 2741a 2741b 2742a 2742b 2743a 2743b 2744a 2744b 2745a 2745b 2746a 2746b 2747a 2747b 2748a 2748b 2749a 2749b 2750a 2750b 2751a 2751b 2752a 2752b 2753a 2753b 2754a 2754b 2755a 2755b 2756a 2756b 2757a 2757b 2758a 2758b

C B1 B1 A1 F C A1s A1 E B1 C A1 F B1 A1 C A1 B1 A1 A1 D* A1 A1 A1 A3 A1 B1 A1 B1 A1 B1 B1 A1 B1 D* B1 A3 A1 A1 C B1 A1 A1 C A1 B1 A1 B1 B1 A1 B3 C C A1 A1 B1

xP#px xxxxPx#P xxPx#xpx Px#Px Iudeono#xPx xxxxxxPsx pXS#px Px#xPx pxsxx#xxP xxxPx#P xxpXSx Px#pxx Iudeon#xxxPsx xxxPS#px Px#xxxPx xxxxxPsx px#xPx xxP#xP Px#xPx Px#Px Px#xPsx px#xxPx P#xxxPx pxx#Px xxxPx Px#Px xxxxpxx#P xPx#xpxx xxxPx#P Px#Px xPx#P xxxxPx#P Px#xPx xxxxxxPx#P Px#xxxxxPSx xxPx#P xxxxxxPx px#xPx px#xPx xxP#px xxxPx#px Px#pxx xPXx#xPx xxxxxpx#pxx xP#xxxxpxx xxxxxxxPx#xpx Px#xxxxPx xxxxPx#P xxxxPx#P Px#Px xxxxxpx xxxxxPSx xxxxxxpxSx Pxx#Px Px#Px xxxxxxxxPx#xP

2759a 2759b 2760a 2760b 2761a 2761b 2762a 2762b 2763a 2763b 2764a 2764b 2765a 2765b 2766a 2766b 2767a 2767b 2768a 2768b 2769a 2769b 2770a 2770b 2771a 2771b 2772a 2772b 2773a 2773b 2774a 2774b 2775a 2775b 2776a 2776b 2777a 2777b 2778a 2778b 2779a 2779b 2780a 2780b 2781a 2781b 2782a 2782b 2783a 2783b 2784a 2784b 2785a 2785b 2786a 2786b

A1 B1 B1 A1 A1 A1 C D A3 E B1 A1 A1 A1 D* C A1 D D* B1 A1 C A1 B1 B1 D B1 A1 A1 B1 D* F A1 A1 D* B1 A1 C A3 C C B1 D* B1 A1 B1 A3 A1 C B1 A1 B1 C A1 C C

419 Px#xPx xxxxxPx#P xxxxpx#xxP P#xPx px#xxxxPx xxxxxPx#Px xxxPsx px#P#xP xxxxPx PSx#xP xxPx#P xPx#pxx P#xxxxPx px#xxPx pxxx#Psx xxpx#Px xPx#xPx PsXx xxPxx#xxPSx xxPx#P Pxx#pxx xxxxxxpx#Px xpx#xxxpxx xxxxxPx#px xxPxx#P pxSSx xxxxPx#P Px#Px xPx#xxxxxPx xxxPx#xP xpxx#xxxPSx xxxxxxxIohannes xxxPx#Px Px#Px xPx#xxPsx xxxxxPx#P xPx#xxxPx xxxxxxxpx#px xxxxxxpXx xxxxxpx#Px xP#Px xxxPXsX Px#xxxPsx xxxxxpx#P Px#xPx xxxP#xP xxxxPx Px#Px xPsx xxxxxxPx#xP px#xxxpxx xxxxxPx#P xxPsx Pxx#Px xPXx xxxxxxxpX#Px

MOSP-Index 1.qxd

5/11/04

420 2787a 2787b 2788a 2788b 2789a 2789b 2790a 2790b 2791a 2791b 2792a 2792b 2793a 2793b 2794a 2794b 2795a 2795b 2796a 2796b 2797a 2797b 2798a 2798b 2799a 2799b 2800a 2800b 2801a 2801b 2802a 2802b 2803a 2803b 2804a 2804b 2805a 2805b 2806a 2806b 2807a 2807b 2808a 2808b 2809a 2809b 2810a 2810b 2811a 2811b 2812a 2812b 2813a 2813b 2814a 2814b

2:49 PM

Page 420

Index to the scansion of the Heliand C A1 A1 B1 B1 B1 A1 B1 C A1 A1 C A1 A1 F A1 A1 A1 D* C E B1 E A1 A1 B1 F A1 A1 B1 A1 B1 C E E B1 E* F D* A1 A1 C D B1 A1 C D* B1 C E C D D B1 A1 B1

xxxpx#Px P#xPx px#xPx xxxxxPx#xpx xxPx#xP xxPx#xP P#xxxpXx xxxxxxPX#P xpXSx Pxx#Px xPxx#Px xxxxxxxxpx#Px P#xPx Px#Px px#xxIohannen xPxx#pxx Px#xPx xxxxxPx#xPx Px#Psx xxxxP#px PSx#P xxPxx#px PSx#P Px#Px Px#Px xxxxxxxPx#px Iohannes#Pxx PXSx PXsxx xxxxPx#P Px#xPx xxxxPx#P xxxPXx PXx#px Pxxx#P xxxxP#xpx Px#Px#xP xxxxIudeono#pX Px#xxPXx Px#Px Px#xPx xxxxxxxxP#px px#Pxx xxxxxPx#P Px#xPx xxPXx xpx#xxPXx xxxPx#P xxxPSx PXxx#P xxxPXx px#Pxx P#Px#P xxpxS#px Pxx#Px xxxxPx#xpx

2815a 2815b 2816a 2816b 2817a 2817b 2818a 2818b 2819a 2819b 2820a 2820b 2821a 2821b 2822a 2822b 2823a 2823b 2824a 2824b 2825a 2825b 2826a 2826b 2827a 2827b 2828a 2828b 2829a 2829b 2830a 2830b 2831a 2831b 2832a 2832b 2833a 2833b 2834a 2834b 2835a 2835b 2836a 2836b 2837a 2837b 2838a 2838b 2839a 2839b 2840a 2840b 2841a 2841b 2842a 2842b

D* B1 C A1 A1 A1 B1 A1 E B1 A1 A3A1 D* A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 D* B3A1 A1 B1 A1 A1 C A1 A1 B1 A1 C F C A1 C A1 B1 A3 A1 A1s A1 B1 A1 A1 B1 B1 D A1 C A1 B1 F C

Px#Pxx xxxxxPx#P xxP#px pxx#Px Px#xPx PX#Px xxPx#P Px#pxx pxSxx#P xxxPX#P Px#xPx xxxxxPx#pxx px#xxxpx#Px xxxxxxPxx#Px xxxxPXx#xxPx#Px xxxxxPx#xPx px#xxxPx#Px xxxxxxxPx#xPx pxx#xPx#xPx xxxxPx#xPx Px#xxpxx#Px xxxxpxx#Px pxx#xpXSx xxxxpx#xPx px#xxxPx xxxPX#P Px#Px xxxxxxpxx#xPx xxxpxSx Px#xPx PSx#Px xxxxxPx#xP Px#xxxPx xxxxxP#px Philippus#P#px xxxxpx#Px Px#pxx xxxxpx#Px px#xxpXx xxxxxPx#P xxxxPx pxXSx P#P#px px#xxxP#x xxxPx#P Px#xPx xPx#xxxxxPx xxxxPx#P xxPxx#P pxSSx Px#Px xxxxpx#Px Px#xPx xxxxxxxPx#xpx Andreas#xxxPx xxxxpxSx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 421

Index to the scansion of the Heliand 2843a 2843b 2844a 2844b 2845a 2845b 2846a 2846b 2847a 2847b 2848a 2848b 2849a 2849b 2850a 2850b 2851a 2851b 2852a 2852b 2853a 2853b 2854a 2854b 2855a 2855b 2856a 2856b 2857a 2857b 2858a 2858b 2859a 2859b 2860a 2860b 2861a 2861b 2862a 2862b 2863a 2863b 2864a 2864b 2865a 2865b 2866a 2866b 2867a 2867b 2868a 2868b 2869a 2869b 2870a 2870b

A1 B1 A1 B1 A1 A1 A1 B1 D* B1 A1 C A1 A1 C B1 B1 C B1 B1 D* B1 A1 D D* B1 D* E A1 B1 C B1 A1 B1 D E A1 A1 D B1 B1 B1 D* B1 B1 A1 A3 C E B1 A1 B1 A1 B1 A1s B1

Pxx#Px xxxxxxPx#xP pxx#xPx xxxPx#P Px#xxxxPx xxPx#Px xP#xxxxxpxx xxxxxPx#P xPx#px#px xxxxpxx#P Px#xxPx xxxxpx#Px Px#xxxxPx PXSx xpx#Pxx xxxxxPxx#P xxxPx#P xxxxP#px xxxPx#P xP#xxP P#xPx#px xxxxxxPx#P Px#Px xxpx#Pxx Px#pXsX xxxxPx#P P#xxPxx#P xxxxxxxxPSx#P px#xxPx xxxxxPxx#P xpx#Px xxpxx#xpx Pxx#Px xxxxxPx#P px#Px#xpx xxpXSx#P P#xPx xPx#xxPx px#Px#P xxxxxPx#P xxxPx#px xxPX#P Px#xPxx xxxxPx#P xxPx#P xpx#xPx xxxPx xxpx#Px PSx#px xPx#P Px#xPx xxpxx#xP pxx#Px xxPx#px P#P#px xxxpxx#xP

2871a 2871b 2872a 2872b 2873a 2873b 2874a 2874b 2875a 2875b 2876a 2876b 2877a 2877b 2878a 2878b 2879a 2879b 2880a 2880b 2881a 2881b 2882a 2882b 2883a 2883b 2884a 2884b 2885a 2885b 2886a 2886b 2887a 2887b 2888a 2888b 2889a 2889b 2890a 2890b 2891a 2891b 2892a 2892b 2893a 2893b 2894a 2894b 2895a 2895b 2896a 2896b 2897a 2897b 2898a 2898b

B1 A1 D B1 A1 C A1 C A1 C D* B1 B1 A1 E A1 A1 A1 C A1 A2 C E* E A3 C A1 B1 A1 C D* B1 A1 B1 D* B1 D* B1 E A1 E A1 A2 A1 E B1 C A1 A1 C D* B1 C B1 A3 A1

421 xxP#xxP pxx#xPx P#PXx xP#xxP xP#xxxPx xxxPXx Px#Px xxpXsX xPx#pxx xxxxxxxxxP#px Pxx#Psx xxxxxP#xpx xxxxPXS Px#Px PSxx#px Px#xPx xxPx#Px pxx#xpxx xxPSx Px#Px Pxx#pXS xxxxxP#px xPx#P#xpx xpx#Px#xP xxxxxPx xPXx xpx#xxxpXx xPx#xP Px#Px xxxpXSx Px#xxPsx xxPx#P Px#xPx xxPx#xP P#xxPsx xxxxxxPx#xpx Px#pXsx xxxxxPx#P pXSx#P xxPXSx pXSx#px xxxxxxxxxPx#Px Px#xxxPS Px#Px pXsXx#px xxxxPx#P xxxxxP#Px xP#xxxxxPx xxxPx#Px xxP#px Pxx#Psx xxxPxx#P xxxxP#Px xxxPx#xP xxxxxPx Px#Px

MOSP-Index 1.qxd

5/11/04

422 2899a 2899b 2900a 2900b 2901a 2901b 2902a 2902b 2903a 2903b 2904a 2904b 2905a 2905b 2906a 2906b 2907a 2907b 2908a 2908b 2909a 2909b 2910a 2910b 2911a 2911b 2912a 2912b 2913a 2913b 2914a 2914b 2915a 2915b 2916a 2916b 2917a 2917b 2918a 2918b 2919a 2919b 2920a 2920b 2921a 2921b 2922a 2922b 2923a 2923b 2924a 2924b 2925a 2925b 2926a 2926b

2:49 PM

Page 422

Index to the scansion of the Heliand D* A1 D B1 A1 A1 A1 B1 A1A1 A3A1 A1A1 B1 C A1 C B1 D A1 D* D A1 D D A1 E B1 B1 E D* C E A1 A1 A1 A1 C D A1 D C B1 B1 A1 A1 A1 C C B1 A1 A1 D* A1 C B1 D* B1

xxP#xPsx xxxxPx#Pxx xP#P#px xxxxPx#xP xxxPx#Px Px#PX PX#xxPx xxxpxx#px Pxx#xxPx#Px xxxxxxPx#xpxx xpxx#xxxPx#Px xxxpx#xP xxxxxpx#PX Px#Px xxxxP#Px xxxxPx#P P#PXS Px#Px Px#P#px P#P#px Px#xxPx xPSXx P#pxx#xP Pxx#Px PSx#xP xxPx#P xxPx#px PXx#P pxx#xPSX xxP#px P#px#xpx pxx#Px P#xPx Px#pxx xpx#xxPx xxP#px px#Pxx#P Px#xPx pxSXx xP#px xxpxx#xP xxxxxPX#P xxxPx#Px Px#Px px#xPx xxxxxxP#Px xxxP#Px xxxxPx#P Px#xPx px#xxPx xxPx#Psx xxxxxxPx#PX xxpx#Px xxxxxPx#P Px#pXsX xxxxxxxxxPx#P

2927a 2927b 2928a 2928b 2929a 2929b 2930a 2930b 2931a 2931b 2932a 2932b 2933a 2933b 2934a 2934b 2935a 2935b 2936a 2936b 2937a 2937b 2938a 2938b 2939a 2939b 2940a 2940b 2941a 2941b 2942a 2942b 2943a 2943b 2944a 2944b 2945a 2945b 2946a 2946b 2947a 2947b 2948a 2948b 2949a 2949b 2950a 2950b 2951a 2951b 2952a 2952b 2953a 2953b 2954a 2954b

A1 B1 A1 C D* C B1 B1 A2 B1 C E F B1 A1 B1 A1 C B1 B1 D* B1 D* B1 A1 C A1 B1 A1 A1 D* B1 D* B1 A1 A1 E C D* C A2 B1 D* B1 A3 B1 A1 A1 A1 A1 A3 B1 A1 B1 A1 A1

Px#xxPx xxPx#px Px#Px xxxP#px xPx#xPSx xxxP#px xPx#px xxxxxxPx#P Px#xxxpxS xxxxxxxPx#xpx xxP#px PSx#px Petrus#xPx xxxPx#px pxx#Px xxxPX#P(xx) Px#xPx xxxxxxpx#Px xxxxxPx#xP xxxxpxx#P Px#xxP#px xxxPx#P pxxx#Psx xxxxPx#P Px#xxxPx xxpx#Px P#xxxPx xxPx#P P#xxPx xP#xPx xxP#xP#px xxxxxxxPx#xP xPx#P#px xxPx#xP xxP#xPx Px#xxPx P#P#xxP xxxxxxpx#pxx xP#xxxpx#Px xxxxxxP#Px P#xxxPS xxxxP#xxP Px#xxxpx#Px xxPx#P xxxxxxpxx xxxPx#P px#xxPx Px#Px xP#xxxxPx xxPxx#Px xxxxxpxx xxxxxPxx#P px#xxPx xxpxx#P xxxPx#Px xxPx#xPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 423

Index to the scansion of the Heliand 2955a 2955b 2956a 2956b 2957a 2957b 2958a 2958b 2959a 2959b 2960a 2960b 2961a 2961b 2962a 2962b 2963a 2963b 2964a 2964b 2965a 2965b 2966a 2966b 2967a 2967b 2968a 2968b 2969a 2969b 2970a 2970b 2971a 2971b 2972a 2972b 2973a 2973b 2974a 2974b 2975a 2975b 2976a 2976b 2977a 2977b 2978a 2978b 2979a 2979b 2980a 2980b 2981a 2981b 2982a 2982b

A1 B1 C B1 A1 D* A1 B1 A1 B1 A1 C A1 B1 B1 C A1 C D C B1 B1 or E A1 A1 A1 C A1 B1 A3 D A1 C B1 B1 A1 B1 B1 B1 A1 D D* D* A1 B1 A1 C A1 B1 C A1 A3 D C D E F

pxS#xPx xxxxxxxxPx#xP xxxxpx#Px xxxxxxPx#px Px#xxxxxPx xP#xxpxSx Px#xPx xxxxxPx#px P#xxPx xxxxPx#px Px#Px xxxxxP#px Px#xxxPx xxxxxPx#xP xxxPx#P xxP#px Px#xPx xxxxpx#Px pxSXx xP#px xxpxx#xP xxxPXx#P Px#xxPx Px#xxPx Px#xxxPx xxpx#Px Pxx#Px xxxxPx#px xxxxPx px#Pxx P#xxxxpXx xxxP#Px xxPXS xxxxxxxxxPx#xP pxx#xPx xxxxxxxPx#px xxpxx#xP xxxxxPX#P Px#xxxPx px#Pxx Px#P#px pxSx#P#px px#xPx xxxPx#P Px#xPx xxxpx#Px Pxx#Px xxxPx#px xxxxPsx Px#Px xxxxPx px#Pxx xxxPxx pXsXx PSx#px xxxxxxxxxPx#Iudeono

2983a 2983b 2984a 2984b 2985a 2985b 2986a 2986b 2987a 2987b 2988a 2988b 2989a 2989b 2990a 2990b 2991a 2991b 2992a 2992b 2993a 2993b 2994a 2994b 2995a 2995b 2996a 2996b 2997a 2997b 2998a 2998b 2999a 2999b 3000a 3000b 3001a 3001b 3002a 3002b 3003a 3003b 3004a 3004b 3005a 3005b 3006a 3006b 3007a 3007b 3008a 3008b 3009a 3009b 3010a 3010b

F B1 D* B1 A1A1 A3A1 F CA1 A1A1 B3A1 A1A1 A3A1 A1A1 B3A1 A1A1 A3A1 F A3A1 A3 A1 E B1 D* B1 A1 B1 D* A1 A3 A1 A1 C D B1 F A1 E B1 A1 B1 A1 A1 E A1 D* A1 F E* A1 B1 D* F A1 B1 C A1

423 Px#xxSidono#P xxxPx#xpx Px#Pxx xxxxPx#P xpx#xPx#Px xxxxpxxxSx Px#xCananeo#Px xxxxPxx#Px PXx#xxxxPx#xPx xxxxxxP#xPx pxx#xxxPxx#Px xxxxxxPx#xPx xpx#xxxPx#Px xxxxP#xPx xPx#xxxxPx#xpxx xxxxxPX#P#x Px#px#Dauides xxxxxxPx#xPx xxxxPx PXSx Psx#xpx xxxxxxPX#P Px#PSx xxxPx#P Pxx#Px xxxxxPx#P Px#xxPXx Pxx#Px xxxxPx xxxxPx#Px Px#xxPx xxxxxP#px px#Pxx xxxxxPx#P PX#xxIsraheles pxx#Px Psx#xpx xxPx#px Px#xxxPx xxPxx#P xPx#xxpxx xPx#Px PXx#P xpx#xxPx Px#xPx#px xxxxxPxx#Px Israhelo#Psx PxSxx#P pxx#xxPx xxxxPx#px Px#pxSx Agaleto#P xP#xxxPx xxxPX#P xxPsx Px#Px

MOSP-Index 1.qxd

5/11/04

424 3011a 3011b 3012a 3012b 3013a 3013b 3014a 3014b 3015a 3015b 3016a 3016b 3017a 3017b 3018a 3018b 3019a 3019b 3020a 3020b 3021a 3021b 3022a 3022b 3023a 3023b 3024a 3024b 3025a 3025b 3026a 3026b 3027a 3027b 3028a 3028b 3029a 3029b 3030a 3030b 3031a 3031b 3032a 3032b 3033a 3033b 3034a 3034b 3035a 3035b 3036a 3036b 3037a 3037b 3038a 3038b

2:49 PM

Page 424

Index to the scansion of the Heliand B1 A1 A1 B1 A1 B1 A1 D* A3 A1 A1 A1 D* B1 A1 B1 E C A1 A1 A1 B1 A1 B1 A1 B1 D* A1 A1 C A1 B1 B1 B1 A3 B1 D* D B1 A1 A1 E C A1 E B1 A1 C B1 F E* A3A1 A1A1 A3A1 A1A1 A3A1

xxxxPx#P Px#Px Px#xPx xxxxPx#xpx Px#xxPx xx(xx)Px#P pxx#xPx P#xxPXx xxxPx Px#xPx Px#xxxPx xxxPx#Px Px#pxSx xxxxxPx#xP P#xxPX(xx) xxxxxPx#px PSx#px xxxxpx#Px xxxxPx#Px Px#Px Pxx#Px xxxxxPx#px xPx#xxPx xxxxxpxS#px Px#xPx xxxxxxxPx#P px#xxP#px Px#Px px#xxxPx xxP#px xxxPx#Px xxxxPx#P xxxxPx#P xxPx#xP xxxxxPx xxxPxx#xP Px#PSx xP#pxxx xxxxPx#P Px#Px Px#xxxPx PXx#P xxxxPxx PXx#Px Psx#xpx xxxxxPX#P Px#xPx xxxxP#Px xxPx#P xxxPx#Iudeono xPXSx#xpx xxxxxxxxPx#Px xPxx#xxxxxxxxPx#xpxx xxxxxPx#xpxx px#xxPx#Px xxxxx(xx)xPx#Px

3039a 3039b 3040a 3040b 3041a 3041b 3042a 3042b 3043a 3043b 3044a 3044b 3045a 3045b 3046a 3046b 3047a 3047b 3048a 3048b 3049a 3049b 3050a 3050b 3051a 3051b 3052a 3052b 3053a 3053b 3054a 3054b 3055a 3055b 3056a 3056b 3057a 3057b 3058a 3058b 3059a 3059b 3060a 3060b 3061a 3061b 3062a

A1 F D* B1 C B1 A1 F D* F D C D* F D* B1 A1 B1 B1 A1 C B1 A1 C A1 B1 B1 A1 D* B1 F B1 A1 B1 A1 B1 A3 E E B1 D C A1 C E B1 F

3062b 3063a 3063b 3064a 3064b 3065a 3065b 3066a

A3A1 B1A1 A3A1 A1 A3A1 A1A1 B3A1 A1A1

Pxx#Px xxxxxIudeo#Px Px#pXSx xxPx#P xxxxxxPSx xP#xpx Pxx#Px xxIudeono#P Px#PSx xxxxxElias#P P#Psx xxxP#Px P#xxxxxPSx xxxxxIohannes#P PS#Pxx#px xxPx#P px#xpxx xxxxPx#px xxPxx#P pxxx#Px xxPsxx xxxPx#P Px#xxPx xxxxxxxxP#px xPXx#xxxpxx xxxPX#P xxxP#xxP(xx) Pxx#Px Px#Psx xxP#xP Simon#Petrus xP#xpx Px#xxPx xxxxPx#P Px#xPx xxPxx#P xxxPx PXx#px PXx#px xxP#xP P#pX#Px xxxxpx#Px Pxx#Px xxxP#Px PXxx#P xxxxxxPx#xpx Px#xxSimon#(xx) px#Ionases xxxxxPx#xPx xPx#xxxPxSx xxxxxxPx#Px Px#xPx xxxxxPX#Px px#xxxpxx#Px xxxP#xPx xPx#xPx#Px

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 425

Index to the scansion of the Heliand 3066b 3067a 3067b 3068a 3068b 3069a 3069b 3070a 3070b 3071a 3071b 3072a 3072b 3073a 3073b 3074a 3074b 3075a 3075b 3076a 3076b 3077a 3077b 3078a 3078b 3079a 3079b 3080a 3080b 3081a 3081b 3082a 3082b 3083a 3083b 3084a 3084b 3085a 3085b 3086a 3086b 3087a 3087b 3088a 3088b 3089a 3089b 3090a 3090b 3091a 3091b 3092a 3092b 3093a 3093b 3094a

EA1 A1A1 A1A1 B1A1 B1 F C D* B1 or E A1 A3A1 A1A1 CA1 B1 A1 A1 E A1 B1 A3 A1 A1 B1 E A1 E B1 A1 C D C A2 C A1 B1 A1 B1 D F D A1 A1 B1 F A1 A1 A1 A1 B1 B1 A1 B1 B1 A1 F C

PSx#xxxP#xPx Px#xxxxxxPx#xPx pxSx#xxxPx#xPx xP#xxxpx#xPx xxxxxpxx#P Px#Peter xxxxxxxxxxpx#Px Px#P#px xxxPXx#P Px#Px xxxxxxxPx#Px xPx#Px#Px xxxxxpXSx#pxx xxxPx#P Px#xPx Pxx#Px px#Px#xP pxx#Px xxxPx#xP xxxxxPx PxSx xPx#Px xxxPx#xP pXSx#xpx xxPx#xxxpxx PXx#P xxxxxxxPx#px xPx#xPx xxxpXSx xpx#Px#P xxP#Px Px#px#P xxxxxxxpx#Px Px#xxxPx xxxxxxxxPx#P Px#xxxpxx xxxPx#P px#Px#P xxxIudeon#P PSXx Px#Px Px#xxPx xxPx#px xxHierusalem Px#Px Px#xxPx Px#Px xPx#xPx xxxxxPx#P xxxPxx#P xPx#xPx xPxx#px xxPx#P Px#xPx Simon#Petrus xxxpx#Px

3094b 3095a 3095b 3096a 3096b 3097a 3097b 3098a 3098b 3099a 3099b 3100a 3100b 3101a 3101b 3102a 3102b 3103a 3103b 3104a 3104b 3105a 3105b 3106a 3106b 3107a 3107b 3108a 3108b 3109a 3109b 3110a 3110b 3111a 3111b 3112a 3112b 3113a 3113b 3114a 3114b 3115a 3115b 3116a 3116b 3117a 3117b 3118a 3118b 3119a 3119b 3120a 3120b 3121a 3121b 3122a

B1 A1 B1 A1 B1 B1 B1 A1 B1 D* B1 B1 A1 A1 B1 E C B1 C A1 A1 A1 B1 A2 B1 E B1 D F F C A1 B1 A1 D* B1 A1 A1 B1 A1 B1 B1 D D* B1 D C A1 B1 D* C C A1 B1 A1 A1

425 xxxxPx#P P#xPx xxxPx#P(xx) PX#Px xxxxxPx#px xpx#xxxxxP xxP#xP Px#Px xxxxxxPx#xpx Px#Px#P xxxxxPx#P xxxpXS#P(xx) Px#Px Px#Px xxxxxPx#P PXx#px xxxxP#px xxxPx#P xxxpx#Px Px#Px xxxxxxxpxx#Px xPx#Px xxxxPx#P Px#xpXS xxpxx#P px#Px#px xxxPxxx#P P#Px#P Simon#Petrus Iacob#xxIohannes xpx#Px Px#xxPx xxxxxxxxxP#xP Px#xxxPx Px#P#px xxPx#P Px#Px PX#xxxpXx xxxxPx#px Px#Px xxxPxx#xP xxPx#P px#Pxx Px#pXsX xxxPx#P px#P#xxP xxxxxxpx#Px px#xxPx xPX#P pXx#PxX xpx#Px xxxPSx Pxx#Px xxPx#P Px#Px PS#pxx

MOSP-Index 1.qxd

5/11/04

426 3122b 3123a 3123b 3124a 3124b 3125a 3125b 3126a 3126b 3127a 3127b 3128a 3128b 3129a 3129b 3130a 3130b 3131a 3131b 3132a 3132b 3133a 3133b 3134a 3134b 3135a 3135b 3136a 3136b 3137a 3137b 3138a 3138b 3139a 3139b 3140a 3140b 3141a 3141b 3142a 3142b 3143a 3143b 3144a 3144b 3145a 3145b 3146a 3146b 3147a 3147b 3148a 3148b 3149a 3149b 3150a

2:49 PM

Page 426

Index to the scansion of the Heliand B1 A3 D* A1 A1 A1A1 C D* A1 D*A1 B1 A1 B1 D* F A1 C A1 A1 E C B1 A1 A1 B1 C B1 E F D B1 D* A1 B1 D A1 A1 E F F B1 D* B1 D* E A1 B1 E B1 E* B1 A1 C A1 B1 C

xxxxxpx#xP xxxxxPx PXSxx px#xxxPx xxxxxPx#Px PXx#xxPx#Px xxxP#px Px#xPsx Px#Px Pxx#xxxPXx#Px xxxPx#xP xP#xpXx xxxPS#P xPx#Px#P Elias#xxMoyses Px#xxPx xxPXx Px#Px xxxPS#Px xP#P#xxpx xxpx#px xxPx#P Px#Px xPx#xxxxxPx xxPx#P xxP#PS xxPx#P pxXx#xP Petrus#xxpXx px#PSx xxxxPx#P Px#xxpx#px P#xxxxpXx xxxxPx#px P#pxSx xxxP#xxxxPx xP#xPx PSx#xpx xxMoysese#Px xxEliase#Px xxPx#P pxx#PsX xxxxxP#xP xxP#xP#xP P#Px#P PXx#Px xxxPx#P pxSx#xP xxxxPx#P Px#Px#px xxxpxx#P Px#Px xxxpx#Px PXxx#PX xxxxPx#P xxxpxSx

3150b 3151a 3151b 3152a 3152b 3153a 3153b 3154a 3154b 3155a 3155b 3156a 3156b 3157a 3157b 3158a 3158b 3159a 3159b 3160a 3160b 3161a 3161b 3162a 3162b 3163a 3163b 3164a 3164b 3165a 3165b 3166a 3166b 3167a 3167b 3168a 3168b 3169a 3169b 3170a 3170b 3171a 3171b 3172a 3172b 3173a 3173b 3174a 3174b 3175a 3175b 3176a 3176b 3177a 3177b 3178a

B1 A1 A1 B1 C B1 A1 D* A1 A1 B1 A1 E A3 A1 C C A1 B1 A1 B1 A1 C A1 B1 D B1 B1 B1 A1 F A1 B1 C A1 A1 B1 D* A1 A1 B1 F C D* B1 A1 C A1 E C A1 D* F A1 B1 A1

xxxPx#px xPx#xxPx xxxxxPxx#Px xPx#px xxP#px xxpxx#P xP#xPx xxxPx#xPSx Px#xPx Pxx#Px xxxxxPx#P xP#xxxPx PXxx#P xxxxxxPx xxxxPx#PX xxxPSx xpx#Px Pxx#Px xxxPx#P px#xxxPx xxxPx#P xpx#xxxPx xxxxP#px Px#Px xxPx#P xpx#pxx#P xxxxxxPxx#P xxxPx#px xP#xxP Pxx#Px xxxxIudeono#P xPx#xxPx xxxPx#P xxPSx xPx#xPx xPx#xxxPx xxxxxxPx#P Px#xxPXS pxx#Px Px#xxxxxpXx xxxxxPX#P xxGalileo#P xxxpxXx Px#xPx#P xxxpxx#P Pxx#pXx xxxP#px xxPxx#xPx PS#xxP xxxpXSx Px#Px xxPx#Pxx xxxxxxIudeono#P Px#xPx xxxxPx#P Px#xPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 427

Index to the scansion of the Heliand 3178b 3179a 3179b 3180a 3180b 3181a 3181b 3182a 3182b 3183a 3183b 3184a 3184b 3185a 3185b 3186a 3186b 3187a 3187b 3188a 3188b 3189a 3189b 3190a 3190b 3191a 3191b 3192a 3192b 3193a 3193b 3194a 3194b 3195a 3195b 3196a 3196b 3197a 3197b 3198a 3198b 3199a 3199b 3200a 3200b 3201a 3201b 3202a 3202b 3203a 3203b 3204a 3204b 3205a 3205b 3206a

C A1 B1 E A1 A3 A1 A1 B1 F F F B1 A1 B1 D B1 F B1 A3 A1 A1 C A1 C A1 B1 A1 A1 A3 B1 A3 C D B1 F B1 A1 B1 D C C C A1 B1 F C A1 B1 A1 B1 A1 B1 A1 B1 A1

xxP#px Px#xxxxPx xxxxxPx#P PXx#px Px#Px xxxxxxPx pxx#Px PXx#xxxxpxx xxxxxPX#P px#xGalilea xxxxIudeono#P Px#xxCafarnaum xxxxxxpXx#px Px#xxxxpxx xxxxxxPx#px pXSxx xxxPx#P Simon#Petrusen xxxxxxPx#px xxxxpxx Px#xpxx xxPXSx xxxxxpx#Px Px#Px xxpx#Px pxx#xPxx xxxxxPx#P Px#Px xxxxxPx#Px xxxxPx xxxPx#P xxxPx xxxxxxxpx#Px pXSxx xxPx#P Simon#Petrus xxxPx#P Px#Pxx xxxxxpx#xP xpx#PX#P xxxxxP#Px xpx#Px xxxpxSx Px#xpxx xxxxxPx#px Simon#Petrus xxxP#Px Px#Px xxxxxxPX#px P#xPx(xx) xxxxxxxxPx#xP xPx#xxxPx xxxxxxPx#px Pxx#Px xxxPx#P xxPx#xxPx

3206b 3207a 3207b 3208a 3208b 3209a 3209b 3210a 3210b 3211a 3211b 3212a 3212b 3213a 3213b 3214a 3214b 3215a 3215b 3216a 3216b 3217a 3217b 3218a 3218b 3219a 3219b 3220a 3220b 3221a 3221b 3222a 3222b 3223a 3223b 3224a 3224b 3225a 3225b 3226a 3226b 3227a 3227b 3228a 3228b 3229a 3229b 3230a 3230b 3231a 3231b 3232a 3232b 3233a 3233b 3234a

A1 A1 C B1 A1 A1 B1 F C A1 B1 A1 B1 A1 B1 A1 C A1 C A1 B1 C C A1 B1 A1 A1 A1 B1 D* C A1 C A1 B1 D* B1 A1 B1 A1 C A1 B1 A3 D* A1 B1 D* B1 A1 B1 A1 B1 A1 C C

427 Px#xxPx Px#xpxx xxxP#Px xxxxxxxPx#P Pxx#Px Px#xPx xPxx#P Simon#Petrus xxxxP#Px Px#xPx xxP#xP P#xPx xxPx#P xP#xxxPx xxxxxPx#P Pxx#Px xxxxxxxpx#px Px#xPx xxxPXx pXS#xPx xxxxxPx#xpx xxPXx xpXSx Px#xxPx xxxxpxx#P Px#Px xxxxxPx#PX xxpx#xxxxPx xxxxxPx#xxpx Px#xxPSx xxxxPsx Px#xPx xxxxpXSx Px#Px xxxxPxx#P xxPx#Pxx xxxpxx#xP(xx) Px#xPx xxxxxPx#xP xxP#xPx xxxxxP#px Px#xxxPx xxxxxP#xP xxxxPx px#xxPxx#P Pxx#pxx xxxxxxPx#P P#xxPSx xxxxxPx#xxP PS#xPx xxxxxPx#P Px#xxxpxx xxxPx#px px#xxPx xxxxxxxxxP#px xxpx#Px

MOSP-Index 1.qxd

5/11/04

428 3234b 3235a 3235b 3236a 3236b 3237a 3237b 3238a 3238b 3239a 3239b 3240a 3240b 3241a 3241b 3242a 3242b 3243a 3243b 3244a 3244b 3245a 3245b 3246a 3246b 3247a 3247b 3248a 3248b 3249a 3249b 3250a 3250b 3251a 3251b 3252a 3252b 3253a 3253b 3254a 3254b 3255a 3255b 3256a 3256b 3257a 3257b 3258a 3258b 3259a 3259b 3260a 3260b 3261a 3261b 3262a

2:49 PM

Page 428

Index to the scansion of the Heliand B1 D* B1 A1 B1 A1 C D* C A1 B1 D A1 A1A1 F B1 A1 A3 C E A1 C A1 A1 A1 A1 B1 D* A1 A1 B1 A1 B1 A1 B1 A1 B1 D* B1 C A1 D* B1 A1 B1 or E E F A1 B1 C C C A1 A1s B1 A1

xxxxxxpxx#px Px#PxS xxxxxpxx#P Px#xPx xxxxPx#xpx xxPx#xxxpxx xxxxxP#px px#xxxxPxx xxxxxxxxxxpx#Px P#xxxxxPx xxxxxxPx#P P#pXsX Px#xPx px#xxxpxx#Px xPxx#Petrus xxxPx#P Px#Px xxxxxPx xxP#Px PS#xP Px#Px xxxpx#Px xxPx#xPx Pxx#Px xxxxxxxPx#Px Px#xPx xxxxPx#P xpx#xpx#px Pxx#Px xPx#xxxpXx xxPx#px Px#xxxPx xxxPx#xP px#Px#pXx xxxPx#xpx Px#xPx xxxxxxPx#px Px#PSx xxxxxxxP#xP xxxxPXx PX#Px pxx#PSx xxxxPx#px Px#Px xxxxxPXx#P xP#P#xpx xxxxxIesu#P Px#xPx(xx) xxxpxx#P xxxxpXSx xpx#Px xxxxPsx Px#xPx PXS#px xxPx#px Px#xxPx

3262b 3263a 3263b 3264a 3264b 3265a 3265b 3266a 3266b 3267a 3267b 3268a 3268b 3269a 3269b 3270a 3270b 3271a 3271b 3272a 3272b 3273a 3273b 3274a 3274b 3275a 3275b 3276a 3276b 3277a 3277b 3278a 3278b 3279a 3279b 3280a 3280b 3281a 3281b 3282a 3282b 3283a 3283b 3284a 3284b 3285a 3285b 3286a 3286b 3287a 3287b 3288a 3288b 3289a 3289b 3290a

C A3 A1 A1 B1 A1 B1 C A1 A3 A1 A3 A1 B1 B1 E D* A1 B1 B1 B1 D B1 A1 B1 A1 B1 A1 B1 D* B1 A3 B1 A1 B1 C C A1 A1 A1 B1 C A1 A1 C C A1 A1 B1 A1 B1 A1 B1 A1 C B1

xxxxxP#px xxxxxxPx xxpxx#Px xPx#xPx xxP#xP pX#xxPx xxxPx#px xxxP#px Px#Px xxxxPx xPxx#Px xxxxxPx Px#xPx xxP#xP xxPx#xpx xpXSx#xP xxPx#xPsx P#xxpxx xxxxP#xpx xPx#xpx xPS#xpx pX#Px#xP xxxPxx#P px#xxPx xxxxPx#P xPXx#xPx xxxxPx#P pxx#Px xxxxPx#px xPx#px#Px xxxxPx#P xxxxxxPx(xx) xxxPx#P Px#Px xxxxP#xxP xxxxPXx xxxxxP#px P#xxPx P#xxP#x(xx) P#xxPx xxxPx#px xxPsx Px#Px Px#Pxx xxxxxP#px xxxPsx Px#xPx Px#Px xxPx#P Px#Px xxxxPx#P P#xpxx xxxxPx#xP Px#xxxxPx xxxxpx#Px xxxPx#P

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 429

Index to the scansion of the Heliand 3290b 3291a 3291b 3292a 3292b 3293a 3293b 3294a 3294b 3295a 3295b 3296a 3296b 3297a 3297b 3298a 3298b 3299a 3299b 3300a 3300b 3301a 3301b 3302a 3302b 3303a 3303b 3304a 3304b 3305a 3305b 3306a 3306b 3307a 3307b 3308a 3308b 3309a 3309b 3310a 3310b 3311a 3311b 3312a 3312b 3313a 3313b 3314a 3314b 3315a 3315b 3316a 3316b 3317a 3317b 3318a

A1 A1 C A1 B1 A1 D* C A1 C D* D A1 D* C D*A1 A1 D* B1 B1 C A1 A1 A1 C A2 A1 A1 C D E F B1 A1 B1 A1 B1 A1 A1 D* B1 A1 C A1 A1 A1 B1 B1 B1 B1 A1 A1 B1 A1 C F

PSxx#Px Px#xPx xxxP#px P#xxPx xxPx#px pxx#xPx Px#xxP#px xxxPSx Px#Px xxpx#Px P#xxPx#P P#pxSx P#xxxxPx xxxPxx#pXSx xxxxpx#Px PSx#Pxxx#Px P#xpXx Px#xxPXx xxxPS#P xPx#P xxxpx#Px Px#xPx xxxxxxPx#xpxx xPxx#Px xxP#px xPx#xxxpXS Px#Px PxSx xxxxxxxxP#px xxPSxx PSx#px Simon#Petrus xxPx#P Px#Px xxxxxxPx#px(xx) Px#xPx xxxxPXS Px#xxPx P#xPx px#xxPXx xxxxPx#xpx Pxx#xxxPx xxxxxpx#Px Px#xPx Px#Px Px#xxPx xxPx#px(xx) xxpxx#P xxxPx#px xxPx#P PXSx Px#xPx xxxxxxxxPxx#P Px#Px xxxxxxpx#Px xxxIsrahelo

3318b 3319a 3319b 3320a 3320b 3321a 3321b 3322a 3322b 3323a 3323b 3324a 3324b 3325a 3325b 3326a 3326b 3327a 3327b 3328a 3328b 3329a 3329b 3330a 3330b 3331a 3331b 3332a 3332b 3333a 3333b 3334a 3334b 3335a 3335b 3336a 3336b 3337a 3337b 3338a 3338b 3339a 3339b 3340a 3340b 3341a 3341b 3342a 3342b 3343a 3343b 3344a 3344b 3345a 3345b 3346a

A1 A1 B1 A3 B1 A1 A1 A1 C A2 B1 B1 B1 D B1 B1 A1 B1 C A1 B1 A1 B1 A1 C A1 C A1 B1 A1 B1 A1 B1 or E D* F A1 C B1 A1 C A1 A1 B1 A1 C A3 B1 C B1 A1 C D* B1 D* C A1

429 pxxSx xPx#xxxxPx xxxxxxPxx#px xxxxxxPx xxxxxxxxpXx#xP xxxPx#Px Px#xPx P#xPx xxxxP#px px#Px#pXS xxxxPx#P xPx#px xxxxxxxpxx#P px#Px#P xxxxxPx#P xxxPx#P xpxx#Px xxxxPx#P xPsx Px#xxxxpxx xxxpxx#xP Px#xPx xxxxPx#P px#xPx xxxpxSx pxx#pxx xxxxxpx#Px Px#xxPx xxxxxPx#P xxxpx#xpxx xxxxPS#P Px#xxPx xxxxxPXx#P xPx#xxPsx Lazarus#xxPx xxxpx#xpxx xxpx#px xxxxPxx#P Px#Px xxPsx Px#Px Px#xPx xxxPx#P xPx#xPx xxxxP#px xxxxxxPx xxxxxPx#px xpx#Px xxxxxPx#px xP#xxxxPx xxxxxxxxpx#Px xxxPXx#xxPx#xP xxxxxxxxPx#P Pxx#xPSx xxPXx Px#pxx

MOSP-Index 1.qxd

5/11/04

430 3346b 3347a 3347b 3348a 3348b 3349a 3349b 3350a 3350b 3351a 3351b 3352a 3352b 3353a 3353b 3354a 3354b 3355a 3355b 3356a 3356b 3357a 3357b 3358a 3358b 3359a 3359b 3360a 3360b 3361a 3361b 3362a 3362b 3363a 3363b 3364a 3364b 3365a 3365b 3366a 3366b 3367a 3367b 3368a 3368b 3369a 3369b 3370a 3370b 3371a 3371b 3372a 3372b 3373a 3373b 3374a

2:49 PM

Page 430

Index to the scansion of the Heliand B1 A1 B1 B1 C D* B1 C D B1 C F A1 A1 B1 A1 B1 B1 A1 B1 A1 A1 B1 C A1 D* A1 F B1 A1 F A1 E D* B1 A1 D F B1 C A1 A1 F A3 C A1 C A1 B1 A3 A1 A1 B1 D A1 A1

xxxxxxPx#P xxxPx#Px xxxxxxxxpxxxsx xxPx#P xPsx xpxx#Px#P xxPx#P xpx#Px px#Pxx xPx#xP xxxxxxP#px xxxAbrahames#P xPx#Px Px#xPx xxxxPx#P px#xPx xxxxPxxsx xxPxx#P PXSx xxxP#xP Px#Px xPxx#xPx xxxPx#P xxP#Px PXx#xPx xPx#xxxpxx#P xxxxxxxPx#Px Abraham#xpx xxPx#P Px#xPx xxLazarus#P Px#xxPx P#P#xP Pxx#xPSx xxxxPxx#P Px#xxxPx P#P#px px#Abraham(xx) xxpxx#P xxxxxxxPsx Px#Px Px#xxxxPx xxxLazarus#px xxxxPx xxP#Px Px#pxx xxP#Px Px#xxxxPx xxxxxxPxx#P xxxxPx xxPxx#Px Px#Px xxPx#px px#Pxx PXSx Pxx#Px

3374b 3375a 3375b 3376a 3376b 3377a 3377b 3378a 3378b 3379a 3379b 3380a 3380b 3381a 3381b 3382a 3382b 3383a 3383b 3384a 3384b 3385a 3385b 3386a 3386b 3387a 3387b 3388a 3388b 3389a 3389b 3390a 3390b 3391a 3391b 3392a 3392b 3393a 3393b 3394a 3394b 3395a 3395b 3396a 3396b 3397a 3397b 3398a 3398b 3399a 3399b 3400a 3400b 3401a 3401b 3402a

C F C A1 B1 A1 B1 A1 B1 A1 A1 F B1 A1 C A1 B1 E B1 A1 B1 A1 C A1 B1 F A1 B1 B1 F B1 A1 B1 A1 C A1s B1 A1 B1 A3 B1 B1 B1 F C C A1 F B1 E B1 A3 C C B1 A3

xxxxxxP#px xxPSxx#xAbraham xxPsx xpx#xxxxxPx(xx) xxPx#P pxx#xpXx xxxxxxxPx#xpx Px#xPx xxxxpxx#P Px#Px Px#pxx Lazarus#xxxPx xxxPx#px Px#xpXx xxxxxpx#Px Px#xPx xxxPx#px PXx#P xxxxxxPx#px pxx#xPx xxxxPxx#P xxPx#xxPx xxxpx#Px Px#xxPx xxxxPx#xxP xxxAbrahame xP#xPx xxxPx#P xxPxx#P xxLazarus xPx#P Px#Px xxxPx#P Pxx#Px xxxPXx PS#px xxxxxxPx#P xPx#xxxxPx xxPx#P xxxxxPx xxxxxxxPx#xP xxPx#P xxxxxPx#P Abraham#Psx xxxxP#px xxxPsx Px#Px Moyseses#xP xxxpxxx#P Psxx#P xxxPxx#P xxxxPx xxxxxxxP#Px xxP#px xxxPx#P xxxPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 431

Index to the scansion of the Heliand 3402b 3403a 3403b 3404a 3404b 3405a 3405b 3406a 3406b 3407a 3407b 3408a 3408b 3409a 3409b 3410a 3410b 3411a 3411b 3412a 3412b 3413a 3413b 3414a 3414b 3415a 3415b 3416a 3416b 3417a 3417b 3418a 3418b 3419a 3419b 3420a 3420b 3421a 3421b 3422a 3422b 3423a 3423b 3424a 3424b 3425a 3425b 3426a 3426b 3427a 3427b 3428a 3428b 3429a 3429b 3430a

C A1 B1 B1 B1 A1 C A1 A1 C B1 B1 A1 A1 A1 B1 A1 D A1 B1 A1 A1 B1 D* B1 B1 A1 E A1 D* B1 A1 B1 B1 A1 A3 B1 E B1 B1 or E B1 A1 B1 C A1 D* A1 D* B1 C A1 A1 B1 A1 A1 A1

xxP#px xPx#xPx xxxxxxxPx#P xxxPx#xP xxxpx#xP P#xPx xxxxxxxPsx Px#Px xxxPxx#Px xxPXx xxxxxxxxxpXx#P xxxpx#xxP Px#Px xPx#xxxPx Px#Px xxxPx#P xxpxx#Px px#PSx Px#Px xxxxxPx#px Px#xPx P#xPx xxxPx#xP xPXx#xPXx xxPS#P xxxxxPxx#xP Px#Px pXXx#P xPxx#pxx px#xxPSx xxxxP#xpx Px#xPx xxxxxPX#P xxxxPx#P P#xxPx xxxxPx xxxpxx#P pXSx#px xxxPx#P xxPXx#P xxxPX#P Px#xPx xxPx#xP xPSx Px#Px xP#xxPx#xP xPx#xPx xPx#PXS xxxPX#px xxxP#Px Px#pxx px#xxPx xxxxPx#xP xP#xPx xxPx#xPx Px#xPx

3430b 3431a 3431b 3432a 3432b 3433a 3433b 3434a 3434b 3435a 3435b 3436a 3436b 3437a 3437b 3438a 3438b 3439a 3439b 3440a 3440b 3441a 3441b 3442a 3442b 3443a 3443b 3444a 3444b 3445a 3445b 3446a 3446b 3447a 3447b 3448a 3448b 3449a 3449b 3450a 3450b 3451a 3451b 3452a 3452b 3453a 3453b 3454a 3454b 3455a 3455b 3456a 3456b 3457a 3457b 3458a

B1 A1 A1 D* C C B1 A1 A1 A1 B1 A3 B1 A1 B1 A1 B1 B1 A1 A1 C D* B1 A1 C B1 A1 B1 B1 or E D* B1 E A1 C A1 A1 B1 E B1 A2 B1 C B1 A1 A1 A1 C B1 A1 ? A1 A1 B1 A1 B1 A1

431 xxxxxPX#P PXx#xxPx Px#xPx xP#xPx#P xpx#Px xxxpxXx xxxPx#P xPx#xPx P#xxPx Px#xPx xxxxxPX#P xxxxxpxx(xxx) xxxxxxpx#xpx pxXSx xxPS#P PXx#Px xxxxxxPx#xP xxxPx#P xxPx#Px Px#xxxPx xxxxxP#px xPXx#xPXx xxxxxxxxPx#xP Px#xxxPx xxxP#Px(xx) xxxxxxPx#P P#xPx xxPx#P xPXx#P Px#xxPx#P xxxxxPx#P xxPSx#P Px#Px xxPSx Px#Px px#xxPx xxxxxxpXx#P PSx#P xxPx#P px#xpx#P xxxxxxpxx#P xxPXx xxxxpxx#P Px#Px pXsx#Px xPx#xPx xxxxxPsx xxP#xpx Pxx#Px px#P pxx#xPx Pxx#Px xxxxxpXx#P Px#xxxxPx xxxxPx#px Px#PX

MOSP-Index 1.qxd

5/11/04

432 3458b 3459a 3459b 3460a 3460b 3461a 3461b 3462a 3462b 3463a 3463b 3464a 3464b 3465a 3465b 3466a 3466b 3467a 3467b 3468a 3468b 3469a 3469b 3470a 3470b 3471a 3471b 3472a 3472b 3473a 3473b 3474a 3474b 3475a 3475b 3476a 3476b 3477a 3477b 3478a 3478b 3479a 3479b 3480a 3480b 3481a 3481b 3482a 3482b 3483a 3483b 3484a 3484b 3485a 3485b 3486a

2:49 PM

Page 432

Index to the scansion of the Heliand C C A1 A1 C A3 C A1 A1 A1 B1 A1 B1 A3 A1 E B1 A1 B1 A1 C A1 B1 C C A3 A1s A1 B1 A1 B1 A1 B1 C A1 D* A1 A1 B1 A2 A1 D* B1 A1 B1 A1 C D* A1 E C A3 B1 A1 B1 A1

xxxxxPsx xxxxxpxxx P#xPx xPx#xPx xpx#Px xxxxpXx xxxPSx Px#xPx PXSx P#xPx xxxxpxx#P Px#xPX xxxxPX#P xxxxPx xpxXSx xPSx#xpx xxpxx#px xPxx#Px xxxPS#P Px#xxpxx xxxPsx Px#xPx xxPS#px xxxPXx xpx#Px xxxxxxpxx px#P#px Px#xxPx xxxxpxx#P Px#xxPx xxxxpX#xP xPx#xxxPx xxxPx#P xpx#Px Pxx#Px xP#xPsx P#xPx Px#Px xxxxPx#P xPx#xxxpx#P Pxx#Px Px#pxSx xxxxPx#P xxPx#Px xxxxPx#xpx xxxPx#xxPx xxxxxxP#px px#xP#xP xPx#xPx P#P#xpx xxxxxxpx#pxx xxxxxxPx xxxPx#P xPxx#xPx xxxxxxpxx#P Px#xxxPx

3486b 3487a 3487b 3488a 3488b 3489a 3489b 3490a 3490b 3491a 3491b 3492a 3492b 3493a 3493b 3494a 3494b 3495a 3495b 3496a 3496b 3497a 3497b 3498a 3498b 3499a 3499b 3500a 3500b 3501a 3501b 3502a 3502b 3503a 3503b 3504a 3504b 3505a 3505b 3506a 3506b 3507a 3507b 3508a 3508b 3509a 3509b 3510a 3510b 3511a 3511b 3512a 3512b 3513a 3513b 3514a

B1 A1 B1 A1 C D C B1 B1 B1 B1 C A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 B3A1 A1A1 A3A1 A1A1 CA1 A1A1 B3A1 A1A1 A3A1 A1A1 A3A1 D* A3A1 A1A1 A3A1 A1A1 CA1 A1 A3A1 D A1 A1 B1 A1 B1 C A1 D B1 A1 B1 C

xxxPx#px xpx#xxPx xxxPxx#px xxPx#xxPx xxP#px P#pXSx xxpx#Px xxxxPx#xpx xxPx#P xxxPx#px xxpXx#P xxxPSx Px#Px P#xxxxPx#xPx xxxxxxPx#Px xxPx#xxpxx#xpxx xxxxxPX#Px xpX#xxxPx#xPx xxxxxxxPx#xPx xPx#xxxxPx#xPx xxxxxPx#xpxx Px#xxxxxxxxpXx#Px xxxxxxPx#xPx xPx#xxxPx#xpxx xxxxxxxpx#xpxx xxP#xPx#Px xxxxxxPx#Px Px#xxxPx#Px xxxxPxx#Px PXx#xxxxPx#Px xxxxxxxxP#xPx xPXx#xxxPx#xPx xxxxxxxxPXSx xPx#Px#Px xxxxxPX#Px Px#pXSx xxxxxPx#Px Px#xxxxPx#xPx xxxxxPxx#Px xpx#xPx#Px xxxxxxPxx#Px xPx#xxxPx xxxxxxxxxPx#xPx P#pXSx xxxPx#Px Px#xPx xxxPx#P Px#xPx xxxpxx#P xxxxPSx Px#Px P#PSx xxxPx#xP Px#xxPx xxxpxx#P xpx#Pxx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 433

Index to the scansion of the Heliand 3514b 3515a 3515b 3516a 3516b 3517a 3517b 3518a 3518b 3519a 3519b 3520a 3520b 3521a 3521b 3522a 3522b 3523a 3523b 3524a 3524b 3525a 3525b 3526a 3526b 3527a 3527b 3528a 3528b 3529a 3529b 3530a 3530b 3531a 3531b 3532a 3532b 3533a 3533b 3534a 3534b 3535a 3535b 3536a 3536b 3537a 3537b 3538a 3538b 3539a 3539b 3540a 3540b 3541a 3541b 3542a

A1 B1 C A3 A1 A1 C A1 A1 A1 C A1 C F F A1 B1 A1 C E* A1 A3 B1 A1 D* A1 A1 A1 C A1 B1 A1 A1 A1 B1 B1 A1 B1 B1 B1 A1 B1 B1 D* C D* B1 A1 B1 A1 B1 E A1 D* B1 A1

P#xPx xxPS#P xxxxxpx#Px xxxxxxPx Pxx#Px pxx#Px xxxxxPXx P#xxPx xxxPx#Px PXSx xxxxxPxx PS#Px xxxxxpx#Px(xx) xxxxxHierusalem xxIudeono#P Px#Px xxxxxPx#P xpx#xxxxPx xxxPsx Px#P#xP Px#xPx xxxxxPx xxxPxx#P pxx#xxxPx xxxxxPx#xPxx Px#xxxxxPx px#xxxxpxx Px#xPx xxPsx PXSx xxpXS#px xPx#xxPx Px#Px xPx#xPx xxxxxPx#P xPxx#P xPx#xPx xPx#px xxxxxxxPx#px xxxxPxS Px#Px xxPxx#xP xxxxxxPx#P Px#xPsx xxxxxxpx#Px Px#xxPSx xxxxxxxPx#px Px#Px xxxxPx#P Px#xxxPx xxPx#px PSx#px Pxx#Px P#xxxPSx xxxxPx#px Px#xxPx

3542b 3543a 3543b 3544a 3544b 3545a 3545b 3546a 3546b 3547a 3547b 3548a 3548b 3549a 3549b 3550a 3550b 3551a 3551b 3552a 3552b 3553a 3553b 3554a 3554b 3555a 3555b 3556a 3556b 3557a 3557b 3558a 3558b 3559a 3559b 3560a 3560b 3561a 3561b 3562a 3562b 3563a 3563b 3564a 3564b 3565a 3565b 3566a 3566b 3567a 3567b 3568a 3568b 3569a 3569b 3570a

D F F A1 B1 E B1 A1 D F C A1 B1 A1 B1 A3 A1 A3 A1 A1 C B1 A1 A1 B1 C A1 A1 B1 F F E A1 A1 B1 A1A1 C A1 B1 D*A1 A3A1 F A3A1 A1 B1 A1 B1 A1 B1 A1 B1 A1 B1 B1 B1 A1

433 P#Pxx xxxxHierusalem Iudeo#Px Px#Px xxxxpxx#xpx pxSx#px xxPx#P Px#Px px#Pxx xxHiericho-S xxpx#px Px#xxxxpxx xxxxxP#xpx Px#xxxPx xxPxx#P xxxPx pxx#PX xxxPx Px#pxx pxx#Px xxxxxxpx#px xxPxx#P pxXSx pxxSx xxxPx#P xxxxPsx pXx#Px PX#xPx xxxxpx#xpx xxxHiesu#P xGalilea-Sx PXxx#P PX#Px Px#xxPx xxPS#px Px#xPx#Px xxxP#px Px#xxxxpxx xxxxxxxPx#P Px#xxxPxx#Px xxxxxPx#xPx Px#Dauides#px xxxxxPx#Px px#xxxxxPx xxxPx#P Px#Px xxpxx#P Px#xxPx xxxxxpxx#P Px#xPx xxxPX#P xPx#xPx xxxxxPx#xP xxxxP#xxP xxxPx#P Px#Px

MOSP-Index 1.qxd

5/11/04

434 3570b 3571a 3571b 3572a 3572b 3573a 3573b 3574a 3574b 3575a 3575b 3576a 3576b 3577a 3577b 3578a 3578b 3579a 3579b 3580a 3580b 3581a 3581b 3582a 3582b 3583a 3583b 3584a 3584b 3585a 3585b 3586a 3586b 3587a 3587b 3588a 3588b 3589a 3589b 3590a 3590b 3591a 3591b 3592a 3592b 3593a 3593b 3594a 3594b 3595a 3595b 3596a 3596b 3597a 3597b 3598a

2:49 PM

Page 434

Index to the scansion of the Heliand E B1 B1 A1 B1 A1 B1 A1 C A3 A1 A1 B1 D* C E A1 A1 B1 B1 A1 A1 B1 C A1 A1 C A1 C A1 B1 A1 C B1 or E A1 A1 B1 A1 B1 C A1 A1 B1 C B1 C B1 E A1 F C A1 B1 A1 C A1

PX#xP xxxPx#P xxxPx#xpx Px#xxPx xxPx#P PSx#xxxpxx xxxxPxx#P(xx) Pxx#Px xxxxxPXx xxxxxPx pxx#xPx xpx#xxPx xxPx#P Px#Px#P xpx#Px pxSx#pX PX#pxx P#xxxxPx xxxPx#xP xxPx#P Px#Px Px#xpxx xxPx#xxpx xP#px xPx#Px P#xxPx xxxxP#px Px#xxPx xxpx#Px Px#Px xxxxPx#xpx Pxx#xPx xxxpx#pxx xxxPXx#P Px#xPx pxx#xPx xxxPXS pxx#xPx xxPx#P xxxpx#Px Px#pxx Px#Px xxxxPx#P xPSx xxPx#P xxxpxXx xxPx#P PSx#P Px#xPx Adam#xxEvan xxxPsx pxx#Px xxxxxpxx#xP PX#xPx xxxpXSx xP#xxPx

3598b 3599a 3599b 3600a 3600b 3601a 3601b 3602a 3602b 3603a 3603b 3604a 3604b 3605a 3605b 3606a 3606b 3607a 3607b 3608a 3608b 3609a 3609b 3610a 3610b 3611a 3611b 3612a 3612b 3613a 3613b 3614a 3614b 3615a 3615b 3616a 3616b 3617a 3617b 3618a 3618b 3619a 3619b 3620a 3620b 3621a 3621b 3622a 3622b 3623a 3623b 3624a 3624b 3625a 3625b 3626a

C A1 B1 B1 A1 A1 D D* D D* D A1 B1 A1 C B1 E A3 E A1 B1 A1 A1 A1 D B1 A1 A1 E A3 B1 B1 A1 B1 A1 B1 A1 D* C B1 E A3 C A1 B1 or E A1 C B1 A1 A3 A1 A1 B1 F F A1

xxPSx P#xPx xxxPxx#px xxxPXS P#xPx pxx#xxPx PSxx Px#PSx px#Pxx xPx#px#Px px#Pxx PXx#Px xxxxxxPx#P xxxPx#Px xxxxxxxxpx#Px xxxxPXS PXxx#P xxxxxxPx PXx#P pxXx#Px xxxxxxPx#xP xPx#xxPx xpX#xxxxPx xPx#xPx xPSxx xPx#px xxxxxxxPxx#Px PXSx px#Px#px xxxxxxxPx xxPX#P xxxPXS Px#Px xPx#px Px#Px xxP#xpx Px#Px pxxx#xPx#P xxxxpxSx xxxPx#P PXx#P xxxxxPx xxxP#Px Px#xxPx xxxPXx#px P#xPx xxpx#Px xxxxPXS pxx#xPx xxxxxPx Px#Px pxx#Px xxxxPx#P Hiericho#Px xxxIudeon#P xpx#xPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 435

Index to the scansion of the Heliand 3626b 3627a 3627b 3628a 3628b 3629a 3629b 3630a 3630b 3631a 3631b 3632a 3632b 3633a 3633b 3634a 3634b 3635a 3635b 3636a 3636b 3637a 3637b 3638a 3638b 3639a 3639b 3640a 3640b 3641a 3641b 3642a 3642b 3643a 3643b 3644a 3644b 3645a 3645b 3646a 3646b 3647a 3647b 3648a 3648b 3649a 3649b 3650a 3650b 3651a 3651b 3652a 3652b 3653a 3653b 3654a

A1 A1 A1 A1 D A1 B1 B1 E A1 A1 C A1 D* B1 C B1 F A1 A1 B1 D* B1 B1 E D* B1 A1 B1 A1 A1 A1 C A1 C D* B1 A1 C D* B1 A1 A1 A1 A1 A1 B1 A1 B1 A1 B1 A3 C D C B1

xxxxxxPx#xPx xxxxPx#Px xxxxPx#xPx xxpx#xpxx P#PXsx px#xxPx xxxxxxpXx#P xxxxPXS PXxx#P px#xxPx Px#Px xxP#Px Px#pxx px#xPxx xxxxP#xpx xxxxP#px xxxxxPx#P xPx#xHiericho xxxxxxxpxx#Px xPx#xPx xxxPx#P xPx#PSx xxxPx#P xxxxPXS PXx#xP P#xxPsx xxxxpxx#P xP#xxxxpXx xxxPx#P Px#xPx xPx#Px pxx#xPx xxxxxxxxxP#px PX#xPx xpXSx P#xxxpXx#P xxxxxPx#P xPx#xPx xxxpx#Px xP#xxxPxx#px xxPx#xxP PX#Px xpxx#xPx xPx#Px xxxxPx#xPx Px#xxxPx xxxxxPx#P xPx#xxPx xxxPX#P Px#Px xxxxxxPx#xP xxPS xpx#Px px#Px#P xxP#px xxPx#P

3654b 3655a 3655b 3656a 3656b 3657a 3657b 3658a 3658b 3659a 3659b 3660a 3660b 3661a 3661b 3662a 3662b 3663a 3663b 3664a 3664b 3665a 3665b 3666a 3666b 3667a 3667b 3668a 3668b 3669a 3669b 3670a 3670b 3671a 3671b 3672a 3672b 3673a 3673b 3674a 3674b 3675a 3675b 3676a 3676b 3677a 3677b 3678a 3678b 3679a 3679b 3680a 3680b 3681a 3681b 3682a

B1 F C A1 B1 A1 B1 A1 C A1 C B1 ? A3 B1 B1 C A3 B1 A1 C E* B1 A1 B1 A1 C E A1 D B1 A3 C C E F B1 A1 D A1 C A1 B1 A1 A1 A1A1 B3A1 C A1 B1 C A1 B1 A1 B1 or E B1

435 xxxxPx#P xxxHiericho-S xxxpx#Px Px#Px xxxxxPx#xpx Px#xxxxPx xxxPx#xpx pxx#xPx xxxxxpx#Px xpx#xxxPx xxxpXSx xxxxPXS PSx xxxxxPx xxxxPx#P xxPx#P xpx#Px xxxPx xxxxxxxPxx#px Pxx#xPx xxpx#Px xxPx#Px#xpx xxxxPx#P Px#xxxxxpXx xxxPX#P xPx#xxPx xxxP#Px px#Px#xP Px#Px P#pxx#P xxxPx#xP xxxxxxPx xxxxxxxpx#Px xPxx PXx#P xPx#xHierusalem xxxxxPx#px pxx#xPx P#PXx xPx#xxpxx xxxxxpx#Px xxP#xxxxPx xxxPx#xpx xPx#Px xxxPx#Px xP#xpxx#Px xxxP#xpxx xxpx#px Px#Px xxxPx#P xxxpx#Px Px#xPx xxPS#xP xpx#xPx xxPXx#P xxPx#P

MOSP-Index 1.qxd

5/11/04

436 3682b 3683a 3683b 3684a 3684b 3685a 3685b 3686a 3686b 3687a 3687b 3688a 3688b 3689a 3689b 3690a 3690b 3691a 3691b 3692a 3692b 3693a 3693b 3694a 3694b 3695a 3695b 3696a 3696b 3697a 3697b 3698a 3698b 3699a 3699b 3700a 3700b 3701a 3701b 3702a 3702b 3703a 3703b 3704a 3704b 3705a 3705b 3706a 3706b 3707a 3707b 3708a 3708b 3709a 3709b 3710a

2:49 PM

Page 436

Index to the scansion of the Heliand F A1 B1 F A1 D* F D* C D* A1 A1 B1 A1 B1 A1 C F B1 A1 B1 B3 A1 A3 E A1 C A1 B1 A1 E D* A1 A1 A1 A1 B1 A1 C F F A3 B1 D* B1 A1 E A3 A1 B1 C F A1 A1 B1 E

px#Dauides Px#xpxx xxxPX#P xPx#xHierusalem pxx#Px Px#xxPx#P xxP#Iudeono Px#Psx xxxxP#px xxxPx#PsX xP#xxxPx px#xxPx xxxxxPxx#P Px#xPx xxPx#px PXSx xxxxpx#Px P#xxHierusalem(xx) xxxPx#xP xPxxSx xxxxPx#px xxxxxxP Px#Px xxxxPx PSx#P PX#xPx xxxxxpx#Px PS#xPx xxxxpxx#P Px#xxPx Psx#P xP#xPsx Px#Px xxPx#xPx Px#Px Px#xPx xxxxpx#xP P#xxPxx xxxxxpx#Px xxHierusalem Iudeo#Px xxxxPx xxpxx#P xxPx#PSx xxxxxPx#px xpx#xxxPx PXx#P xxxxxxxxpxx Px#Px xxPx#P xxxP#px xxHierusalem xxpxx#Px P#xxxPx xxxxxxPx#P P#Px#xpx

3710b 3711a 3711b 3712a 3712b 3713a 3713b 3714a 3714b 3715a 3715b 3716a 3716b 3717a 3717b 3718a 3718b 3719a 3719b 3720a 3720b 3721a 3721b 3722a 3722b 3723a 3723b 3724a 3724b 3725a 3725b 3726a 3726b 3727a 3727b 3728a 3728b 3729a 3729b 3730a 3730b 3731a 3731b 3732a 3732b 3733a 3733b 3734a 3734b 3735a 3735b 3736a 3736b 3737a 3737b 3738a

A1 D* A1 A1 A1 A1 B1 A3 B1 A1 B1 F F F A1 D A1 F C B1 A1 B1 A1 A1 C B1 A1 A3 A1 A1 B1 A1 B1 C C A3 B1 E A1 B1 A1 C A1 A3 C A1 C A2 F E A1 C C D* A1 A1

Pxx#Px pxx#xxPx#P Px#pxx Px#xPx xP#xxPx xP#xxPx xxPxx#P xxPx xxxxpxx#P xxxpxx#pxx xxxxP#xpx xxxHiesu#P xGalileo#Px xNazareth-S PX#Px px#Psx xxpxx#xPx xxxIudiun xxxxpx#Px xPx#xpx P#xPx xxxxPx#xpx PS#Px Px#xxPx xxxPSx xxxPX#P Px#Px xxxxxxPx Px#Px Px#xPx xxxxP#xpx Px#xPx xxxxxpxx#P(xxx) xxPSx xxxxP#px xxxxPx(xx) xxxxxPx#P PXx#P Px#Px xxxxPx#P Px#Px xxxxPsx pxx#Px xxxxxPx xxxxP#px Px#xxxxxpXx xxxxxP#Px P#xxpx#P xxIudeono#px PSx#P pxx#xPx xxxPsx xpx#Px Pxx#xxxpxx#P pxXx#Px xxxPx#Px

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 437

Index to the scansion of the Heliand 3738b 3739a 3739b 3740a 3740b 3741a 3741b 3742a 3742b 3743a 3743b 3744a 3744b 3745a 3745b 3746a 3746b 3747a 3747b 3748a 3748b 3749a 3749b 3750a 3750b 3751a 3751b 3752a 3752b 3753a 3753b 3754a 3754b 3755a 3755b 3756a 3756b 3757a 3757b 3758a 3758b 3759a 3759b 3760a 3760b 3761a 3761b 3762a 3762b 3763a 3763b 3764a 3764b 3765a 3765b 3766a

B1 A1 C A1 C A1 B1 C F A3 A1 C A1 B1 A1 B1 A1 A2 B1 C F A1 A1 B1 B1 D* B1 A1 C B1 B1 A1 B1 A1 C A1 B1 B1 B1 A3 E D B1 A1 C B1 A1 A1 C A1 B1 D* B1 D B1 A3

xxxxP#xpx px#xpXx xxxxpx#Px P#xPx xxP#px Px#xxxpxx xxxxPxx#P xxxpx#Px P#Israheles xxxxxxxxxPx Pxx#Px xxpxSx Pxx#Px xxPx#P Psx#Px xxPx#px Px#Px PS#PS xxxxxPx#xpx xxpx#Px Iudeo#Px xPx#xxxxpxx Px#Px xPxx#P xxxPx#P pxxx#PSx xxxpxx#P Pxx#xPx xxxxpx#Px xxPx#P xxP#xpx P#xPx xxP#xpx Px#xPx xxxxP#px PXx#xxpxx xxxxxPx#P xxxxxPx#P xxxxxxP#xpx xxxxxxxxPx PXx#P P#Px#P xxxxxxPx#px xxPx#xxpxx xxpx#px xxPx#P Px#Px px#xPx xxxPSx Px#pxx xxPx#P pxx#PSx xxxxxpxx#P px#Psx xxxxxpxx#P xxxxxpXS

3766b 3767a 3767b 3768a 3768b 3769a 3769b 3770a 3770b 3771a 3771b 3772a 3772b 3773a 3773b 3774a 3774b 3775a 3775b 3776a 3776b 3777a 3777b 3778a 3778b 3779a 3779b 3780a 3780b 3781a 3781b 3782a 3782b 3783a 3783b 3784a 3784b 3785a 3785b 3786a 3786b 3787a 3787b 3788a 3788b 3789a 3789b 3790a 3790b 3791a 3791b 3792a 3792b 3793a 3793b 3794a

A1 A1 B1 A1 B1 D* C A1 B1 B1 A1 A1s B1 A1 B1 B1 A1 A1 B1 A1 C A1 B1 C C B1 B1 A3 E A1 A1 A1 C F B1 A1 B1 A1 C A1 B1 B1 D A3 A1 E B1 C A1 A1 A1 A1 C A1 B1 or E F

437 Px#Px PXx#Px xxxPS#px Px#Px xxPX#P xpx#xxPxx xxxxpx#Px Px#pxx xxxxxPx#px xxPxx#P(xx) Px#Px PXS#px xxxxP#xP pxx#xPx xxxxpxx#P xxxxxxpxx#P P#xxPx pxx#xPx xxxxP#xxP Px#xxxPx xxxxxpx#Px PXx#Px xxxxxxxxPx#px xxxxpx#Px xxxP#px xxPS#P xxxxxxPx#px xxxxxxxxxPx PXx#P xxxpx#xpxx Px#Px Px#xPx xxxpx#Px P#P#Iudeono xxxxPx#P Px#Px xxPx#P Px#xxxxpxx xxxxxxP#px Pxx#xxPx xxPx#P xxpXx#P P#Pxx xxxxxxPx Px#Px PXx#P xxpXS#px xxxxxxxpxsx PX#Px pXx#xxxPx Px#Px Px#xxxPx xxxpXsx xpxx#xPx xPXx#P Erodeses#px

MOSP-Index 1.qxd

5/11/04

438 3794b 3795a 3795b 3796a 3796b 3797a 3797b 3798a 3798b 3799a 3799b 3800a 3800b 3801a 3801b 3802a 3802b 3803a 3803b 3804a 3804b 3805a 3805b 3806a 3806b 3807a 3807b 3808a 3808b 3809a 3809b 3810a 3810b 3811a 3811b 3812a 3812b 3813a 3813b 3814a 3814b 3815a 3815b 3816a 3816b 3817a 3817b 3818a 3818b 3819a 3819b 3820a 3820b 3821a 3821b 3822a

2:49 PM

Page 438

Index to the scansion of the Heliand B1 A1 A1 A3 B1 D* A1 A1 B1 A1 C D* A1 C A1 D* C C A1 B1 C A3 A1 A1 B1 A1 B1 A1 C F B1 A1 C B1 A1 A1 C B1 A1 E B1 A3 B1 A1 C A1 C C E A3 C A1 B1 C F B1 or E

xxPS#P Px#Px xxxxP#xxPx xxxxxPx xxxxxpxx#P xPx#pxSx Px#Px Px#Px xxxxxxPx#P Px#xPx xxxxP#px Px#pXsx Px#Px xxxPsx(xxx) Px#Px Px#Px#xpx xxP#PX xxPXx Px#xPxx xxxPxS xxxxxxP#px xxxxxpx#P pxx#xPx Px#xxxPx xxxPx#P Px#xxxxxPx xxxPx#px Px#Px xxP#px xPx#xRumu xxxxxxxPx#px Px#Px xxxP#px xxxxPx#px Px#xpxx PXSx xxxxxxxxxpx#Px xxP#xP P#xPx PSx#P xxxxxPxx#P xxxxxxxPx xxxxxxPx#P xxPx#Px xxxPsx(xx) Px#xxPx xxxxxpx#Px xxPXx PXx#P xxxxPx xxxP#px xPXx#xPx xxPxx#P xxP#px Iudeon#Px xxpXXx#P

3822b 3823a 3823b 3824a 3824b 3825a 3825b 3826a 3826b 3827a 3827b 3828a 3828b 3829a 3829b 3830a 3830b 3831a 3831b 3832a 3832b 3833a 3833b 3834a 3834b 3835a 3835b 3836a 3836b 3837a 3837b 3838a 3838b 3839a 3839b 3840a 3840b 3841a 3841b 3842a 3842b 3843a 3843b 3844a 3844b 3845a 3845b 3846a 3846b 3847a 3847b 3848a 3848b 3849a 3849b 3850a

D* A3 B1 A1 B1 D* B1 C C A3 D F B1 A1 B1 A1 C E B1 A1 B1 A1 F A1 C A1 B1 A3 B1 A1 B1 A1 B1 A3 C A3 B1 A1 B1 A2 A1 C B1 B1 A1 A1 B1 A3 A1 A1 B1 A3 C A3 A1 A3

Px#pxx#P xxxxpxx xxPx#P xPxx#pxx xxxxxPx#P xxPx#PXS xxxxxxPxx#P xxxxxxPSx xpx#Px xxxxxPx pXSxx xRumu-S xxxxxxPx#px xP#xxxxpXx xxxxxxPx#P(xx) Px#Px xxxxP#px pXSx#xxP xxPX#px Px#xxP#x xxxxxPx#px pxx#Px xxxxIudeono#px xPx#xxxPx xxxxPsx Px#xPx xxxPx#P xxxxxPx xxxxpxS#px Pxx#xxPx xxxP#xpx PS#Px xxxxxxPxx#xP xxxxxPx xxxxpx#Px xxxxxxxPx xxxxPx#P xP#xxxpxx xxxP#xpx PS#PS xpx#xxPx xxpXSx xxxPx#px xxpxx#P pxx#xPx Px#xxPx xxxxxP#xpx xxxxxxxPx Px#Px Px#xxxPx xxxxxxPx#P xxxxpxx xxxpx#Px xxxxxxxxPx xPx#Px xxxxxxpxx(xxx)

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 439

Index to the scansion of the Heliand 3850b 3851a 3851b 3852a 3852b 3853a 3853b 3854a 3854b 3855a 3855b 3856a 3856b 3857a 3857b 3858a 3858b 3859a 3859b 3860a 3860b 3861a 3861b 3862a 3862b 3863a 3863b 3864a 3864b 3865a 3865b 3866a 3866b 3867a 3867b 3868a 3868b 3869a 3869b 3870a 3870b 3871a 3871b 3872a 3872b 3873a 3873b 3874a 3874b 3875a 3875b 3876a 3876b 3877a 3877b 3878a

F A1 B1 C A1 A3 A1 A1 B1 A1 A1 D* A1 A3 C A1 F C A1 A2 B1 A1 B1 A1 C A3 E A1 A1 A1 B1 B1 A1 A1 B1 A1 B1 A1 B1 A3 A1 A1 F A1 B1 C A1 B1 A1 A1 C A3 A1 A1 B1 A1

Moyses#xP Px#Px xxxxPx#xpx xxpXSx Px#xPx xxxxxxPx px#xPx Px#Px xxxxxxPx#P xPx#xPx px#xxxPx Px#xxxpXsx Px#xPx xxxxPx xxxpx#Px pxx#xxpxx xxxxP#Iudeono xxxxxxPxx P#xxPx xxPx#PS xxxxxxPx#xpx xpx#xxxpxx xxxxxxxxxPxx#px xPx#xxPx xxxxxP#px xxxxpxx PXx#P xxPx#xPx xxxxxxpxx#xPx xPx#xPx xxxPx#P xxPx#xpx PxSx xxPx#Px xxxxxpxx#P xPx#xPx xxxxxxPx#P(xx) Px#Px xxxxxPx#P xxxxPX P#xxPx P#xxPx xPx#Iudeon Px#xxpxx xxxpx#xP xxPsx pXsx#Px xxxPx#xpx PxSx xPx#Px xxxxxpx#Px xxxxxPx xxPx#xPx P#xPx xxxxPx#P Px#xPx

3878b 3879a 3879b 3880a 3880b 3881a 3881b 3882a 3882b 3883a 3883b 3884a 3884b 3885a 3885b 3886a 3886b 3887a 3887b 3888a 3888b 3889a 3889b 3890a 3890b 3891a 3891b 3892a 3892b 3893a 3893b 3894a 3894b 3895a 3895b 3896a 3896b 3897a 3897b 3898a 3898b 3899a 3899b 3900a 3900b 3901a 3901b 3902a 3902b 3903a 3903b 3904a 3904b 3905a 3905b 3906a

C D* F A1 B1 A1 B1 D A1 D* A1s B1 F C B1 B1 A1 A1 B1 A1 B1 A3 C A1 A1 A1 B1 A1 A1 C C B1 A1 A1 D* A1 F A2 C A1 C B1 A1 B1 A1 C A1 A1 E C B1 A1 B1 A1 C A1

439 xxxP#px Px#PSx Iudeo#Px P#xxPxx xxxxxPx#xP xPXx#Px xxxPx#P xxpx#PXsx Px#Px xxP#xxxPxx pxS#px xxxpxx#P xxxxIudeono#P(xx) xxpXsx xxxPx#xP xxxPx#P Px#xxPx xPx#Px xxxxPx#xpx Px#xPx xxxxxxP#xpx xxxxxPx xxPXx Pxx#Px P#xxpxx Px#xPx xxxPX#P Px#Px xxxxxPx#PX(xx) xxxP#px xxxxxxpx#Px xxxP#xxP Px#xPx Px#xxxxPx Px#P#px xpx#xxpxx xxxP#Iudeono pxx#PS xxPXx Px#Px xxPsx xxpxS#px Px#Px xxxPx#xP xxxxPx#xPx xxpx#Px Px#Px Pxx#pxx xpx#Px#P xxPXx xxxxPx#xP Pxx#xPx xxxPx#P Px#xxxxxPx xxpXsX xPx#xxPx

MOSP-Index 1.qxd

5/11/04

440 3906b 3907a 3907b 3908a 3908b 3909a 3909b 3910a 3910b 3911a 3911b 3912a 3912b 3913a 3913b 3914a 3914b 3915a 3915b 3916a 3916b 3917a 3917b 3918a 3918b 3919a 3919b 3920a 3920b 3921a 3921b 3922a 3922b 3923a 3923b 3924a 3924b 3925a 3925b 3926a 3926b 3927a 3927b 3928a 3928b 3929a 3929b 3930a 3930b 3931a 3931b 3932a 3932b 3933a 3933b 3934a

2:49 PM

Page 440

Index to the scansion of the Heliand C C C E B1 A1 A1 A1 C D* B1 A3 A1 B1 A1 A1 B1 B1 A1 A1 B1 C E E A1s D* A1 A1 B1 A3 E D E B1 A1 A1 C D B1 A3 A1 A1 B1 F C A1 B1 B1 A1 E C A1 C B1 B1 A1

xxxxxxpx#px xxxP#Px xxxpx#Px pXSx#xP xxPx#P Px#xPx Px#Px Pxx#Px xP#px pxx#px#px xxxxxP#xP xxxxxPx xPx#Px xxxxxxxPx#xP(xx) px#xpxx Px#Px xxPx#P xxxxxPx#xP Px#Px Px#xxxxxPx xxxxPx#px xxPsx PXx#P PXx#px pxS#px px#xxpx#Px xxpx#xxPx Px#xPx xxxxxxPx#xP xxxxxpxx PXx#P P#pXsX PXx#P xxxpxx#P xPx#Px P#xxPx xxP#Px P#pXSx xxPx#px xxxxPx xxxPx#Px xxxP#xxPx xxPx#P PSx#Iudeon xxP#px Px#xxxPx xxxxxxxPx#P xxxPxx#P Pxx#Px PSx#P xxxpx#Px(xxx) Px#xpxx xxxxpx#Px xxxxxxPX#px(xxx) xxPx#P Px#xxPx

3934b 3935a 3935b 3936a 3936b 3937a 3937b 3938a 3938b 3939a 3939b 3940a 3940b 3941a 3941b 3942a 3942b 3943a 3943b 3944a 3944b 3945a 3945b 3946a 3946b 3947a 3947b 3948a 3948b 3949a 3949b 3950a 3950b 3951a 3951b 3952a 3952b 3953a 3953b 3954a 3954b 3955a 3955b 3956a 3956b 3957a 3957b 3958a 3958b 3959a 3959b 3960a 3960b 3961a 3961b 3962a

B1 A1 B1 A1 A1 C D A1 B1 A1 C A1 B1 C A1 A3 A1 D* B1 B1 C A1 B1 B1 A1 A1 B1 D* B1 B1 A1 A3 C A3 C F C A1 B1 A1 D F A1 A1 B1 F B1 A1 B1 A1 C A1 A1 A1 A1 B3

xxxpxx#P Px#xxxxpXx xxPxx#P PXx#Px xxxxxxxxpx#xPx xxPXx xpx#pxSx px#xxPx xxxxxPx#P xxxPx#Px xxxP#px Px#xxPx xxxxxxPsx#P xpx#Px xxP#xxPx xxxxxPx pxx#xxPx xPxx#xPsx xxxpxS#px xxxxPx#xpx(xx) xpx#Pxx Px#xxPx xxxxxPx#P xxxPx#px P#xxPx xPx#xPx xxxxxxxPx#xpx Px#pXsx xxxxxxPx#xP(xxx) xxxPx#P Px#Px xxxxxxxPx xxxxP#px xxxxPx xxxxP#px px#xxxIudeon xxxP#Px Px#Px xxxxxP#xP px#xxxPx P#pxSx xxxxxIudeono#xP Px#Px Pxx#Px xxxxxPx#P xxIordanes#P xxPx#xP xxPxx#xPx xxPx#xP Px#pxx xxpx#Px px#xxxxpxx Px#Px Px#xPx xPx#xPx xxPxx#P

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 441

Index to the scansion of the Heliand 3962b 3963a 3963b 3964a 3964b 3965a 3965b 3966a 3966b 3967a 3967b 3968a 3968b 3969a 3969b 3970a 3970b 3971a 3971b 3972a 3972b 3973a 3973b 3974a 3974b 3975a 3975b 3976a 3976b 3977a 3977b 3978a 3978b 3979a 3979b 3980a 3980b 3981a 3981b 3982a 3982b 3983a 3983b 3984a 3984b 3985a 3985b 3986a 3986b 3987a 3987b 3988a 3988b 3989a 3989b 3990a

B1 A1 B1 A3 A1 F B1 B1 or E A1 F D B1 A1 B1 B1 A1 B1 A1A1 B1 F B1 F A1 A3 D A1 B1 A1 B1 F A1 A1 B1 C B1 A3 D E A1 F F A1 C A1 D F F A1 B1 A1 C A1 B1 A3 A1 A1A1

xxxPx#P Px#xpxx xxxxP#xP xxxxxxxPx pxx#Px px#xBethaniu xxxxxPx#px xxxxPXx#px pxx#Px Maria#xxMartha px#PSx xxPsx#P xPx#xPx xxxxPx#P xxPx#P Pxx#xxPx xxxxPx#px xP#xxxxPx#Px xxxPx#P xpx#xBethaniu xxxPx#P Lazarus#pXS xxxxxPx#xPx xxxxxPx P#pxSx Px#xPx xxxxxPx#P Px#xxPx xxxP#xpx xxLazaruses px#xPx xP#xxPx xxxPxx#P(xx) xpx#Px xxxxxxPx#xP xxxxPx px#Pxx P#P#xxpx xP#xxxPx xxxxHierusalem Iudeo#Px Px#Px xxxP#Px Px#xxxPx px#Pxx xxxxxIordan Iudeo#Px Px#Px xxxxP#xpx Px#Px xxxxxP#px(xxx) P#xxpXx xxxP#xP xxxxxxPx Px#Px xxxxPx#Px#xPx

3990b 3991a 3991b 3992a 3992b 3993a 3993b 3994a 3994b 3995a 3995b 3996a 3996b 3997a 3997b 3998a 3998b 3999a 3999b 4000a 4000b 4001a 4001b 4002a 4002b 4003a 4003b 4004a 4004b 4005a 4005b 4006a 4006b 4007a 4007b 4008a 4008b 4009a 4009b 4010a 4010b 4011a 4011b 4012a 4012b 4013a 4013b 4014a 4014b 4015a 4015b 4016a 4016b 4017a 4017b 4018a

CA1 A1 B1 or E D* A1 F B1 D* C A1 B1 A1 B1 B1 A1 A1 D* A1 B1 A1 B1 D* C E A1 D* A1 A1 B1 A1 B1 F B1 A1 C A3 C E B1 A1 C D* B1 F D A1 B1 F C A1 B1 F F C A1 A3

441 xxxxxxPxx#Px Px#xpXx xxPXx#xP Px#pXSx xP#xxpXx Thuomas#xPx xxxPx#P PS#Pxx#px xxxxxxP#px(xx) xPx#xxxPx xxxxPx#P pxx#xxxPx xxPx#P xxxxPx#px Px#xPx Px#xxxxPx P#xPx#P Px#xxxxPx xxxxxpx#xP Px#Px xxxxxpxx#xP Px#xxxPxx xxxxxP#Px P#P#xpx xxxxPx#Px Px#pXsxx xPSx#Px Px#xPx xxxPx#P Px#xxPx xxPx#P Lazarus#xxpxx xxxP#xpx xpx#xxPx xxxxxP#px xxxxxPx xxxxxxxpX#px PXx#P xxxxxxPx#xxP PS#xPx xxxxxxxxP#px xPx#px#px xxxxxPx#P xxBithaniu P#Pxx Px#xxxPx xxxPx#P Maria#xxMartha xPsx Pxx#Px xxxPx#px xHierusalem Iudeo#Px xxP#Px Px#Px xxxxpxx

MOSP-Index 1.qxd

5/11/04

442 4018b 4019a 4019b 4020a 4020b 4021a 4021b 4022a 4022b 4023a 4023b 4024a 4024b 4025a 4025b 4026a 4026b 4027a 4027b 4028a 4028b 4029a 4029b 4030a 4030b 4031a 4031b 4032a 4032b 4033a 4033b 4034a 4034b 4035a 4035b 4036a 4036b 4037a 4037b 4038a 4038b 4039a 4039b 4040a 4040b 4041a 4041b 4042a 4042b 4043a 4043b 4044a 4044b 4045a 4045b 4046a

2:49 PM

Page 442

Index to the scansion of the Heliand E F B1 A1 C A1 B1 C B1 A1 B1 B1 A1 B1 A1 D B1 C E F C A1 B1 C B1 E A1 A1 C D* B1 F B1 A1 E A1 E B1 A1 B3 D A1 B1 D B1 C B1 A1 C B1 C B1 A1 C B1 B1

PSx#P Lazaruses#xP xxxPx#P P#xxPx xxxxpx#Px px#xxPx xxxPx#P xxxP#Px xxPx#P xPx#xxPx xxxPX#P xpxS#px pXx#Px xxPx#P Pxx#Px px#Pxx xxPx#P xxPXx PXx#P Martha#Psx xxPXx Px#Px xxxPX#P xxxpx#Px xxxPx#P(xx) PXxx#P Px#Px PX#xPx xxxxxxxxP#px Px#Psx xxxxxxPx#P Lazarus#xxxxPx xxxxxxPx#P pxx#xPx xxP#P#xP Px#xPx PXxx#P xxxxPx#px Px#Px xxxxxxpx P#pxSx xPx#xxPx xxxPX#P xxpx#PSx xxxxxPx#P(xx) xxpx#Px xxPx#P Px#Px xxxxP#Px xxxPx#P xxP#px xPxx#P xPx#xPx xxPsx xxxxxPx#P(xx) xxxxPx#P

4046b 4047a 4047b 4048a 4048b 4049a 4049b 4050a 4050b 4051a 4051b 4052a 4052b 4053a 4053b 4054a 4054b 4055a 4055b 4056a 4056b 4057a 4057b 4058a 4058b 4059a 4059b 4060a 4060b 4061a 4061b 4062a 4062b 4063a 4063b 4064a 4064b 4065a 4065b 4066a 4066b 4067a 4067b 4068a 4068b 4069a 4069b 4070a 4070b 4071a 4071b 4072a 4072b 4073a 4073b 4074a

C B1 C B1 A1 B1 B1 C E A1 B1 D A1 B1 D B1 A1 C B1 A1 B1 B1 A1 A1 B1 A1 A1 B1 B1 A1 B1 D B1 A1 C B1 A1 B1 A1 F A1 F C D* B1 C A1 F A1 D C A1 A1 A1 B1 A3

xxxxpX#Px xxxPx#P xxP#px xxxxPx#P P#xPx xxxPx#px xxxxPx#px xP#px PSx#xpx xPx#xPx xxxPx#P xxpx#pxSx pxx#Px xxPx#P px#Pxx xxxP#xP Px#Px xxPXx xxPx#xP P#xPx xxxPx#xP xxxPxS Px#xPx Px#xPx xxP#xP xP#xxPx xpx#xxPx xxPx#xP xxxxxP#xpx xP#xxxPx xxxxxxxPx#P(xx) P#px#px xxxxPx#P px#xxxPx xxxP#px xxPxx#xpx pxx#xxPx xxxxxxxxpXx#P Px#Px Maria#Psx xxxxpxx#Px Iudeo#Px xxxxpx#Px Px#PXS xxxxPx#xP xxxpx#Px Px#Px Lazaruses#xP Px#Px P#PXx xxxxpx#Px px#xxPx Px#Px Px#xPx xxxxxPx#P xxxxPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 443

Index to the scansion of the Heliand 4074b 4075a 4075b 4076a 4076b 4077a 4077b 4078a 4078b 4079a 4079b 4080a 4080b 4081a 4081b 4082a 4082b 4083a 4083b 4084a 4084b 4085a 4085b 4086a 4086b 4087a 4087b 4088a 4088b 4089a 4089b 4090a 4090b 4091a 4091b 4092a 4092b 4093a 4093b 4094a 4094b 4095a 4095b 4096a 4096b 4097a 4097b 4098a 4098b 4099a 4099b 4100a 4100b 4101a 4101b 4102a

F A1 B1 E B1 A1 C D C F B1 or E A1 C A1 C B1 B1 A1 C B1 E* C E A1 B1 B1 B1 B1 D D* B1 D* B1 A1 A1 B1 A1 A1 B1 A3 F B1 C A1 F A1 B1 C C C C A1 B1 E B1 E

xLazarus#P Px#xPx xxxpx#xpx P#P#xpx xxxPxx#P xPx#xPx xxxxxP#px P#Pxx xxxxxxxpx#Px Martha#xxxpxx xPXx#P P#xxPx(xx) xxxxpx#px xxP#xPx xxxxxxxP#px xPx#P xxxPx#P Px#Px xxxxP#Px xxxxxPx#P Px#P#xxpx xxxPsx PSx#P PX#xxPx xxxxxxxPx#P(xx) xxxPx#px xxxP#xP xxxxPx#P P#Pxx xpxx#P#px xxxpxx#P xPx#Px#P xxxPxx#P P#xxPx Px#Px xxxxxpX#xP xxxP#xPx(xx) pxSx#Px xxxxxPx#P xxxxxxxxPx Iudeono#Px xxxxPx#px xxxxxxpX#Px xxPx#xPx xxxLazaruse#P Pxx#Px xxxxxPx#P xxxxpx#Px xxxP#px xxxPsx xxxxpx#Px xP#xxxxxPx xxxxxPx#xP xPSx#xpx xxxPx#P PXx#P

4102b 4103a 4103b 4104a 4104b 4105a 4105b 4106a 4106b 4107a 4107b 4108a 4108b 4109a 4109b 4110a 4110b 4111a 4111b 4112a 4112b 4113a 4113b 4114a 4114b 4115a 4115b 4116a 4116b 4117a 4117b 4118a 4118b 4119a 4119b 4120a 4120b 4121a 4121b 4122a 4122b 4123a 4123b 4124a 4124b 4125a 4125b 4126a 4126b 4127a 4127b 4128a 4128b 4129a 4129b 4130a

D A1 D* F B1 B1 A1 D A1 F C A1 B1 A1 D F C A1 B1 A3 A1 A1 B1 A1 C D* A1 A1 B1 or E A1 B1 A1 A1 D* B1 A1 B1 A1 B1 E C D* C A1 C A1 A1 F F D* B1 D A1 D* C A1

443 px#Px#P xPx#xxPx Px#P#xP Lazarus#xxxxPx xxxxP#xpx xxxPXsx Px#Px px#PSx xpxxx#Px Maria#xxMartha xxxP#Pxx xPx#xPx xxxPx#P Pxx#xxxPx px#Pxx Iudeo#Px xxxxxxxpx#Px Px#xPx xxxxP#xP xxxxPx Px#xxPx Px#Px xxxxxPx#P P#xPx xxpXsXx xpxx#P#px Px#xpxx pxx#xPx xxxPXx#P Px#Px xxxxxxPx#xpx xxxxpxxx#Px P#xxPx xPx#pxSx xxxxPxx#P Px#xPx xxxpx#xP Px#xpXx xxxxpxx#xpx xPSx#P xxxxP#px xPx#PSx xxxxP#px Px#xxxPx xxxPXx Px#xPx xxxPx#Px xHierusalem xIudeono#P Px#Psx xxPXsx P#Psx Pxx#Px xPx#xxPx#P xxxxpx#Px xxP#xxxPx

MOSP-Index 1.qxd

5/11/04

444 4130b 4131a 4131b 4132a 4132b 4133a 4133b 4134a 4134b 4135a 4135b 4136a 4136b 4137a 4137b 4138a 4138b 4139a 4139b 4140a 4140b 4141a 4141b 4142a 4142b 4143a 4143b 4144a 4144b 4145a 4145b 4146a 4146b 4147a 4147b 4148a 4148b 4149a 4149b 4150a 4150b 4151a 4151b 4152a 4152b 4153a 4153b 4154a 4154b 4155a 4155b 4156a 4156b 4157a 4157b 4158a

2:49 PM

Page 444

Index to the scansion of the Heliand B1 A1 E* A1 A1 A1 C A2 A1 F B1 D A1 E B1 or E A1 C A3 B1 A1 B1 B1 C F B1 A1 B1 A1 B1 C B1 C A1 F B1 C F C A1 A1 B1 A1 B1 B1 F B1 A1 B1 A1 A3 A1 C A1 C C A1

xxPx#P Px#xPx Px#P#xxpx P#xPx xxxxxxxPx#Px xxPx#xpxx xxxxxxpX#px xP#xxpxS Px#Px Iudeo#Px xxxxPsx#P px#Pxx xxPx#Px pXSx#xP xPXx#P Px#xPx xxP#Px(xxx) xxxxpxx xxxxxPx#xpx xPx#xxxPx xxPx#px xPXS xxxxpXSx Px#xRumu xxxxPx#px Px#Px xxxxxxxPx#px pxx#xxxPx xxxxxPx#P xxP#px xxxpxx#P xxxP#Px Px#xxPx Kaiphas#xxPx xxxxxpxx#xP xxxPsx Iudeo#Px xxxpx#Px Px#Px Px#xPx xxxPx#px(xx) Px#Px xxxpxx#xP xxxxPx#xpx px#Iudeono xxxpxx#P Px#xpxxx xxxPx#P Px#xPx xxxxxxxPx Px#Px xxxxPsx px#xPx xxxPsx xpx#Px xP#xxxPx

4158b 4159a 4159b 4160a 4160b 4161a 4161b 4162a 4162b 4163a 4163b 4164a 4164b 4165a 4165b 4166a 4166b 4167a 4167b 4168a 4168b 4169a 4169b 4170a 4170b 4171a 4171b 4172a 4172b 4173a 4173b 4174a 4174b 4175a 4175b 4176a 4176b 4177a 4177b 4178a 4178b 4179a 4179b 4180a 4180b 4181a 4181b 4182a 4182b 4183a 4183b 4184a 4184b 4185a 4185b 4186a

B1 A1 D A1 C B1 C F A1 A1 B1 A1 C D* A1 A1 B1 B1 A1 A1 E C E F A1 A1 C A3 A1 A3 C B1 B1 B1 A1 A1 B1 B1 A1 E C B1 A1 C B1 F C B1 B1 A3 A1 C A1 E B1 F

xxxP#xP xP#xxxPx px#PSx xPx#xxxpxx xxxxxxxxP#px xxPxx#P xxxP#px xxHierusalem xPx#Px Px#xPx xxxxP#xP Px#xxPx xxxxP#px Px#PXS xxPx#Px xxPx#xPx xxxxxxxxPx#P xxxpxx#xP Pxx#Px px#xxPx PXx#P xxxPSx pXSx#P px#Iudeono xxxxxPx#xPx Px#Px xxxxxxxxP#px xxxxxxxxxPx Px#Px xxxxxPx xxP#Px xxxPx#P xxxxxxxxpxx#P xxxPx#P xPx#Px px#xPx xxxPX#P xxPx#xpx PxSx pxSx#px xxxxxxpx#PX xxxxPXS xxxxxxxpxx#Px xxpXSx xxxPx#P Px#xxxIudeon xxpx#px xxpXx#P xxxPS#P xxxxxPx pxx#Px xxpx#Px Px#xxPx xPSx#px xxxxxxxPx#P xxxxxxEffrem

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 445

Index to the scansion of the Heliand 4186b 4187a 4187b 4188a 4188b 4189a 4189b 4190a 4190b 4191a 4191b 4192a 4192b 4193a 4193b 4194a 4194b 4195a 4195b 4196a 4196b 4197a 4197b 4198a 4198b 4199a 4199b 4200a 4200b 4201a 4201b 4202a 4202b 4203a 4203b 4204a 4204b 4205a 4205b 4206a 4206b 4207a 4207b 4208a 4208b 4209a 4209b 4210a 4210b 4211a 4211b 4212a 4212b 4213a 4213b 4214a

E B1 A1 A1 B1 F A1 D* F A1 C A1 C A1 C D* C A1 C A1 B1 C A1 C F E C F F C A1 A3 A1 F C A1 B1 A1 C F B1 D* A1 B1 B1 A1 B1 C B1 or E D*A1 B1 A1 B1 A1 A1 F

pxSx#P xxxPx#P Px#Px pxx#xxpxx xxxxxPx#P xxBethania Px#xpxx xxxPx#Psx Iudeon#xPx#P Px#xpxx xxxxxxpx#px Px#xPx xpx#Px(xxx) xxPx#xPx xxP#px xPx#xxPx#P xP#px Px#xxxPx xxxxpX#Px Px#xxxPx xxxxPx#P xxxxPsx xPx#xPx xxxxxxP#px xxBethania P#Px#P xxPXx xxHierusalem Iudeo#Px xxPsx Px#Px xxxxPx xPxx#Px Iudeono#Px xxpx#px Px#xxxxpxx xxPx#P pxx#xxxPx xxxxxxP#Px Maria#xxMartha xPx#px Pxx#xxPSx Px#Px xxPS#P xxPx#xP Pxx#pxx xxPx#xP xxxpx#Px xxPXx#P xpxx#xxPsx#Px xxxxPXsx xPx#xxPx xxxxxPX#P P#xxPx pxx#Px xxHierusalem

4214b 4215a 4215b 4216a 4216b 4217a 4217b 4218a 4218b 4219a 4219b 4220a 4220b 4221a 4221b 4222a 4222b 4223a 4223b 4224a 4224b 4225a 4225b 4226a 4226b 4227a 4227b 4228a 4228b 4229a 4229b 4230a 4230b 4231a 4231b 4232a 4232b 4233a 4233b 4234a 4234b 4235a 4235b 4236a 4236b 4237a 4237b 4238a 4238b 4239a 4239b 4240a 4240b 4241a 4241b 4242a

F D* B1 A1 B1 A1 B1 A1 C C A1 A1 A1 D* C A1 A1 A1 C A1 B1 C E A3 B1 A1 C C A1 A1 B1 A1 C E* C B1 B1 A1 B1 E B1 C B1 A1 F F B1 E B1 C F A1 B1 A1 D D*

445 xIudeono#P pxS#Psx xxxPxx#P pxx#xxxPx xxpxx#xpx Pxxx#Px xxxxPx#P Px#Px xxxxpx#Px xxxPsx Px#xPx xPx#xxPx Px#Px Px#PSx xxxPsx PS#xPx xpx#xPx Px#Px xxxxPXx Px#xxPx xxxpxx#xpx xxPsx PSxx#P xxxxxpx#P xxPx#P pxx#xPx xxxxpXsx xxxxPsx Px#xxPx xpx#xxxxpxx xxPx#P xxxPx#Px xxP#px pxx#Px#xpx xxP#Px xxPx#P xxxPx#xP Px#xPx xxpxx#P PSx#px xxxxPx#P xxxP#Px xxP#xxP Px#xxPx xxxxIudeo#Px Oliueti#xpx xxxP#xP PXx#P xxxxP#xP xxxxxxPxx xxxxIudeono#Px xPx#xPx xxxxxPx#P Px#Px xP#Pxx#P xP#xPsx

MOSP-Index 1.qxd

5/11/04

446 4242b 4243a 4243b 4244a 4244b 4245a 4245b 4246a 4246b 4247a 4247b 4248a 4248b 4249a 4249b 4250a 4250b 4251a 4251b 4252a 4252b 4253a 4253b 4254a 4254b 4255a 4255b 4256a 4256b 4257a 4257b 4258a 4258b 4259a 4259b 4260a 4260b 4261a 4261b 4262a 4262b 4263a 4263b 4264a 4264b 4265a 4265b 4266a 4266b 4267a 4267b 4268a 4268b 4269a 4269b 4270a

2:49 PM

Page 446

Index to the scansion of the Heliand C A1 A1 B1 A1 E* B1 A1 B1 A1 B1 C C B1 B1 C A1 A1A1 B1 C C A1 A1 D* E A1 C D B1 A1 C A1 D D* D A1 E A1 B1 A1 B1 F B1 E – A1A1 C B1 A1 D* B1 A1 C C A1 D*

xxxpx#Px Pxx#Px xxxxxxpXx#Px xxxxPXS Px#xPx Px#Px#xP xxxPxx#px Px#xPx xxxxxxpxx#xP PX#xxxPx xxxxxxPx#xP xxxpxxx xpx#Px xxxPx#xpx xxxxxxxPx#px xxPxx Pxx#xPx Px#xxxxPx#pxx xxPx#xP xxxxPxx xxP#px Px#xxxPx P#xPx pxx#pXSx PSx#px px#xxxxPx xxxxxxpXSx px#Px#P xxxpxx#xpx xPx#xxxPx xxxxP#px Px#xPx pXsXx xPx#P#px px#Pxx Px#Px xxpXSx#P PXx#xPx xxPx#px Px#Px xxPx#P Iudeo#Px xxPx#px PSx#px – xPx#xPx#xPx xxxxxP#px xxPx#P px#xPx xPx#xPx#P xxxPx#xpx Px#xPx xxxxpx#pxx xxpXSx Px#Px P#xxxxpx#px

4270b 4271a 4271b 4272a 4272b 4273a 4273b 4274a 4274b 4275a 4275b 4276a 4276b 4277a 4277b 4278a 4278b 4279a 4279b 4280a 4280b 4281a 4281b 4282a 4282b 4283a 4283b 4284a 4284b 4285a 4285b 4286a 4286b 4287a 4287b 4288a 4288b 4289a 4289b 4290a 4290b 4291a 4291b 4292a 4292b 4293a 4293b 4294a 4294b 4295a 4295b 4296a 4296b 4297a 4297b 4298a

B1 A1 B1 B1 D A1 B1 A1 C A2 D A1 B1 B1 C A1 B1 D A1 A3 C B1 A1 A1 C A1 B1 E B1 D* B1 A3 B1 A1 C B1 A1 B1 C B1 A1 C E A1 B1 E B1 C E E C A3 A1 B1 E A1

xxxPxx#xpx PX#xxxPx xxxPx#P xxxxxxxP#xP P#Pxx P#xxxxxxPx xxxPx#px Pxx#Px xxxxxxxxxxP#px xpx#xxxpx#P xxxxxxPSxx px#xxPx xPx#P xPx#xP xpXSx px#xPx xxPx#P P#pXsX Px#xPx xxxxPx(xx) xxxxxP#px xxxPx#xP P#xxPxx xxPx#xPx xxP#px Px#Px xxxxPS#P xPSx#xP xxxxxxxxpXx#xpx xPx#Px#P xxxxxxPxx#P xxxxxxPx xxxxPx#P(xxx) xpX#xPx xxxxP#px xxPx#P Px#Px xPXS xxxxxxP#px xxxPXS Px#Px xxPXx Px#xxpx P#xxPx xxxpxS#px PXx#P xxxPx#px xxPSx pxSx#P PS#xP xpx#Px xxxxxPx(xx) Px#xPx xxPx#xpx pXSx#px PX#xxxpXx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 447

Index to the scansion of the Heliand 4298b 4299a 4299b 4300a 4300b 4301a 4301b 4302a 4302b 4303a 4303b 4304a 4304b 4305a 4305b 4306a 4306b 4307a 4307b 4308a 4308b 4309a 4309b 4310a 4310b 4311a 4311b 4312a 4312b 4313a 4313b 4314a 4314b 4315a 4315b 4316a 4316b 4317a 4317b 4318a 4318b 4319a 4319b 4320a 4320b 4321a 4321b 4322a 4322b 4323a 4323b 4324a 4324b 4325a 4325b 4326a

B1 B1 B1 A1 A1 D C A1 B1 A1 B1 B1 A1 A1 A1 A1 C A1 C A3 B1 D* C B1 B1 B1 A1 A1 A1 D C D* B1 D* B1 A1 D B1 C A1 C D* C E C A1 B1 A1s B1 D B1 C A1 D* C E

xxpx#xP xxPX#P xxPx#P xPx#xxxxpXx xxxxPx#xPx px#Pxx xxxxpXSx Px#Px xxxxPx#xpx xPx#xxxPx xxxPx#px xxxxxxxPXS Px#Px pxx#Px px#xxPx Px#xpxx xxxxxpx#Px px#xxPx xxpx#Px xxxxxPx xxxPx#xpx xPx#PXS xxxxxxpX#px xxxPx#px xxxxxxxxPx#P xxxxPx#xpx xPx#xPx xpXx#xxxPx Px#Px P#pXSx xxpx#Px px#xPx#pX xxxxxxPx#px Px#xPx#P xxxpxx#P pxx#xxPx PSXx xPx#xP xxxP#px P#xxPx xxpx#Px xPx#P#xpx xxxxpX#Px pxS#xpx xxpx#Px Px#xxPx xxpXx#xP pXS#px xxpxxx#P px#Psx xxpXS#P xxxxP#px P#xPx Px#P#xpx xxxxpX#Px PSxx#P

4326b 4327a 4327b 4328a 4328b 4329a 4329b 4330a 4330b 4331a 4331b 4332a 4332b 4333a 4333b 4334a 4334b 4335a 4335b 4336a 4336b 4337a 4337b 4338a 4338b 4339a 4339b 4340a 4340b 4341a 4341b 4342a 4342b 4343a 4343b 4344a 4344b 4345a 4345b 4346a 4346b 4347a 4347b 4348a 4348b 4349a 4349b 4350a 4350b 4351a 4351b 4352a 4352b 4353a 4353b 4354a

B1 A1 B1 A1 C A1 D A2 B1 E C A1 B1 B1 B1 A1 A1 B1 A1 A1 C A1 B1 D* B1 A3 A1 A1 C A1 B1 B1 A1 A2 C A3 B1 B1 A1 A3 B1 D* B1 A1 C A1 B1 D B1 A1 B1 D* C A1 B1 D*

447 xxxxxxxxPXS Px#xPx xxPx#P Px#xxPx xxxxP#Px Px#xxxpxx px#PS#P Px#pxS xxpxx#P pxSxx#P xxPXx xxPx#xxxxpXx xxxPx#px xPx#px xxxxxPx#xpx xPx#xxxxpXx xxxxxxPx#xPx xxxPx#P Px#Px Px#xPx xxP#px pXSx#Px xxxPxx#px Pxx#xxPxx xxxxxPx#px xxxPx pxx#xPx xxPx#xxPx xxpx#Px P#xPx xxxPx#P xxxP#xxP px#xPx P#xxPS xxpx#Px xxxxxxxxPx xxxPx#P xxPx#P Px#Px xxxxxxPx xxxpx#xP xpx#xPsx xxxxxPx#P xxP#xPx xxP#px pxx#xxPx xxxxPx#P P#PSx xxxxxxPx#P xPx#xxxxPx xxxxxPx#xpx px#xPSx xxPsx PS#xPx xxxxPxx#P xpxx#pXSx

MOSP-Index 1.qxd

5/11/04

448 4354b 4355a 4355b 4356a 4356b 4357a 4357b 4358a 4358b 4359a 4359b 4360a 4360b 4361a 4361b 4362a 4362b 4363a 4363b 4364a 4364b 4365a 4365b 4366a 4366b 4367a 4367b 4368a 4368b 4369a 4369b 4370a 4370b 4371a 4371b 4372a 4372b 4373a 4373b 4374a 4374b 4375a 4375b 4376a 4376b 4377a 4377b 4378a 4378b 4379a 4379b 4380a 4380b 4381a 4381b 4382a

2:49 PM

Page 448

Index to the scansion of the Heliand B1 A1 B1 C C A1 C A1 E B1 C A1 C A1 B1 C C C A1 F B1 C A1 B1 C A1 A1 F E A1 B1 F B1 D* B1 C B1 or E A1 A1 D* B1 B1 B1 A1 C A1 C A3 B1 D* C A1 A1 D* B1 D*

xxxPx#P xP#xxxpXx xxxxPx#px xxxPXx xPSx PXx#xxPx xpXSx Px#Px PSx#px xPx#P xxP#px Px#xxPx xxxxP#Px xPx#xxPx xxxxxPx#xpx xxxxxP#px xPsx xxxpxSx Px#xpxx xNoeas#Px xxxxxxpxx#P xxPXx Px#Px xxPx#P xxxxP#px P#xpxx xxPx#Px xxSodomo#P P#Px#xP P#xxPx xxxxpxx#xxP xxxLoth#Px xxxPx#px Pxx#Pxx xxxPx#P xxxP#Px xxxxPXx#P xP#xPx Px#xpxx xPXx#xxP#px xxxxP#xpx xxxxPx#P xxxxxxPx#xpx Px#xxxxPx xxP#px Px#xpxxx xxxxxxxxP#Px xxxxxxPx xPX#P Px#Px#px xxxP#px px#xxPx pXx#PX Px#xxPx#P xxpx#xpx Px#xPxx

4382b 4383a 4383b 4384a 4384b 4385a 4385b 4386a 4386b 4387a 4387b 4388a 4388b 4389a 4389b 4390a 4390b 4391a 4391b 4392a 4392b 4393a 4393b 4394a 4394b 4395a 4395b 4396a 4396b 4397a 4397b 4398a 4398b 4399a 4399b 4400a 4400b 4401a 4401b 4402a 4402b 4403a 4403b 4404a 4404b 4405a 4405b 4406a 4406b 4407a 4407b 4408a 4408b 4409a 4409b 4410a

B1 A1 B1 E A1 A1 B1 A1 B1 C A1 A1 B1 B1 B1 C B1 A1 B1 A1A1 CA1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1 C B1 A1 A1 B1 D* B1 A1 C D* B1 A1 B1 A1 C A1 C A1 B1 A3 B1 A1 B1 A3 B1 A1

xxPx#P Pxx#xpxx xxxxxpxx#P pxSx#px Px#xPx PXxx#Px xxxxxxxxPx#P pxx#xPx xxxxPx#P xxxPSx Px#Px xPx#xxxxPx xxxxxxPx#P xxPx#px xxpXx#P xxxxxPxx xxPxx#P Px#xxxPx xxxxPx#px px#x(xxx)xxxpxx#Px xxxxxxPxx#Px xPx#xxxpxx#Px xxxpxx#Px xP#xxxxpXx#Px xxxxPx#Px px#xxxpxx#Px xxxxxxpxx#Px xPx#xxPx#Px xxxxxPx#pxx xPx#xPx xxxxxxxxpx#Px xxxPx#P Px#xxPx Px#xPx xxxpxx#P xPx#xPxx xxxxpxx#px Px#xxxPx xxxxxxxxpx#Px Pxx#xPSx xxxxxxxpx#xpx P#xxPx(xxx) xxxxxPx#P xPx#xxxxpxx xxxxxxxP#px Px#Px xxxxP#Px xPx#xxxxpxx xxxxxxxPx#xP xxxxxxPx xxxxPx#P xPx#xxxxpXx xxxxxPX#P xxxxPx(xxx) xxxPxx#px Px#xPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 449

Index to the scansion of the Heliand 4410b 4411a 4411b 4412a 4412b 4413a 4413b 4414a 4414b 4415a 4415b 4416a 4416b 4417a 4417b 4418a 4418b 4419a 4419b 4420a 4420b 4421a 4421b 4422a 4422b 4423a 4423b 4424a 4424b 4425a 4425b 4426a 4426b 4427a 4427b 4428a 4428b 4429a 4429b 4430a 4430b 4431a 4431b 4432a 4432b 4433a 4433b 4434a 4434b 4435a 4435b 4436a 4436b 4437a 4437b 4438a

C D*A1 A3A1 C A1 A1A1 A3A1 D* A3A1 B1D CA1 A1 C A1 B1 A1A1 B3A1 D* C D* C B1 D A1 C A3 B1 A1 B1 D B1 A1 B1 E B1 A1 B1 A1 C A3 B1 A1 B1 D C A1 B1 A1 C A1 B1 C A1 D* C A1

xpx#Px xPx#xxPXx#Px xxxxxxxxpxx#Px xxxPSx Px#Px px#xxxxPx#pxx xxxxxxxxpxx#xPx xPx#xPxx xxxxxPx#Px xPx#xxpXsXx xxxxxxPxx#Px Px#xxxPx xxxP#Px Px#xxxPX xxpXx#P Px#xxxPx#Px xxxxxxxxP#xPx xP#xxPxS xxxP#px(xxx) px#xxPxx xxP#Px xxxpxx#P px#Psx PXx#Px xpXSx xxxxPx xxPx#xxP Px#xPx xxxxPx#P P#PXS xxPx#P xxPx#xxxxxPx xxxPx#P xpxSx#xpx xxxpx#xP Px#Px xxxxxxPx#P Px#xPx xxxP#PX xxxxPx xxxxPx#px px#xPx xxxxxxxP#xpx px#PX#P(xxx) xxxxxxpx#px Px#xxxpxx xxxxPx#P pxx#Px xxxxxxxpx#Px px#xxxxpXx xxxxPX#P xxxPXx(xxx) PxSx Px#xPXx xxxxPsx pxx#xPx

4438b 4439a 4439b 4440a 4440b 4441a 4441b 4442a 4442b 4443a 4443b 4444a 4444b 4445a 4445b 4446a 4446b 4447a 4447b 4448a 4448b 4449a 4449b 4450a 4450b 4451a 4451b 4452a 4452b 4453a 4453b 4454a 4454b 4455a 4455b 4456a 4456b 4457a 4457b 4458a 4458b 4459a 4459b 4460a 4460b 4461a 4461b 4462a 4462b 4463a 4463b 4464a 4464b 4465a 4465b 4466a

C A1 B1 A1 B1 D* C B1 A1 D* B1 A3 B1 A1 B1 B1 A1 B1 A1 D C D A1 B1 C D* A1 B1 A1 B1 A1 C B1 A1 B1 A1 B1 A1 C C A1 F C A1 C B1 C A1 B1 D C A1 A1 F C C

449 xxxxxxxxpx#Px xPx#xxxxPx xxxxxxxPx#xpx xPxx#xxxxxpxx xxxxxxPX#P xPx#px#Px xxxxP#px xxxPx#P Px#Px Px#pXsx xxxPx#xpx xxxxPx xxxpx#xP xPx#xxxpxx xxxxpxx#P xxPx#P PXSx xxPx#px Px#xPx px#PxS xxP#px P#pXsX xPxx#Px xxPsx#P xxP#Px xpxx#px#Px Pxx#Px xxxxxxPx#P Px#Px xxxxxpXx#xP Px#Px xxP#px xxxxxpxx#P Px#Px xxxxxxPx#P xPx#xxxPx xxxxxPxx#P Px#Px xxpx#Px(xx) xxxxP#P#x Px#pxx Giudeono#Px xxxxxxpx#Px px#xxxPx xxxP#Px xxxxPx#px xxxpXSx Px#xPx xxxPx#xpx px#PSx xxxpx#px PS#xPx PXSx Iudeono#Psx xxxxxxpx#Px xxPsx

MOSP-Index 1.qxd

5/11/04

450 4466b 4467a 4467b 4468a 4468b 4469a 4469b 4470a 4470b 4471a 4471b 4472a 4472b 4473a 4473b 4474a 4474b 4475a 4475b 4476a 4476b 4477a 4477b 4478a 4478b 4479a 4479b 4480a 4480b 4481a 4481b 4482a 4482b 4483a 4483b 4484a 4484b 4485a 4485b 4486a 4486b 4487a 4487b 4488a 4488b 4489a 4489b 4490a 4490b 4491a 4491b 4492a 4492b 4493a 4493b 4494a

2:49 PM

Page 450

Index to the scansion of the Heliand A1 C C A3 A1 A2 F A1 C A3 A1 B1 A1 A1 C A1 B1 A1 B1 A1 F C A1 F A1 A1 B1 F B1 A1 B1 A1 C B1 B1 A1 B1 A1 C A3 A1 A1 B1 B1 A1 A1 B1 A1 B1 A1 C A2 C A3 A1 A1

Px#pxx xP#px xxxPXx xxxxpxx Px#Px Px#pxS xCaiphas#P Px#xxPx xxxxxxP#px xxxxxPx Px#Px xxxxxxxxxPxx#px Px#xPx xxxxPx#pxx xxxxxpXSx pxx#xPx xxxxpxsx#px xPx#xPx xxPx#px Px#xpxx xxP#Iudeono xxxPsx P#xxPx xP#xxxIudas#P Pxx#Px P#xxpXx xxpxx#P Iudeono#Psx xxxxxPx#P Px#Px xxxxxPx#P(xx) Px#xPx xxxxxP#px xxP#xxxxP xxxpxx#px xxPx#xPx xxxxxPx#P(xxx) xP#xPx xxxP#px xxxxxxPx Px#Px Pxx#Px xxxxxxPsx#P xxPx#P pxXSx Px#xPx xxxxxxPx#xP pXx#Px xxxxxPx#xP Px#xxxxxxpxx xxP#px pxx#xxPS xxxxxxxP#px xxxxxxxPx Pxx#Px PXx#Px

4494b 4495a 4495b 4496a 4496b 4497a 4497b 4498a 4498b 4499a 4499b 4500a 4500b 4501a 4501b 4502a 4502b 4503a 4503b 4504a 4504b 4505a 4505b 4506a 4506b 4507a 4507b 4508a 4508b 4509a 4509b 4510a 4510b 4511a 4511b 4512a 4512b 4513a 4513b 4514a 4514b 4515a 4515b 4516a 4516b 4517a 4517b 4517’a 4517’b 4518a 4518b 4519a 4519b 4520a 4520b 4521a

B1 D C A1 C D* A1 A1 B1 D* A1 D* B1 A1 B1 A1 B1 A1 C A1 B1 A1 B1 A1 C D* B1 F B1 A1 B1 A1 B1 A1 B1 C B1 A3 B1 A1 B1 C A1 F B1 A1 A1 B1 A1 A1 B1 A1 C A3 B1 A1

xxxxpxS#px P#PX#P xxxxpX#Px xpx#xxPx xxxxxxpx#Px xpx#xpXSx xxxxxxpxx#Px Px#Px xxxxxPx#xP xxxPx#Pxx Px#Px Px#xPSx xxxPx#px Pxx#Px xPx#P Px#xPx xxPx#xP PX#xxPx xxpx#px Px#xPx xxxxxPxx#P xPx#xxPx xxxxPxxx#P Px#xxPx xxxxxxpx#Px Pxx#xPSx xxxPx#P Simon#Petrus xxxxxPx#P(xx) P#xxPx xxxxPx#px xxxxPxx#Px xxxxxxPx#xpx PX#xxPx xxxPx#xpx(xx) xxPXx xxxxPx#px xxxxPx xxxxxxPx#P P#xPx xxxxxxxxP#xP xpXSx px#xxxPx Simon#Petruse xxxxPx#xP(xx) P#xxPx Px#xxPx xxxxPx#xpx Px#Px Px#xpXx xxxxxxPx#P Px#Px xxpXSx xxxPx xxxPx#px xpx#xxxPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 451

Index to the scansion of the Heliand 4521b 4522a 4522b 4523a 4523b 4524a 4524b 4525a 4525b 4526a 4526b 4527a 4527b 4528a 4528b 4529a 4529b 4530a 4530b 4531a 4531b 4532a 4532b 4533a 4533b 4534a 4534b 4535a 4535b 4536a 4536b 4537a 4537b 4538a 4538b 4539a 4539b 4540a 4540b 4541a 4541b 4542a 4542b 4543a 4543b 4544a 4544b 4545a 4545b 4546a 4546b 4547a 4547b 4548a 4548b 4549a

A1 B1 A1 A1 B1 A1 B1 D* A1s A3 B1 E C A1 E D* B1 C A1 A3 A1 A1 B1 F B1 C C E C B1 A1 A1 B1 A3 B1 A3 C A1 B1 A1 B1 A1 B1 A1 B1 A1 C A1 B1 F A1 A1 E* E C A1

Pxx#Px xxPXsx Px#pxx Px#xxPx xxxxxxPx#px Px#xxPx xxxxxxxPx#P pxx#xxPXx pxS#px xxxxxxPx xxxxPx#P xxxxxxxPSx#P xxP#px Px#xPx PXx#P Px#xPxx xxxxxxxxPx#P xxxPsx Px#Px xxxxPx xPxx#Px Px#xxxPx xxxPx#P xpx#Hierusalem xxxPx#px(xx) xxP#Px xxP#px pXSx#xP xxxxxxP#px xxPx#px Px#pxx P#xPx xxxPx#px xxxxxPx xxxxPx#xpx xxxxxPx xxpx#Px Px#Px xxxPx#px xxpxXx#xxPx xxxxxxPS#P Px#pxx xxxPx#P pxx#pxx xxPx#px Psx#Px xxxPsx Px#xxxxPx xxxP#xxP xxHierusalem Pxx#Px PS#xPx Px#P#xxP PSx#P xxxxP#Px xpxxx#xxPx

4549b 4550a 4550b 4551a 4551b 4552a 4552b 4553a 4553b 4554a 4554b 4555a 4555b 4556a 4556b 4557a 4557b 4558a 4558b 4559a 4559b 4560a 4560b 4561a 4561b 4562a 4562b 4563a 4563b 4564a 4564b 4565a 4565b 4566a 4566b 4567a 4567b 4568a 4568b 4569a 4569b 4570a 4570b 4571a 4571b 4572a 4572b 4573a 4573b 4574a 4574b 4575a 4575b 4576a 4576b 4577a

C A1 C C A1 D* F D* C C E C B1 A1 C C A1 A1 A1 C A1 A1 A1 B1 A1 A1 F A1 B1 A1 B1 B1 A1 A1 A1 A1 B1 A1A1 B1 A1 B1 A1 C B1 B1 D* C D C C B1 A1 B1 B1 A1 A3

451 xxpx#px Px#Px xxP#px xxxPSx Px#Px xPx#px#xpx xxIudeono#P P#xxPsx xPsx xxxxxxxxPXx pxSX#P xxxpx#Px xxxxPx#xpx pxx#Px xxxPXx xxxPsx Px#Px xPx#xxxPx Px#xxPx xxpxSx Px#Px Px#xxxxxPx P#xxPx(xx) xxpx#xP Px#Px Pxx#Px Iudeono#Px Px#xxxPx xxxxxPxx#P Px#Px xxxxxxpXx#xP xPx#P Px#xPx Px#xpxx xxxPx#Px pxx#Px xxxPx#P Px#xxPXSx xxxxxxpxx#P px#xxxxPx xxxxxxPx#P Px#Px xxxxxpx#Px xxxxPx#px xxxxxxPx#P xPx#xxxPxx xxxpx#Px(xx) xP#pxx#P xxxxPSx xxpXsx xxxxxxPx#P xPx#xxPx xxxxxxPx#P xxxxxpXx#P Pxx#Px xxxxPx

MOSP-Index 1.qxd

5/11/04

452 4577b 4578a 4578b 4579a 4579b 4580a 4580b 4581a 4581b 4582a 4582b 4583a 4583b 4584a 4584b 4585a 4585b 4586a 4586b 4587a 4587b 4588a 4588b 4589a 4589b 4590a 4590b 4591a 4591b 4592a 4592b 4593a 4593b 4594a 4594b 4595a 4595b 4596a 4596b 4597a 4597b 4598a 4598b 4599a 4599b 4600a 4600b 4601a 4601b 4602a 4602b 4603a 4603b 4604a 4604b 4605a

2:49 PM

Page 452

Index to the scansion of the Heliand F A1 C A1 B1 A1 B1 A1 B1 C A1 A1 B1 A1 A1 A1 C D A1 B1 A1 E C A1 B1 A1s D B1 A1 E B1 B1 D* C A1 A1 B1 A3 A1 C E F B1 F C C A1 A1 A1 A1 B1 A1 A1 A1 B1 A3

xxxPx#Iudeono xPx#xxpXx xxxxxxxP#px Px#Px xxxxxPx#xP Px#Px xxxxxPx#P Px#xPx xxxxPx#xpx xxxxPXx Px#Px xP#xxxPx xxxxxPxx#P pxx#pxx xxxxpx#xPx PXx#xxxxPx xxxP#px px#Pxx PXSx xxxxxPx#xpx xPxx#Px PXx#px xxP#px Px#xxxxPx xxxxxPx#P PS#px xpx#Pxx xxxxxxpXx#xP Px#Px PXx#px xxxxxPx#P xPx#xxxP xxxxPx#Pxxx xxxPSx Px#xpXx PxSx xxxxPx#xpx xxxxxPx Px#xxPx xxxxPxx PSx#px Simon#Petrus xxxxxPx#px xIohanne#xxPx xxxxpx#Px xP#px Px#PX P#xPx xxxxxxxxPxx#Px Px#Px xxxxPx#P pxx#xxPx xxxxxxPx#xPx Px#xPx xxxxxPxx#P xxxxxPx

4605b 4606a 4606b 4607a 4607b 4608a 4608b 4609a 4609b 4610a 4610b 4611a 4611b 4612a 4612b 4613a 4613b 4614a 4614b 4615a 4615b 4616a 4616b 4617a 4617b 4618a 4618b 4619a 4619b 4620a 4620b 4621a 4621b 4622a 4622b 4623a 4623b 4624a 4624b 4625a 4625b 4626a 4626b 4627a 4627b 4628a 4628b 4629a 4629b 4630a 4630b 4631a 4631b 4632a 4632b 4633a

B1 B1 A1 B1 or E B1 A1 C E C A1 B1 D B1 A1 B1 A1 B1 A1 C F B1 A1 B1 A1 A1 A1 B1 A1 A1 A1 C B1 B1 A2 D C A1 F A1 A1 B1 A1 B1 B1 A1 C E F B1 A1 B1 A1 D B1 B1 D*

xxxP#xpx(xx) xxxPx#px pXx#PX xxxxxPXx#P xxxxpxS#px PX#xpXx xxxxxP#px PXx#P xxxxxxxP#px xxPx#xxxPx xxxPxS px#Px#px xxxxpxx#xP PXx#xPx xxxxPx#P Px#Px xxxPx#P xPx#xxPx xxxxxxPsx Iudase#xP xxxxxPx#P Px#xxxxPx xxxxPx#P px#xxxxPx px#xxPx(xx) P#xxP#x xxxxPx#P Px#Px xP#xxPx xPx#xxxPx xxxPsx xP#xP xxxxPx#xP xxP#xxxxpx#P px#P#xpx xxxPsx Px#Px xxxSatanas Px#xPx Px#xxxPx xxxxxPx#px xP#xxxxPx xxxxPx#P xxxxxxxpxx#P Px#Px xxxxxP#px PSx#P Iudas#Px xxxxPx#px px#xxPx xxxPx#P Px#xPx px#Pxx xxxxxPx#P xxxPxx#P PX#P#xxP

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 453

Index to the scansion of the Heliand 4633b 4634a 4634b 4635a 4635b 4636a 4636b 4637a 4637b 4638a 4638b 4639a 4639b 4640a 4640b 4641a 4641b 4642a 4642b 4643a 4643b 4644a 4644b 4645a 4645b 4646a 4646b 4647a 4647b 4648a 4648b 4649a 4649b 4650a 4650b 4651a 4651b 4652a 4652b 4653a 4653b 4654a 4654b 4655a 4655b 4656a 4656b 4657a 4657b 4658a 4658b 4659a 4659b 4660a 4660b 4661a

A1 D* B1 D* D A3 B1 A1 C A1 C B1 B1 A1 B1 A1 C D* C B1 A1 A3 B1 A1 B1 C A1 A1 A1 B1 A1 A1 C D* B1 A1 A1 A3 B1 A1 B1 A1 B1 B1 A1 C A1 A3 A1 A2 C F B1 A1 B1 D*

Pxx#Px Pxxx#pXsX xxPx#P P#xxxxxPxx xxpx#Pxx xxxPx xxP#xP pX#xxPx xxxP#px xPx#xxPx(xx) xxxxPsx xxxP#xpx xxxxxPx#px px#xxPx xxxPx#P px#xxPx xxxxpx#Px Px#xxxPsx xP#Px xxpxx#P xPx#xPx xxxxPx xxxxxPx#P Px#xxpxx xxPx#P xxxxxxxxPxx Pxx#Px Px#xxxxPx Px#pxx xxPxS Px#Px px#xxxxpXx xxpx#Px P#xxxxPXS xxxxxxPx#xP Px#xPx xPx#xPx xxxxxPx xxxxPXsx Px#Px xxPx#px Px#xxxxxPx xxPx#P xxPXS Px#xPx xxxPXx Pxx#Px xxxxPx xxxxPx#PX PX#pxS xxxxpx#Px Satanas#Px xxxxxxPxx#px Px#Px xxxPx#xpx px#xxPxS

4661b 4662a 4662b 4663a 4663b 4664a 4664b 4665a 4665b 4666a 4666b 4667a 4667b 4668a 4668b 4669a 4669b 4670a 4670b 4671a 4671b 4672a 4672b 4673a 4673b 4674a 4674b 4675a 4675b 4676a 4676b 4677a 4677b 4678a 4678b 4679a 4679b 4680a 4680b 4681a 4681b 4682a 4682b 4683a 4683b 4684a 4684b 4685a 4685b 4686a 4686b 4687a 4687b 4688a 4688b 4689a

C D* A1 D* B1 B1 A1 D* C B1 A1 A3 A1 C B1 A1 C A1 C A1 C A1 B1 A1 F A1 E A1 B1 A1 B1 A3 A1 C C C A1 C A1 A1 B1 B1 A1 A1 B1 B1 A1 A1 B1 B1 C C C A2 B1 A3

453 xxxxxxpx#Px xxxpx#xPsx P#xPx xPXx#xxxxxPXx xxxxxxPx#P xxxxPx#P P#xxPx Px#xxxxxxPsx xxxxxxP#px xxxxPx#P xPx#xPx xxxxxPx xPx#Px xxpXSx xxxPx#P Px#xPx xxxP#px Px#xPx xxP#px Px#xPx xxP#px Px#xxxxPx xxxxPx#P Px#xPx Simon#Petrus#P px#xxPx PSx#P xPx#xxPx xxxxpxx#P(xx) xPx#xxxPx xxPx#xP xxxpxx pxx#Px xxpx#Px xxP#Px xxxxxPXx Px#xPx xxxxPXx Px#Px xxPx#xPx xxxPx#px xxxxPx#xP Px#Px Px#xxxPx xxxxPx#P xxPx#P Px#Px P#xxPx xxxxpx#xxP xPx#P xxP#PX xxPXx xxxxxP#px px#xxPS xxxxxPx#xpx xxxxPx(xx)

MOSP-Index 1.qxd

5/11/04

454 4689b 4690a 4690b 4691a 4691b 4692a 4692b 4693a 4693b 4694a 4694b 4695a 4695b 4696a 4696b 4697a 4697b 4698a 4698b 4699a 4699b 4700a 4700b 4701a 4701b 4702a 4702b 4703a 4703b 4704a 4704b 4705a 4705b 4706a 4706b 4707a 4707b 4708a 4708b 4709a 4709b 4710a 4710b 4711a 4711b 4712a 4712b 4713a 4713b 4714a 4714b 4715a 4715b 4716a 4716b 4717a

2:49 PM

Page 454

Index to the scansion of the Heliand A1 A1 B1 A1 B1 A2 B1 B1 A1 C B1 A2 B1 A3 A1 B1 A1 D* C C A1 F B1 B1 A1 B1 A1 D C A3 A1 A1 B1 B1 C A1 B1 C E A3 C A1 C A1 B1 A1 B1 B1 B1 A1 A1 E B1 A1 A1 C

Pxx#Pxx Pxx#Px xxxPx#px Px#Px xxxxxxxxxPx#P xxPx#xPS xxxxPx#P xxxxPx#xP Px#xPx xpxSx xxxxxxxPx#xP xxxP#xxPS xxxxP#xpx xxxxpXx(xx) xPx#Px xxpx#xxP Px#Px Px#PSx xxxxxxP#px xxxxPxx Px#Px Px#xxxIuðeon xxxxxxPx#P xxxxxPx#xP Px#Px xxxxxPx#xP PX#Px P#pXsX xxxxxxxxpx#Px xxxxPx Px#xPx xPx#xxPx xxxxPxx#P xpXx#xpx xxpx#Px Px#Px xxxPx#P xpXSx PXx#P xxxxxPx xxxpx#Px px#xxxPx xxxpx#Px Px#xPx xxxxxP#xP Psx#Px xxPx#P xP#xxxP xxxxxxxpXx#xP xP#xxxPx xxxpxx#Px PXx#P xxxxxPx#px Px#xxxPx Pxx#Px xxPXx

4717b 4718a 4718b 4719a 4719b 4720a 4720b 4721a 4721b 4722a 4722b 4723a 4723b 4724a 4724b 4725a 4725b 4726a 4726b 4727a 4727b 4728a 4728b 4729a 4729b 4730a 4730b 4731a 4731b 4732a 4732b 4733a 4733b 4734a 4734b 4735a 4735b 4736a 4736b 4737a 4737b 4738a 4738b 4739a 4739b 4740a 4740b 4741a 4741b 4742a 4742b 4743a 4743b 4744a 4744b 4745a

A1 A1 B1 F B1 A1 F E B1 D* A1 A3 C A1 F A1 A1 A1 C A1 C A1 B1 D* B1 E C A2 B1 A1 B1 A1 C C A1 B1 A1 F F E B1 A1 C A1 C C A1 C A1 A1 C D* A1 B1 D D*

Px#Px PXSx xxxxxxPx#xP Oliueti-S xxxP#xpx Px#xxPx xxxIudas#P pxSx#P xxxxxxpxx#xP xPx#xpx#px Px#Px xxxxPx(xx) xxxxP#px xPX#xxxPX xxxxIuðeon#xxPx Px#xpxx xxxxxPx#Px xpX#xxPx xxxxxP#px Px#Px xxxxP#px xPx#xxxxPx xxxPx#px Px#xPXsx xxPx#xpx xxPSx#xP xxxxxxxP#px Px#xPS xxxxxPx#px pxx#Px xxxxPx#P Px#xxxxPx xxxxpx#Px xxPsx Px#Px xxP#xP Px#Px Iacobe#xxIohannese xxxxPx#Petruse PSx#px xxxxxPx#px Px#Px xxxxpx#px xPx#Px xpx#Px xxP#Px Px#Px xxxxxPXxx P#xPx Pxx#Px xxxpXsx xPx#xPsx P#xPx xxxxPx#xP px#Pxx Px#xPsx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 455

Index to the scansion of the Heliand 4745b 4746a 4746b 4747a 4747b 4748a 4748b 4749a 4749b 4750a 4750b 4751a 4751b 4752a 4752b 4753a 4753b 4754a 4754b 4755a 4755b 4756a 4756b 4757a 4757b 4758a 4758b 4759a 4759b 4760a 4760b 4761a 4761b 4762a 4762b 4763a 4763b 4764a 4764b 4765a 4765b 4766a 4766b 4767a 4767b 4768a 4768b 4769a 4769b 4770a 4770b 4771a 4771b 4772a 4772b 4773a

A1 A1 D A1 E A1 C C A1 A1 A1 D* C A1 B1 C D* A3 C D* B1 A1 B1 A1 B1 B1 A1 D* E A1 B1 E B1 D* B1 A1 B1 A3 B1 A1 A1 D* B1 A1 B1 A1 B1 A3 B1 A1 C D* C C A1 C

pXx#PX PS#xPx px#pxSx Px#Px PSx#P PXSx xxxxpx#Px xxxPXx P#xPx xP#xxPx Px#xxPx P#xPS#P xxP#px Px#xPx xxxPx#P xxxpx#Px xP#xxxPsx xxxPx xPsx xP#xpx#Px xxPx#P Psx#Px xxxxP#xpx xPx#xxxPx xxxxxPxx#P xP#xxP PXx#Px Px#pXsx PXx#P PX#xxPx xxPx#xP(xx) PSx#xpx xxxxPx#px Px#Psx xPx#P xPXx#xPx xxxxPx#P xxxxxxPx xxxxxpx#xP Px#xxxxPx Px#P#x Px#Psx xxxPx#P Px#xPx xPx#P Px#Px xxxxP#xxP xxxxxPx xxxxPxx#P Px#xxxxPx xxxP#px Px#PXx xxP#px xxxxxPxx Px#Px xxxPsx

4773b 4774a 4774b 4775a 4775b 4776a 4776b 4777a 4777b 4778a 4778b 4779a 4779b 4780a 4780b 4781a 4781b 4782a 4782b 4783a 4783b 4784a 4784b 4785a 4785b 4786a 4786b 4787a 4787b 4788a 4788b 4789a 4789b 4790a 4790b 4791a 4791b 4792a 4792b 4793a 4793b 4794a 4794b 4795a 4795b 4796a 4796b 4797a 4797b 4798a 4798b 4799a 4799b 4800a 4800b 4801a

A1 B1 A1 A1 B1 A1 A1 A3 B1 D A1 B1 C A1 B1 B1 C A1 A1 D* A1 A1 B1 A1 B1 C A1 C C A1 B1 A1 E A1 D A1 C A1 C A1 B1 A1 B1 A1 B1 A1 D* A1 C A1 D* E* C B1 D E

455 Px#xpxxx xxxPx#P Pxx#Px xpx#xxxPxx xxxxPxx#P Px#xPX xxPx#Px xxxxxPx(xx) xxxpx#xP px#Px#P xP#xxPx xxxxPx#P xxP#px xPxx#Px xxxxxxPx#px xP#xpx xpx#Px P#xpXx xP#xxPx px#xxPsx P#xxxPx Px#xpXx xxPx#P xxpx#xPx xxxPx#px xxxxxP#px PXSx xxxP#Px xpx#Px Px#Px xxxxpx#xP Pxx#Px px#Px#P Px#xpxx xpx#Pxx Pxx#xxPx xxxxxpx#Px P#xPx xxxpx#Px PX#xxPx xxxpx#xP(xx) Px#Px xxxxxPx#P PSx#pxx xxPx#P Px#Px xP#xxxP#px Px#xxPx xxPXx Px#xPx P#xxP#px Px#Px#xpx xxxPsX xxPx#P px#Pxx xxxpxSx#px

MOSP-Index 1.qxd

5/11/04

456 4801b 4802a 4802b 4803a 4803b 4804a 4804b 4805a 4805b 4806a 4806b 4807a 4807b 4808a 4808b 4809a 4809b 4810a 4810b 4811a 4811b 4812a 4812b 4813a 4813b 4814a 4814b 4815a 4815b 4816a 4816b 4817a 4817b 4818a 4818b 4819a 4819b 4820a 4820b 4821a 4821b 4822a 4822b 4823a 4823b 4824a 4824b 4825a 4825b 4826a 4826b 4827a 4827b 4828a 4828b 4829a

2:49 PM

Page 456

Index to the scansion of the Heliand C D* A1 C E D* A1 A1 B1 A1 B1 E* A1 A1 C C A1 D* F E F D* B1 D* A1 E C A1 F B1 A1 A3 C B1 B1 D* B1 A1 C B1 A1 A1 C A1 B1 A1 D A3 A1 F B1 A1 B1 A1 B1 F

xxP#px pxx#PXx Px#pxx xxPSx PXx#P P#xxxxxxxPxx Px#xPx Px#xxxPx(xx) xxxPx#px px#xPx xxxPx#px Px#Px#xP xPx#Px pxx#xxxxPx xxxxxxxpx#px xxxP#Px Px#xpxx Px#PXsX Px#xIudas PSx#P Iudeon#xxpx PXx#Psx xxP#xxP Px#xPsx Px#xPx PXx#xP xxxxxP#Px px#xPx xpx#xxJudas#P xxxPx#P Px#Px xxxxxxPx xxxP#Px xxPx#xpx xxxxxxPx#P Px#Px#P xxxxxxPX#P(xx) Px#xxxxPx xxP#Px xxxPx#px Px#Px Px#xxxxxxPx xxxxxP#px Px#xxxPx xxPx#px xxPx#xPx px#Pxx#P xxxxPx pxx#Px P#P#Iudeono xxxxPxx#P Px#Px xpxxxsx Pxx#Px xxxxPS#P Iudas#xPx

4829b 4830a 4830b 4831a 4831b 4832a 4832b 4833a 4833b 4834a 4834b 4835a 4835b 4836a 4836b 4837a 4837b 4838a 4838b 4839a 4839b 4840a 4840b 4841a 4841b 4842a 4842b 4843a 4843b 4844a 4844b 4845a 4845b 4846a 4846b 4847a 4847b 4848a 4848b 4849a 4849b 4850a 4850b 4851a 4851b 4852a 4852b 4853a 4853b 4854a 4854b 4855a 4855b 4856a 4856b 4857a

C A1 A1 D* C A1 B1 A1 A1 A1 B1 A1 B1 D* A1 A1 F A1 C C B1 A3 A1 E C A1 B1 B1 A1 A1 C F C A1 B1 B1 F F B1 or E A1 B1 A3 F A3 C D* A1 E C A1 B1 A1 D D* A1 A1

xxxxxpx#Px P#xxPx xxxPx#Px Px#xxPXx xxxpx#Px Px#xxxxpxx xxxxxPx#xP xpxx#xxxPx Px#Px PX#xxxpXx xxxxxxxPx#P Pxx#xxPx xxxxxxxxPx#xP xxPx#xxxxxPx#P xxxxxxxPx#Px xPx#xxxPx xxxPx#Iudeono Px#xxxxxpxx xxxxxxP#px xxpx#Px xxxxxPx#P xxxxxxPx Px#Px xPSx#xP xxxxP#Px Px#pxxx xxxxxxpxx#xpx xxxxPX#P xxxPx#Px xPx#Px xxxP#px Iudeo#Px xxxxpx#px Px#Px xxxxxPx#px xxxxPx#xpx xxxGalileo#Px xNazareth-S xxxxPXx#P Px#xPx xxxPx#P xxxxxPx P#Iudeono xxxxpxx xxxxP#Px Px#Px#P Px#xPx pXSx#xpx xxxxP#px xPx#xPx xxxxPxx#P xPx#xxxxxPx px#Pxx Px#PxS xPxx#Px Px#xPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 457

Index to the scansion of the Heliand 4857b 4858a 4858b 4859a 4859b 4860a 4860b 4861a 4861b 4862a 4862b 4863a 4863b 4864a 4864b 4865a 4865b 4866a 4866b 4867a 4867b 4868a 4868b 4869a 4869b 4870a 4870b 4871a 4871b 4872a 4872b 4873a 4873b 4874a 4874b 4875a 4875b 4876a 4876b 4877a 4877b 4878a 4878b 4879a 4879b 4880a 4880b 4881a 4881b 4882a 4882b 4883a 4883b 4884a 4884b 4885a

B1 or E A1 B1 C A1 A1 A1 A1 A1 C A1 A1 B1 C A1 A1 B1 D F D* C A1 B1 A1 B1 B1 A1 A1 C A1 B1 A1 A1 A1 A1 F A1 B1 A1 A1 B1 C A1 E D A1 B1 D* B1 B1 C F C A1 C B1

xxxxxPXx#P pxx#xPx xxPx#P xxPXx Pxx#Px xxxxxpXx#Px xxxxxPxx#Px Px#xxxPx(xxx) PX#P#x xxxxxpx#Px Px#Px Px#Px xxxxxP#xP xxxxxxxxPxx Px#Px pXx#Px xxPx#P P#Psx Simon#Petrus P#xxPx#px xxxxxxxP#px xP#xxxxxPx xxxPx#P Px#Px xxxPx#P xxPS#px xxPx#Px P#xxPx xxxxxpx#Px P#xxPx xxxP#xP P#xPx P#xxxPx xxxpXx#PX Px#Px xxMalchus#P Px#Px xxPxx#P Px#xPx xP#xxxxPx xxxxPx#P xxxpxSx P#xxPx pxSx#P P#Px#P P#xPx xxxxPx#P xpXx#xxPXx xxxP#xP xxxxxPx#px xxxP#px Px#xSimon#Petruse xxxxP#px P#xPx xxxxxpx#Px(xx) xxxpxx#xP

4885b 4886a 4886b 4887a 4887b 4888a 4888b 4889a 4889b 4890a 4890b 4891a 4891b 4892a 4892b 4893a 4893b 4894a 4894b 4895a 4895b 4896a 4896b 4897a 4897b 4898a 4898b 4899a 4899b 4900a 4900b 4901a 4901b 4902a 4902b 4903a 4903b 4904a 4904b 4905a 4905b 4906a 4906b 4907a 4907b 4908a 4908b 4909a 4909b 4910a 4910b 4911a 4911b 4912a 4912b 4913a

A1 A1 E E C B1 A1 A1 B1 A1 C A1 B1 A1 B1 E A1 B1 B1 A1 B1 A1 B1 D* A1 B1 A1 D* B1 A1 A1 A1 A1 A1 B1 B1 C B1 B1 B1 A1 A3 B1 A1 C A1 B1 A2 A1 A1 B1 B1 B1 B1 B1 D*

457 Psx#Px xpxx#xxxPx PXx#P PXx#px xpXSx xxxxxxxPx#px pxx#Px Px#xPx xxxxxxPXsx P#xPx xxxxxxxxpx#px Px#xPx xxxxP#xxP Px#Px xxxxPX#P pxSx#px xPx#xPx xxxpxx#px xxxxxPx#P Px#Px xxxxPx#P Px#xxxxPx xxxxPx#P Px#Psx xxPx#Px xPx#xxP Px#Px Px#xxPxx xxxxPx#xpx P#xPx xxxxxxPx#Px Px#xPx P#xPx PXSx xxxxPx#P xPx#px xxxxP#px xxPx#px xxxPx#px(xx) xxxPx#P Px#Px xxxxxPx xxxxxxxPx#P xxxPx#Px xxxP#px PS#Px xxPx#P PS#px#P xxxxxxPx#PX Px#xxxxPx xxxxxxxxPx#P xPx#P xxxPx#P xxxxPx#px xxxxPx#px xPx#Psx

MOSP-Index 1.qxd

5/11/04

458 4913b 4914a 4914b 4915a 4915b 4916a 4916b 4917a 4917b 4918a 4918b 4919a 4919b 4920a 4920b 4921a 4921b 4922a 4922b 4923a 4923b 4924a 4924b 4925a 4925b 4926a 4926b 4927a 4927b 4928a 4928b 4929a 4929b 4930a 4930b 4931a 4931b 4932a 4932b 4933a 4933b 4934a 4934b 4935a 4935b 4936a 4936b 4937a 4937b 4938a 4938b 4939a 4939b 4940a 4940b 4941a

2:49 PM

Page 458

Index to the scansion of the Heliand F D* A1 E A1 D* A1 D* A1 A1 C A1 D A1 C B1 A1 A1 C B1 B1 B1 A1 A3 F A1 B1 C A1 A1 B1 B1 C C A1 E B1 A1 B1 A3 C A1 B1 E B1 A1 A1 F C D* B1 F C A1 C B1

px#Iudeono px#xxxxpx#px Px#Px emended: PXxx#P Px#xxPx Px#Px#P Px#xPx Px#pxSx Px#xPx Px#xpXx xxxxxxxpXSx P#xxpXx PSXx xPXx#xxPx xxxxxpx#px xxPx#P Px#Px px#xPx xpXSx xxxPx#px xxxxP#xxP xxxxxPXS Px#Px xxxxxPx Px#P#Iudeono xPx#xxxPx xxxxPxx#P xpxSx Px#Px Px#xpXx xPX#xxpx xxxPx#xP xxP#px xxxxpxsx Px#xPx PXx#xpx xxxxxxPx#P xPx#xpxx xxxxxPx#xP xxxxxxxxxPx xxxP#px Px#xPx xxxxPx#xpx Psxx#P xxxxxPx#P xxxPx#xxxPx xxxxxpxx#Px Iohannes#xxPetrus xpx#Px Pxx#Pxx xxpxS#px xxPx#Iudeon xxpx#Px xxxxPxx#Px xxxpx#Px xxxPx#xP

4941b 4942a 4942b 4943a 4943b 4944a 4944b 4945a 4945b 4946a 4946b 4947a 4947b 4948a 4948b 4949a 4949b 4950a 4950b 4951a 4951b 4952a 4952b 4953a 4953b 4954a 4954b 4955a 4955b 4956a 4956b 4957a 4957b 4958a 4958b 4959a 4959b 4960a 4960b 4961a 4961b 4962a 4962b 4963a 4963b 4964a 4964b 4965a 4965b 4966a 4966b 4967a 4967b 4968a 4968b 4969a

B1 B1 B1 A1 C D* A1 A1 B1 F C A1 C E F B1 or E C A1 A1 F B1 A1 B1 F B1 A2 B1 A1 F B1 B1 D C F B1 A1 B1 F A1 A3 A1 C A1 A1 A1 C C A1 E A1 B1 A1 B1 D* B1 C

xxxPx#P xxPx#P xxxxxPx#P Px#xxpxx xxP#px P#xPsx xxPx#xPx xP#xxxpxx xxxxxPx#P Iudeo#Px xxxxpx#px Px#xPx xxP#px PSxxx#P Iohannes#xP xxPXx#P xxxxxxxxP#Px Px#xxxPx xxxxPx#Px Petrus#xPx xxxxxPx#P Px#xPx xxxxPx#xP Iohannes#xxxxIudeon xxxxPx#P P#xxxPS xxxxPx#P Px#xPx xxxIudeon#P xxPxx#P xxxxxxPx#P px#PSS xxxxP#px(xx) Pxx#xGalilea xxxpxx#P Px#xPx xxPx#P Simon#Petrus#P P#xxPx xxxxPx P#xxPx xxPxx px#xPx P#xxxxxpxx xxxxxP#xxPx xxxxxpx#Px(xx) xxxxP#px xPXx#xxxPx pxXx#P P#xxxxPx xxxxxxPx#P xP#xxxPx xxxxP#xP Px#pXSx xxxx(xx)xxxPX#px xxPXx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 459

Index to the scansion of the Heliand 4969b 4970a 4970b 4971a 4971b 4972a 4972b 4973a 4973b 4974a 4974b 4975a 4975b 4976a 4976b 4977a 4977b 4978a 4978b 4979a 4979b 4980a 4980b 4981a 4981b 4982a 4982b 4983a 4983b 4984a 4984b 4985a 4985b 4986a 4986b 4987a 4987b 4988a 4988b 4989a 4989b 4990a 4990b 4991a 4991b 4992a 4992b 4993a 4993b 4994a 4994b 4995a 4995b 4996a 4996b 4997a

A1 B1 B1 D* C D* A1 D* B1 A1 B1 F C A1 B1 E A1 B1 B1 C C A1 B1 D* B1 A1 B1 A1 C A1 A1 A1 B1 A1A1 B1 B1 A1 A1 A1 A1 B1 E B1 A1 B1 F D A1 B1 F A1 A1 C A1 B1 B1

Pxx#Px xPx#xP xxxxxP#xxP Px#Psx xxxxPSx Pxx#PXx#P xxxxPx#Px xP#xxxxPSx(xxx) xxxxxxxxxPx#xpx xxxPx#xxxxxxPx xxxxpxx#xP xxxgalileisk#P xxxxxxpx#PX xP#xxxPx xxPx#P PSx#xP xxxxPx#xPx xxxxPx#xP xxxxPx#P xxxxPxx xPSx xPx#xxxxpXx xxxxxxxxPx#P xPx#Psx xxxxxPx#xP Px#xPx xxxxxPx#P xxxPx#Px xxxxxPSx Px#Pxx Px#Px Pxx#xPx xxxPx#px xPxx#xxPx#Px xxxxxxxxPx#px xxxxPx#P PXSx xPx#xPx xxxxPx#Px Pxx#xxPx xxxxxPx#P pxS#xpx xxxPxx#P Px#xPx xxxPx#P Px#xSimon#Petruse px#Pxx xxxPx#xxxPx xxxxxPx#P Simon#Petruse P#xxPx P#xxPx xxxpx#Px Px#xxxxPx xxxPx#xP xxxxxPx#P

4997b 4998a 4998b 4999a 4999b 5000a 5000b 5001a 5001b 5002a 5002b 5003a 5003b 5004a 5004b 5005a 5005b 5006a 5006b 5007a 5007b 5008a 5008b 5009a 5009b 5010a 5010b 5011a 5011b 5012a 5012b 5013a 5013b 5014a 5014b 5015a 5015b 5016a 5016b 5017a 5017b 5018a 5018b 5019a 5019b 5020a 5020b 5021a 5021b 5022a 5022b 5023a 5023b 5024a 5024b 5025a

B1 A1 B1 C A1 A1 D* A1 B1 A1 C A1 B1 E A1 C A1 A1 B1 A1 B1 A1 B1 B1 A1 B1 A1 A1 B1 A3 B1 A1 B1 A1 C A1 C A1 B1 A3 A1 C A1 A1 B1 A1 B1 C A1 A1 B1 A1 B1 D B1 B1

459 xxxxPX#P Px#Px xxxxxPx#P xpxSx xPx#Px Px#xPx xP#xxxPx#P Px#xxPx xxxxxxxPx#px xP#xxxpxx xPsx Px#xPx xxxPx#P PSx#P xxxxPx#Px xxPsx Px#Px Pxx#xxPx xxxxxxxxxxPx#P pXSx#Px xxxxPx#px Px#Px xxxpxx#xP xxPx#px P#xPx xPx#P Px#Px px#xxPx xxPx#P(xx) xxxxxxxPx xxxxxpXx#xP Px#Px xxxxPx#P Px#Pxx xxpXSx Px#pxx xxxxxxxP#px Px#Px xxxxxxxPx#P xxxxxPx PX#P#x xxxxxxPxx Px#Px xPx#xxxxxPx xxxPx#P Px#xxxxPx xxxxxP#xP xPxx pxx#Px P#xxPx xxxxxPx#P Px#xPx xxxxxPx#P px#Pxx xxxPx#P xxPx#P

MOSP-Index 1.qxd

5/11/04

460 5025b 5026a 5026b 5027a 5027b 5028a 5028b 5029a 5029b 5030a 5030b 5031a 5031b 5032a 5032b 5033a 5033b 5034a 5034b 5035a 5035b 5036a 5036b 5037a 5037b 5038a 5038b 5039a 5039b 5040a 5040b 5041a 5041b 5042a 5042b 5043a 5043b 5044a 5044b 5045a 5045b 5046a 5046b 5047a 5047b 5048a 5048b 5049a 5049b 5050a 5050b 5051a 5051b 5052a 5052b 5053a

2:49 PM

Page 460

Index to the scansion of the Heliand A1 C A1 B1 A1 A1 B1 E* A1 D* A1 A3 C B1 or E C A3 B1 A1 B1 A1 B1 A3 A1 A1 B1 E A1 B1 A1 E B1 A1 B1 D* C A1 C A1 B1 A1 B1 A1 B1 A1 B1 A1 B1 D A1 B1 A1 C A1 F C A1

P#xPx xxxPSx Px#Px xxxPx#P Px#PX xPxx#xPx xxxxxxPx#xP pxx#Px#xpx xxxxxxpXx#Px PX#xxxPXx Px#Px xxxxPx xxxP#px xPXx#P xxxP#px xxxxPx xxPx#xP Px#xPx xP#xP Px#xpxxx xxP#xpx xxxxxPx Px#Px pxx#xxPxx xxxPx#px pXSx#P PxSx xxxPx#P pxx#xPx pxSx#P xxxxPx#px xPx#xxPx xxxxP#xxP Psx#Pxx xxxpX#px Px#xxPx xxxxPSx xP#xxpxx xxxxxxPx#P Px#Px xxxxPxx#xP Px#Px xxxxxPx#P xPx#xxxPx xxxxPx#P P#xxPx xxxPX#P P#pXsX Px#xPx xxxxxPx#P Px#pxx xPSx xxxxpxx#Px Iudeono#Px xxP#px Px#xxxPx

5053b 5054a 5054b 5055a 5055b 5056a 5056b 5057a 5057b 5058a 5058b 5059a 5059b 5060a 5060b 5061a 5061b 5062a 5062b 5063a 5063b 5064a 5064b 5065a 5065b 5066a 5066b 5067a 5067b 5068a 5068b 5069a 5069b 5070a 5070b 5071a 5071b 5072a 5072b 5073a 5073b 5074a 5074b 5075a 5075b 5076a 5076b 5077a 5077b 5078a 5078b 5079a 5079b 5080a 5080b 5081a

B1 A1 C A1 C A1 D F C A1 C B1 A1 A2 E A1 C A1 B1 C C D* B1 or E A1 B1 A1 A1 A1 C D* B1 A1 C A1 B1 C E A1 B1 A3 A1 A3 C A1 C B1 A1 B1 B1 A1 B1 A1 B1 A1 B1 E

xxxPx#P pxx#xxPx xxxxxxP#px Px#Px xxxP#px Px#xPx px#Pxx px#Iudeono xxxpx#Px PS#xPx xxPsx xPXS px#xPx Px#xxPS PSx#P Px#Px xxxxP#px Px#xPx xxxxPx#P xxxPxx xPSx Px#Pxsx xPXx#P xxPXx#Px xxPx#P xxxxxxPXSx Px#Px xPx#xPx xxxxxxxpx#Px xP#xPsx xxxxPx#xP xPx#xPx xxP#Px Px#xPx xxxxxPx#P xxxP#px PSx#P Px#Px xxxxxxPx#P xxxxxxPx Px#xPx xxxxxPx xxP#px xxxPx#PX xxxxPsx xxPx#P P#xPx xPx#px xxxxxxxxPx#P xpxx#xxpxx xxxxxxP#xpx xPx#xpxx xxxxPx#xpx Px#Px xxxxxxpxx#xP pxSx#P

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 461

Index to the scansion of the Heliand 5081b 5082a 5082b 5083a 5083b 5084a 5084b 5085a 5085b 5086a 5086b 5087a 5087b 5088a 5088b 5089a 5089b 5090a 5090b 5091a 5091b 5092a 5092b 5093a 5093b 5094a 5094b 5095a 5095b 5096a 5096b 5097a 5097b 5098a 5098b 5099a 5099b 5100a 5100b 5101a 5101b 5102a 5102b 5103a 5103b 5104a 5104b 5105a 5105b 5106a 5106b 5107a 5107b 5108a 5108b 5109a

A1 A1 B1 B1 A1 D* B1 A3 C B1 or E B1 D B1 A1 B1 D* F E B1 A1 B1 C B1 B1 D* D* C E C E A1 A1 B1 A1 B1 A1 B1 A1 B1 D* C A1 B1 A1 B1 A1 F A1 B1 A1 B1 F B1 A1 A1 F

Px#xxPx xpXx#xPx xxPxx#P xxxxxxPx#xP Px#Px Px#xxxpx#px xxPx#P xxxxxPx xxpx#Px xPXx#px xxP#xP P#pX#Px xxxxxxPx#P xxxxPx#xxxxPx xxxxxxPx#xpx xPx#px#px xxxxxxxIudeon#P PSx#px xxxPx#P xxxPx#xxxPx xxxxxPx#xP xxxxP#Px xxxxxxPx#P xxxxxPx#xpx xxPxx#P#px Px#Px#px xpXSx xpxSx#px xxxxP#px xpXSx#px xxxxxpxx#Px xxPx#xPx xxxxxPx#P xP#xxxPx xxPx#px Pxx#xxxPx xxxxPx#P P#xxPx xxxxxPx#P(xx) xpx#xPSx xxxxP#px PSx#xxPx xxxxxxPx#px Px#xxxxpxx xxxxxPx#px px#xxP#x xxxxIudeon#P xPx#xPx xxPx#P Px#xxxxPx xpx#xxP P#Iudeono xxxxxPx#px Px#xPx xxxxxxxPx#xPx xxxxxHierusalem

5109b 5110a 5110b 5111a 5111b 5112a 5112b 5113a 5113b 5114a 5114b 5115a 5115b 5116a 5116b 5117a 5117b 5118a 5118b 5119a 5119b 5120a 5120b 5121a 5121b 5122a 5122b 5123a 5123b 5124a 5124b 5125a 5125b 5126a 5126b 5127a 5127b 5128a 5128b 5129a 5129b 5130a 5130b 5131a 5131b 5132a 5132b 5133a 5133b 5134a 5134b 5135a 5135b 5136a 5136b 5137a

F D A1 A1 B1 F C B1 A1 A1 A1 A1 A1 A2 A1 A1 C D* D* A1 B1 A1 A1 B1 B1 C C A3 B1 A2 C D* B1 F B1 D* F A1 C F F A1 C C A1 C E D* F B1 or E B1 C A1 A1 F A2

461 Iudeo#Px px#Pxx Px#Px xPx#xPx xxxxPx#P Iudeo#Px xxxxpx#Px xxxPxx#P Px#xPx xPx#xxxxpxx xxxxxxPx#Px xxP#xxxPx xxxxxxPx#xPx Pxx#xxpXS PXx#pxx PXSx xxP#px P#xxPXx xxxxxPx#xPxx pxx#xxPx xxxxxxPx#P Px#Px xP#xxPX xxpxx#xP xxxxxPx#P xxPxx xP#px xxxxxPx xxxPx#P xxPx#PS xpx#px Px#xxxxpxsx xxxxPx#px xRumu-S xxxxPx#xP px#xxxxxPxx xxxxxxxxPx#Iudeono xPXx#xPx xxPsx Pilatus#xxPx xxxPonteo#Px Px#Px xxxxP#px xxxPSx P#xPx xP#px PSx#P xPx#xxxpx#px Iudeo#Px xxPXx#P xxxxxxxPx#px xxxxPxx Px#Px Px#Px xxxxpx#Iudeono Px#xxPS

MOSP-Index 1.qxd

5/11/04

462 5137b 5138a 5138b 5139a 5139b 5140a 5140b 5141a 5141b 5142a 5142b 5143a 5143b 5144a 5144b 5145a 5145b 5146a 5146b 5147a 5147b 5148a 5148b 5149a 5149b 5150a 5150b 5151a 5151b 5152a 5152b 5153a 5153b 5154a 5154b 5155a 5155b 5156a 5156b 5157a 5157b 5158a 5158b 5159a 5159b 5160a 5160b 5161a 5161b 5162a 5162b 5163a 5163b 5164a 5164b 5165a

2:49 PM

Page 462

Index to the scansion of the Heliand B1 A1 C E B1 D A1 A3 A1 A1 F C E A1 B1 F B1 A1 B1 A1 B1 E* B1 A1 B1 F B1 A1 B1 A1 C C A1 A3 F A3 B1 A1 A1 A3 C D* B1 A1 D* F C A1 B1 C B1 A1 B1 or E A1 B1 A1

xxP#xxP Pxx#xxxxpxx xxxxxxP#px xpxSxx#P xxxPS#P xxxpx#Px#P xPx#xxPx xxxxxxxPx Pxx#Px xxxxPx#Px Pilatus#xP xxPsx PXx#P Px#Px xxPx#P Iudases#px xxxpx#xP xPx#xPx xxxxxxP#xxP xxPx#Px xxxxxPx#P Px#Px#xP xxxxxpxx#xP Px#Px xxxxxxxPxx#P P#xxxxxIudiun xxxxPx#P Px#Px xxxxpxx#P Px#xxpXx xxxxxPSx(xx) xxPxx Px#xPx xxxxxxxPx P#Iudeono xxxxxxPx xxxxxP#xxP xxpxx#Px Px#Px xxxxPx xpx#Px xPx#xPx#P(xxx) xxxxxxPx#xP xP#xxxxxpxx xxP#xxP#px Iudas#Px xxxpx#Px Px#xPx xxxpxx#P xxxpx#Px xxxxxPx#P P#xxxxxPx xxxPXx#P Pxx#pxx xxxPx#px px#xxpxx

5165b 5166a 5166b 5167a 5167b 5168a 5168b 5169a 5169b 5170a 5170b 5171a 5171b 5172a 5172b 5173a 5173b 5174a 5174b 5175a 5175b 5176a 5176b 5177a 5177b 5178a 5178b 5179a 5179b 5180a 5180b 5181a 5181b 5182a 5182b 5183a 5183b 5184a 5184b 5185a 5185b 5186a 5186b 5187a 5187b 5188a 5188b 5189a 5189b 5190a 5190b 5191a 5191b 5192a 5192b 5193a

A1 B1 A1 A2 C A1 B1 D A1 D B1 D* A1 C B1 D* A1 C A1 B1 D F F A1 B1 C C B1 F A1 F C A1 A1 B1 A1 B1 A1 B1 B1 A1 A1 B1 A1 A1 C D A1 B1 D* C A1 A1 A1 B1 D

xxxP#xPx xxxxPx#P Px#Px P#xxpxS xPXx px#xPx xxPx#xP P#Px#xP P#xxPx P#Px#px xxxxxxPx#P xP#xP#px Px#pxx xxxPSx xxxP#xxP Px#PSx Px#Px xxxxxxpXSx Px#Px xxxxPx#xP px#Pxx xRumu-S xxxxxxxPx#Iudeono Px#Px xxpxx#P xxxxpx#Px xxxxxP#px xxxPxsx Pilatus#xP Px#Px xxxP#Iudeono xxxP#Px Px#xPx Px#xPx xxxxxxPx#P xxxxPx#Px xxxxxxxxPx#xpx Px#xPx xxxxxxxPx#P xxxxP#xpx pxx#xPx Px#xPx xxxxxpxx#xpx xPx#xxPx xxxxPx#Px xxxxxpx#Px xxxxxxxxpx#Pxx Px#Px xxxxxxxPx#P xPx#xPsx xxxxP#px px#xxP#x pX#xxxPx xpx#xxxPx xxxxPx#P px#Pxx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 463

Index to the scansion of the Heliand 5193b 5194a 5194b 5195a 5195b 5196a 5196b 5197a 5197b 5198a 5198b 5199a 5199b 5200a 5200b 5201a 5201b 5202a 5202b 5203a 5203b 5204a 5204b 5205a 5205b 5206a 5206b 5207a 5207b 5208a 5208b 5209a 5209b 5210a 5210b 5211a 5211b 5212a 5212b 5213a 5213b 5214a 5214b 5215a 5215b 5216a 5216b 5217a 5217b 5218a 5218b 5219a 5219b 5220a 5220b 5221a

C A3 A1s D* B1 A3 B1 A3 A1 A3 A1 B1 C A1 C A1 E D B1 F C A1 C A1 C A1 C D B1 B1 A1 A2 B1 A2 B1 A1 B1 F B1 D* B1 D* F A1 B1 B1 A1 A1 B1 E B1 A1 B1 C B1 C

xxxPSx(xx) xxxxxpxx PS#px xPx#xxxxxxxxPsx xxxxPx#px xxxxxxPx xxxPx#P xxxxxxPxx P#xPx xxxxxxxPx Px#xPxx xxPxx#P xPsx Px#xPx xxxPsx xPx#xxxxxpxx PSx#P px#Pxx xxxxPx#P px#xRumu-S xxxxxP#px Px#Px xxxxPSx Pxx#Px xxxxxP#pX xpxx#Px xxxxxP#px px#Pxx xxxxxxPxx#px(xx) xxxPx#P Px#Px Px#xxxxpXS xxxxPxx#px P#xxPS xxxPX#P Pxx#xxpxx xxxxxPx#px(xx) Giudeo#Px xpxX#P xxxPx#Psx xxxxpxx#xpx xPx#xxxpxXx#P Iudeo#Px Px#xPx xxxxPx#xP xxxPx#P Px#pxx P#xxxxxPx xxxxxP#xpx PXxx#P xxxPx#P xxxpxx#Px xxPx#px(xx) xxxxpXSx xxxPx#P xxxxPSx

5221b 5222a 5222b 5223a 5223b 5224a 5224b 5225a 5225b 5226a 5226b 5227a 5227b 5228a 5228b 5229a 5229b 5230a 5230b 5231a 5231b 5232a 5232b 5233a 5233b 5234a 5234b 5235a 5235b 5236a 5236b 5237a 5237b 5238a 5238b 5239a 5239b 5240a 5240b 5241a 5241b 5242a 5242b 5243a 5243b 5244a 5244b 5245a 5245b 5246a 5246b 5247a 5247b 5248a 5248b 5249a

C C A1 A3 F E C A1 B1 B1 A1 A1 B1 A1A1 B3A1 A1 B1 B1 D D* B1 A1 F A1 A1 C A1 C A1 C B1 A1 C F C A1 C F F E D D* B1 C A1 B1 A1 A1 F B1 C E A1 B1 A1 A1

463 xxPsx xxpx#Px Pxx#Px xxxxPx Iudeo#Px PXx#xP xpxSx xPXx#xPx xxxxxxxxpXx#xpx xxxPsx#P Px#Px xxxPx#Px xxxxPx#P xpx#xxxPx#pxx xxxxP#xPx xPx#xxPx xxxxPxx#P xxPx#px px#Pxx Px#Px#P xxxPx#xP Px#Px xxxxxxxpx#Iudeono Px#Px xxxxpxx#Px xxP#px xxxxxPx#Px xxxpXSx Px#xPx xxPSx xxxxxPx#px Px#Px xxxPSx Iudeo#Px xxxxpx#px Px#Px xxxxxP#PX xPx#xGalileo#Px xxxxIudeon#P pXSx#px px#Pxx Px#Psx xxxPx#P xxxxPxx Px#Px xxxxxxxPx#P Px#xPx xPx#xxxPx px#Iudeono xPx#px xxpxsx PSx#P Px#xPx xxxxxPx#P P#xPx Px#xPx

MOSP-Index 1.qxd

5/11/04

464 5249b 5250a 5250b 5251a 5251b 5252a 5252b 5253a 5253b 5254a 5254b 5255a 5255b 5256a 5256b 5257a 5257b 5258a 5258b 5259a 5259b 5260a 5260b 5261a 5261b 5262a 5262b 5263a 5263b 5264a 5264b 5265a 5265b 5266a 5266b 5267a 5267b 5268a 5268b 5269a 5269b 5270a 5270b 5271a 5271b 5272a 5272b 5273a 5273b 5274a 5274b 5275a 5275b 5276a 5276b 5277a

2:49 PM

Page 464

Index to the scansion of the Heliand A1 F C A1 F A2 B1 F B1 A1 C A1 C F C A1 B1 B1 A1 F F B1 A1 C C F A1 A1 B1 B1 A1 B1 E A1 C B1 C B1 E A3 B1 F C D* B1 B1 C B1 A1 A1 B1 C F D* D A1

xxxxxPx#Px xPx#xGalilea-Sx xxPsx pXxx#Px Erodes#xP#P PXx#pXS xxxxxxPx#xP xPx#xRumu xxxPx#xpx xpxx#xxxxPx xxpx#Px Px#xPx xxxxxxpx#Px xHierusalem xxPsx xxpxx#xxxPx xxxxPx#P xxxxPxx#P Px#Px Px#Iudeono Pilatus#xP#P xxxPx#P pxx#Px xxPxx xP#px xxxxxErodese Px#Px Px#xPx xxxxpxsx#P xxpxx#xP PX#pxx xxPx#P PXx#P Px#xpXx xxxPsx xxxPx#P xxxxxpx#Px xPx#P xpxSx#P xxxxxPx xxxxPx#P pX#Erodes xxxxxP#px Px#PXx xxPx#px xxxPx#P xpx#Px xxxxxxPx#P Px#Px Px#xxPx xxpxx#px xxPSx Iudeo#Px Pxx#xxxxPsX pxSXx pxx#Px

5277b 5278a 5278b 5279a 5279b 5280a 5280b 5281a 5281b 5282a 5282b 5283a 5283b 5284a 5284b 5285a 5285b 5286a 5286b 5287a 5287b 5288a 5288b 5289a 5289b 5290a 5290b 5291a 5291b 5292a 5292b 5293a 5293b 5294a 5294b 5295a 5295b 5296a 5296b 5297a 5297b 5298a 5298b 5299a 5299b 5300a 5300b 5301a 5301b 5302a 5302b 5303a 5303b 5304a 5304b 5305a

C A1 C A1 B1 A1 C F A1s A1 B1 F C A1 C C C A1 C A1 C A1 D B1 A1 C A1 B1 A1 A1 A1 C B1 A1 F A3 A1 D* C F B1 D* A1 A2 B1 A1 C C A1 B1 A1 D* C A1 F C

xxxPsx P#xxPx xxxpx#Px Px#xPx xxPx#P pxx#xxpxx xxxxPsXx Erodese#xxPx PS#px Px#xPx xxxPx#P Iudeo#Px xxxxpx#px Px#xxPx xxxxxpXsX xxpx#Px xxxxpxsx xPx#xxxxxPx xxxxP#px pxXx#Px xxxxxpx#Px pxx#xPx P#Pxx xxPx#P P#xxPx xxPSx P#xpxx xxxxxPxx#P Px#Px xPx#xxxPx P#xPx xxxpx#Px xxxxxxPx#P Px#xPx Iudeon#pxxx xxxxxPx Px#xPx Px#pXSx xxxxxP#px Erodes#xpX xxPx#P xPx#xxxPx#P xxPx#Px Pxx#xpXS xxxpxx#P xpx#xPx xxxpx#Px xxxxxxPSx P#xpxx xxxxxpxx#P PXSx P#xxPsx xxxxxxxxxP#Px xxpXx#Px xPilatus#P xxxPsx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 465

Index to the scansion of the Heliand 5305b 5306a 5306b 5307a 5307b 5308a 5308b 5309a 5309b 5310a 5310b 5311a 5311b 5312a 5312b 5313a 5313b 5314a 5314b 5315a 5315b 5316a 5316b 5317a 5317b 5318a 5318b 5319a 5319b 5320a 5320b 5321a 5321b 5322a 5322b 5323a 5323b 5324a 5324b 5325a 5325b 5326a 5326b 5327a 5327b 5328a 5328b 5329a 5329b 5330a 5330b 5331a 5331b 5332a 5332b 5333a

A1 A1 A1 A1 B1 B1 A1 A1 C F B1 A1 C A1 B1 D C D B1 C B1 B1 A1 A1 B1 A1 B1 C A1 F B1 A2 A1 B1 A1 A1 B1 A1 A1 D* B1 A1 F D* A1 C A1 A1 B1 A1 B1 A1 B1 D* B1 A1

Px#xPx Px#xPx px#xPx PxSx xxPx#xP xxPx#P Px#xPx Px#xPx xxPsx Iudeon#xxPSx xxxpxx#P xpx#xPx xxxxpx#PX Pxx#Px xxxxPx#px px#Pxx xxP#px P#pxsx xxxxxPx#P(xx) xxxpx#Px xxPx#xpx xxxxpxx#xpx xPx#Px xPx#xxPx xxxxxPx#xP Px#xxPx xxxpxx#P xxxxxpx#Px P#xxPx Erodes#Px xxxP#xP xxxPx#PS xxxxxPx#xPx xxxxxxxxPx#xpx Px#Px P#xPx xxxxxxxxxPx#P xPx#xPx Px#Px Px#xxPsx xxxxPx#P pxx#xPx P#Iudeono Px#xPx#px Pxx#Px xxPSx pxx#Px P#xPx xxxPx#px Px#xPx xxxPx#px Px#xPx xxxxPx#P PXx#px#px xxxPx#P PXSx

5333b 5334a 5334b 5335a 5335b 5336a 5336b 5337a 5337b 5338a 5338b 5339a 5339b 5340a 5340b 5341a 5341b 5342a 5342b 5343a 5343b 5344a 5344b 5345a 5345b 5346a 5346b 5347a 5347b 5348a 5348b 5349a 5349b 5350a 5350b 5351a 5351b 5352a 5352b 5353a 5353b 5354a 5354b 5355a 5355b 5356a 5356b 5357a 5357b 5358a 5358b 5359a 5359b 5360a 5360b 5361a

B1 C A1 A3 B1 A1 C A3 A1 D* C D* C C A1 D* A1 A3 C D* B1 B1 B1 F B1 C C A1 C A3 A1 A1 B1 A1 B1 A1 B1 A1 B1 A3 B1 A1A1 B1 E A1 D B1 A1 A1 F B1 B1 C A1 B1 A1

465 xxxxxPx#xpx xxxxxpXsx pxx#Px xxxxpxx xxPx#xP pxx#xxPx xxxxxxP#px xxxxPx Px#xPx px#xxPsx xxxxpx#px xP#xxxpxsx xxP#Px xxxPsx Px#Px Px#xxpx#px xxxxxxxpxx#Px xxxPx(xx) xxxxxxxP#px Px#PxS xxxxxxxxPx#P xxxxPx#xpx xxxxxxPx#xpx px#Iudeono xxxPx#P xxxPXx xpx#Px xxPXx#xPx xpx#Px xxxxxPx Pxx#Px xxPXx#xxxPx xxxxpxS#px P#xxxPx(xx) xxxP#xxP Px#xPx xxxxxxPx#P Px#xPx xxxxPxx#P xxxxPx xxPXsx xPx#xPx#Px xxxxxP#xxP PSx#P Px#xPx px#Pxx xxxxxxxxPx#xP xxpxx#xxxPx Px#xpxx Px#Iudeono xxx(xxx)xPxx#P xxPx#P xxxxpx#Px Px#xPx xxxxpxx#P Px#xPx

MOSP-Index 1.qxd

5/11/04

466 5361b 5362a 5362b 5363a 5363b 5364a 5364b 5365a 5365b 5366a 5366b 5367a 5367b 5368a 5368b 5369a 5369b 5370a 5370b 5371a 5371b 5372a 5372b 5373a 5373b 5374a 5374b 5375a 5375b 5376a 5376b 5377a 5377b 5378a 5378b 5379a 5379b 5380a 5380b 5381a 5381b 5382a 5382b 5383a 5383b 5384a 5384b 5385a 5385b 5386a 5386b 5387a 5387b 5388a 5388b 5389a

2:49 PM

Page 466

Index to the scansion of the Heliand C A1 B1 E B1 D* B1 A1 C A1s A1 D* B1 D* F A1 B1 A1 B1 C B1 A1 F B1 A1 A1 C D* A1 F B1 A3 B1 A1 C A1A1 F D B1 or E A1 B1 or E C A1 B1 C A1 B1 A1 B1 C A1 F C A1 C B1

xxxxxxP#px xpx#xxxPx xxxxPx#px pXSx#px xxxxxxxPx#px xPx#xxpXSx xxxP#xpx xP#xxxxPx xxxxxxP#px PS#px xxxxxxPx#Px xxxxPx#Psx xxxxxxPx#xpx xxPx#xpxsx xPx#Iuðeono Px#xxPx xxxxxPsx#P Px#xPx xxPx#P xxxP#pxx xPX#P Px#xxPx Pxx#Iudeon xxxxPxx#P Px#xPx px#xPx xxxxxpX#Px xPx#xxxxpxSx xxxxPx#Px xRumu-S xxxxPx#xxP xxxxxxxxPx xxxxxxxPx#xpx xP#xxxxxxPx xxP#px Px#xxPXSx px#Iudeono xpx#PS#P xPXx#P Px#xPx xPXx#P xxPSx xPSxx#PX xxxPx#P xxxxpX#Px Px#xxPx xxxxxxxPx#P Px#xPx xxxxPx#P xxxxpXSx Px#Px Iudeo#Px xxxP#Px xPx#xxxPx xxxxxxP#Px xxxxPXS

5389b 5390a 5390b 5391a 5391b 5392a 5392b 5393a 5393b 5394a 5394b 5395a 5395b 5396a 5396b 5397a 5397b 5398a 5398b 5399a 5399b 5400a 5400b 5401a 5401b 5402a 5402b 5403a 5403b 5404a 5404b 5405a 5405b 5406a 5406b 5407a 5407b 5408a 5408b 5409a 5409b 5410a 5410b 5411a 5411b 5412a 5412b 5413a 5413b 5414a 5414b 5415a 5415b 5416a 5416b 5417a

C A1 C A1 C D A1 A1 B1 A1 B1 D* B1 C A1 A3 C D B1 A1 E A2 C A3 A1 F B1 C A1 C F B1 C B1 B1 A1 B1 C A1 C F D C A3 A1 A1 B1 D* F B1 A1 C A1 A1 B1 A1

xxxxxxPsx xPx#xxxPx xxxxxxxP#px Px#xPx xxxxpXSx xpx#Px#P Px#Px xxP#xxxxxPx xxxPx#P px#xxpXx xP#xxxP Px#P#px xxPx#P xxxpXSx Px#Px xxxxxPx xxxP#Px xP#pXsx xxxxxxPx#xpx Px#xPx xxPSx#xpx xPx#pXS xxxxxpx#Px xxxxPx Px#xPx Barrabas#xxPx xxxxPx#P xxxPSx pxx#xPx xxPSx Px#Iudeono xxPx#xP xpx#Px xxPxx#px xxPx#P xPx#Px xxxxPx#P xxPsx px#xPx xxxxpxsx xPx#Iudeono xP#Pxx xxxpx#Px xxxxxxPx Px#Px pxx#Px xxxpxx#P P#xxxxpxSx xPx#Iudeono xxxxpxx#P Px#xpxx xxxxPsx P#xPx xPxx#xPx xxxPx#P P#xpXx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 467

Index to the scansion of the Heliand 5417b 5418a 5418b 5419a 5419b 5420a 5420b 5421a 5421b 5422a 5422b 5423a 5423b 5424a 5424b 5425a 5425b 5426a 5426b 5427a 5427b 5428a 5428b 5429a 5429b 5430a 5430b 5431a 5431b 5432a 5432b 5433a 5433b 5434a 5434b 5435a 5435b 5436a 5436b 5437a 5437b 5438a 5438b 5439a 5439b 5440a 5440b 5441a 5441b 5442a 5442b 5443a 5443b 5444a 5444b 5445a

B1 A1 B1 A1 C D* C A1 C C E A1 B1 A1 B1 A1 C E* B1 B1 E F B1 F A1 A1 B1 D B1 A1 C C A1 A1 C F A1 A1 B1 B1 A1 A1 C B1 A1 A1 F C E C B1 D A1 A1 C A1

xxPX#P pxx#xPx xxxP#xxP xxxxxPx#xPx xxxxxP#Px Px#xPxx#P xxxpxsx Px#xPx xxxpx#Px xxxxPsx PXx#P pxx#xPx xxxxPx#xP pXx#xxPx xxPx#xP P#xxxPx xxP#Px Px#Px#P xxxxxpX#xP xxxxPx#xpx Psxx#P Satanas#Px xxPx#P Iudases#xP Pxx#Px xPx#xxPx xxxPX#P P#Pxx xxxPx#P Px#xxPx xxxxxxpX#Px xxPXx Px#xPx Px#xPx xP#px xxSatanase P#xPx xxP#xxPx xxxPx#P xxPx#P P#xxPx xpxx#xPx xxxxxxP#Px xxxpxx#P Px#xPx Pxx#pxx Satanas#xP#xP xxpxsx PXx#P xxxP#Px xxxxPx#xP xxpx#pXSx PSx#PX Px#Px xxxPSx Px#Px

5445b 5446a 5446b 5447a 5447b 5448a 5448b 5449a 5449b 5450a 5450b 5451a 5451b 5452a 5452b 5453a 5453b 5454a 5454b 5455a 5455b 5456a 5456b 5457a 5457b 5458a 5458b 5459a 5459b 5460a 5460b 5461a 5461b 5462a 5462b 5463a 5463b 5464a 5464b 5465a 5465b 5466a 5466b 5467a 5467b 5468a 5468b 5469a 5469b 5470a 5470b 5471a 5471b 5472a 5472b 5473a

C A1 B1 A1 B1 B1 A1 B1 A1 A1 A1 B1 C E B1 A1 B1 A1 B1 B1 B1 A1 B1 E B1 D* A1 B1 A1 A1 B1 C A1 C B1 A1 B1 B1 C C A1 D* B1 A3 A1 B1 A1 B1 B1 A1 F A1 A1 A1 B1 D*

467 xxxxP#Px Px#Px xxxxxPx#xpx Px#xxPx xxxxxxxP#xpx xxxxxxxPXS xpxx#xPx xxPx#pX xP#xxxpXx Px#xpxx xxxxxPx#Px xxxPx#P xpx#Px xpXSx#xpx xxxxxPx#xP xP#xxxPx xxxxPx#P Px#Px xxxxxxPx#P xxxxPxx#P xxxPx#P Px#xPx xxxxxpx#xxpx PSx#xpx xxxxxPx#px Pxx#Px#xP pxx#xPx xxPx#P pxx#Px xP#xxxPx xxxPx#P xxpxsx xpxx#Px xxPsx xxPx#P pxx#xPx xxxxPx#P xxxxPx#P xxxP#px xpxsx pxxx#xPx xPx#PxS xxPx#P xxxxxPx Px#Px xxxxPx#P xPx#xxPx xxxpxx#P xxxPx#P px#xPx xxxxPx#Iudeono xPXx#xxPx xPxx#xPX xPx#Px xxxxxPx#xpx P#xxxxPx#px

MOSP-Index 1.qxd

5/11/04

468 5473b 5474a 5474b 5475a 5475b 5476a 5476b 5477a 5477b 5478a 5478b 5479a 5479b 5480a 5480b 5481a 5481b 5482a 5482b 5483a 5483b 5484a 5484b 5485a 5485b 5486a 5486b 5487a 5487b 5488a 5488b 5489a 5489b 5490a 5490b 5491a 5491b 5492a 5492b 5493a 5493b 5494a 5494b 5495a 5495b 5496a 5496b 5497a 5497b 5498a 5498b 5499a 5499b 5500a 5500b 5501a

2:49 PM

Page 468

Index to the scansion of the Heliand A1 A1 B1 A1 D D B1 B1 A1 A1 B1 B1 A1 A1 B1 D F A1 C A1 B1 A1 B1 A1 B1 C B1 F A1 E C A1 C A1 D E A1 A1 B1 A3 B1 A1 B1 A3 A1 A1 A1 D* A1 B1 C D* A1 A1 B1 A1

Px#Px px#xPx xxxxxpxx#P P#xxxxxxPx px#Pxx P#pxsx xxxxxxPx#P xxxxxxxPxx#P pxx#Px Pxx#Px xxxxxPx#px(xx) xxxxPxx#P xPx#xxPx xPx#xPx xxxxxPx#xP xP#P#px pxsx#Iudeono xpxx#pxx xxxxxxxxxxP#px pxxx#Px xxxP#xxP xP#xxxpxx xxxxxxP#xpx xxxxpxx#xPx xxxxPx#px(xxx) xxxxpx#Px xxxxxxPx#xP xpx#xxxxxxIudeon xxxpxx#Px PXx#xP xpxSx pxx#xPx xxxPsx PX#xPx P#px#xxP Psxx#px Px#Px pxx#xPx xxxxxPx#px xxxxxxPx xxxxpxx#P Px#Px xxxxxxPx#px xxxxxPx xxxxxPx#Px px#xxPx xxxxxPx#xPx Pxx#xxxpXsx Px#pxx xxxxPx#P xxPSx Px#xPXS Pxx#Px Px#Px xxxPX#P Px#Px

5501b 5502a 5502b 5503a 5503b 5504a 5504b 5505a 5505b 5506a 5506b 5507a 5507b 5508a 5508b 5509a 5509b 5510a 5510b 5511a 5511b 5512a 5512b 5513a 5513b 5514a 5514b 5515a 5515b 5516a 5516b 5517a 5517b 5518a 5518b 5519a 5519b 5520a 5520b 5521a 5521b 5522a 5522b 5523a 5523b 5524a 5524b 5525a 5525b 5526a 5526b 5527a 5527b 5528a 5528b 5529a

B1 D* C A1 B1 A3 A1 A1 A1 A3 A1 A1 A1 A1 B1 D* A1 A1 A1 D* F A1 B1 A1 B1 A1 A1 A1 D F A1 D* B1 A1 B1 A1 B1 A3 C C B1 B1 A1 A1 C B1 E F B1 A1 B1 A1 B1 A1 A1 A1

xxxxxxxPx#P Px#xxxpXSx xxxxP#Px px#xxxxPx xxxxxPx#xP xxxxxpxx Px#Px Px#xxxPx Px#Px xxxxPx Px#Px pxx#xxxPx Px#Px Pxx#Px xxxxxPx#P Px#P#px Px#Px px#xxxxxPx xxxPxx#Px Px#Pxx#P Pxx#Iudeon px#xPx xxPX#P Px#xPx xxxxxpxx#P PS#xPx PXx#xPx xxP#xPx px#Pxx xxGalilea xxPx#Px Pxx#xxPsx xxxxPx#P Px#xpxx xxPx#P Px#xPx xxxxP#xP xxxxPx xxxP#px(xx) xxxPSx xxxPx#px xxPx#P Px#Px Px#Px xxxxP#px xxPx#P PXx#P Px#Iudeono xxP#xP Px#xPx xxxxPS#px Px#xPx xxxPx#P xxP#xPx Px#Px Px#xPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 469

Index to the scansion of the Heliand 5529b 5530a 5530b 5531a 5531b 5532a 5532b 5533a 5533b 5534a 5534b 5535a 5535b 5536a 5536b 5537a 5537b 5538a 5538b 5539a 5539b 5540a 5540b 5541a 5541b 5542a 5542b 5543a 5543b 5544a 5544b 5545a 5545b 5546a 5546b 5547a 5547b 5548a 5548b 5549a 5549b 5550a 5550b 5551a 5551b 5552a 5552b 5553a 5553b 5554a 5554b 5555a 5555b 5556a 5556b 5557a

A1 A1 B1 C B1 A3 A1 A1 F A1 C A1 C A1 A1 A1 A1 A1 A1 D* C F C E A1 A1 B1 C A1 A1 E A1 A1 A3 E A3 A1 A1 A1 D* A1 D* A1 A1A1 F F B1 A1 C B1 A1 C A1 A1 B1 F

P#xxxxPx Pxx#xxxPx xxxPx#P xxPXx xxxxxxPx#px xxxxPx Px#Px xxPx#Px P#Iudeono P#xpxx xxxxxP#px pxx#xPx xxP#PX Px#Px Px#Px Px#xpxx xxxPx#xxxxxPx Px#Px xP#xxPx P#xxxPxx xxxxxxP#px Px#xxIudeon xxxP#px PXx#P xxxxxxPx#Px xpxx#xPx xxxpx#xxP(xx) xxPXx xPx#Px Pxx#Px pXx#P xPx#xPx xPx#xPx xxxxPx PSx#xpx xxxxxpxx Px#Px px#xxxxPx xPxx#Px xxxxPx#PsX xpxx#Px P#xxpxsx xxxPx#Px Px#xPx#Px xxxxpX#Iudeono Iesus#xNazareth-S xxPx#P xPx#Px xxPsx xPx#P xPx#xPx xP#Px xxxxxxxxPx#Px Px#Px xxxxxxPx#xP pX#xxxxIudeon

5557b 5558a 5558b 5559a 5559b 5560a 5560b 5561a 5561b 5562a 5562b 5563a 5563b 5564a 5564b 5565a 5565b 5566a 5566b 5567a 5567b 5568a 5568b 5569a 5569b 5570a 5570b 5571a 5571b 5572a 5572b 5573a 5573b 5574a 5574b 5575a 5575b 5576a 5576b 5577a 5577b 5578a 5578b 5579a 5579b 5580a 5580b 5581a 5581b 5582a 5582b 5583a 5583b 5584a 5584b 5585a

B1 D B1 E B1 A3 F D* C A1 C C A1 A1 A1 A1s A1 A1 A1 A1 B1 D B1 A1 A1 C B1 A1 B1 F B1 A1 C B1 A1 B1 D E B1 B1 B1 A1 B1 A1 B1 A1 B1 A1 A1 A1 B1 B1 C D B1 A1

469 xxxxPxx#px P#pxsx xxxxxxxPx#xpx PSx#xpx xxxxPx#xP xxxxxPx px#Iudeono Px#xPx#P xP#Px Px#xPx xxxP#px xxpXSx Px#xPx Pxx#Px xPx#Px PS#px Pxx#Px Px#xxxPx xxxxxpxx#xPx px#xxxPx xxxpX#xxP(xxx) px#Pxx xxxxPx#xpx px#xxxxPx xxPx#xPx xxP#px xxxxxpxx#P xxPx#xPx xxxxPx#P xxPS#Iudeo xxxxPx#P P#xxxxpXx(xx) xxxxxxxP#Px xxxxxxxxPx#px P#xPx xPx#P pXsXx PSx#P xxxPx#xP xPx#px xxxxxxxxPx#P xxPx#Px xxxxxxPx#P Px#xPx xxxPx#P pxx#xPx xxxxxPx#P xxPx#Px xxxxPx#xPx Px#Px xxxPx#P xPx#xP xxxPsX(xx) P#px#px xxxxxPx#px Px#xxxPx

MOSP-Index 1.qxd

5/11/04

470 5585b 5586a 5586b 5587a 5587b 5588a 5588b 5589a 5589b 5590a 5590b 5591a 5591b 5592a 5592b 5593a 5593b 5594a 5594b 5595a 5595b 5596a 5596b 5597a 5597b 5598a 5598b 5599a 5599b 5600a 5600b 5601a 5601b 5602a 5602b 5603a 5603b 5604a 5604b 5605a 5605b 5606a 5606b 5607a 5607b 5608a 5608b 5609a 5609b 5610a 5610b 5611a 5611b 5612a 5612b 5613a

2:49 PM

Page 470

Index to the scansion of the Heliand C A1 C A1 B1 A1 A1 C B1 DA1 C A1 B1 A1 B1 A1 B1 A1 B1 B1 B1 A1 B1 A1 A1 B1 B1 C A1 A3 B1 E B1 A3 B1 or E A1 B1 B1 C A3 D C B1 F A1 A1 C A1A1 B1 C A1 F A1 A1 C A1

xxxpx#Px P#xxPx xxxpXsX PX#xxxpXx xxxxxxxPx#P Px#xxxxxpxx xxxxPx#Px xxxPXx xxxPx#P P#PXSx xxxxxxP#px Px#xxxPx xxxxPx#P xPx#xPx xxPx#px P#xxxxPx xxxxxPxx#P Px#xPx xxxPx#P xxxPxx#px xxPx#P pxx#xxpxx xxxxxxxxPx#P PXx#xxxxpXx Px#xPx xxxxxPx#P(xx) xxxxxxPx#P xxpx#px Px#Px xxxxPx xxxPx#P PXxx#P xxxxPx#px xxxxPx xxxxPXx#P Px#xPx xxxxxPx#P(xx) xxxPx#P xpXSx xxpx px#P#px xxxpxXx xxxxxxxxPx#P xxxxMaria Px#Px P#xxxPx xxxxP#px Px#PXSx xxxxP#xpx xxPxx Px#pxx xxxxIohannes Px#Px Px#xxxPx xxxpx#Px Pxx#xxPx

5613b 5614a 5614b 5615a 5615b 5616a 5616b 5617a 5617b 5618a 5618b 5619a 5619b 5620a 5620b 5621a 5621b 5622a 5622b 5623a 5623b 5624a 5624b 5625a 5625b 5626a 5626b 5627a 5627b 5628a 5628b 5629a 5629b 5630a 5630b 5631a 5631b 5632a 5632b 5633a 5633b 5634a 5634b 5635a 5635b 5636a 5636b 5637a 5637b 5638a 5638b 5639a 5639b 5640a 5640b 5641a

B1 A1 B1 A1 B1 B1 C F B1 A1 B1 D B1 B1 B1 B1 A1 E* C C B1 A1 B1 A1 B1 A1 B1 A1 B1 C A1 B1 B1 A1 A1 C B1 A1 B1 A1 B1 D* B1 A1 D A3 A1 D* B1 E B1 or E A1 F A1 B1 A1

xxPx#P Px#xxxPx xxxxPxx#P Px#xPx xxxpXS#P xxxxxPx#px xxxxxxpx#Px Px#xxIohannes xxxxxxPx#P Pxx#xxPx xxxPx#P px#PSx xxxxxPx#xP xxPx#px xxxPx#xP xxxxPx#P Px#Px PXS#xpX xxxxpX#Px xxxxpx#px xxxPx#P P#xxPx xxxP#xxP xxPx#xxPx xxxPx#P Px#xPx xxpx#xP P#xxPx xxxxPx#px xxxxpx#PX Px#Px xxxxPx#pX xxxxPX#P P#xxxPx pXx#PX xP#px xxpx#xP xxP#xxxPx xxPx#P Px#xpxx xP#xxpx xxxpXx#PxX xxxxxPx#P Px#xPx px#pxSx(xx) xxxxxxPx Px#Px Px#pXsX xxxxPx#px PXx#xP xxxxxxxPXx#P Px#xPx px#Iudeono Px#xxxxPx xxxxxPxx#P Px#xxxPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 471

Index to the scansion of the Heliand 5641b 5642a 5642b 5643a 5643b 5644a 5644b 5645a 5645b 5646a 5646b 5647a 5647b 5648a 5648b 5649a 5649b 5650a 5650b 5651a 5651b 5652a 5652b 5653a 5653b 5654a 5654b 5655a 5655b 5656a 5656b 5657a 5657b 5658a 5658b 5659a 5659b 5660a 5660b 5661a 5661b 5662a 5662b 5663a 5663b 5664a 5664b 5665a 5665b 5666a 5666b 5667a 5667b 5668a 5668b 5669a

A1 C A1 D* B1 B1 A1 C A1 D* C B1 B1 A1 B1 D* A1 A1 B1 A1 A1 A1 A1 A1 C D* A1 D* B1 A1 A1 A1 A1 D* B1 A1 B1 E* B1 or E B1 A1 C A1 A1A1 E* A1 E* B1 B1 A1 A1 D B1 A1 B1 A1

Px#Px xxxxPxx xPx#xPx Px#pXsx xxPx#px xxxPx#P Px#Px xxxPSx px#xxPx xPx#xPsx xxP#px xxPx#px xxxxxxpx#xP xpx#xPx xxxxxxPx#P Px#xPXx xxxxxPx#Px xPx#xxxPx xxxxxxPx#px Pxx#xPx xxxxxxPx#Px xPx#xxPx Px#xPx xxPx#xPx xxxP#px Px#xxpxXx#px xxxxPx#xPx(xx) xxP#xpx#Px xxxpx#xP P#xpXx pxx#Px xPxx#xxPx Pxx#Px P#xxxPsx xxxPx#P P#xxPx xxP#xxP PXSx#xpX xxPXx#P xpXx#xpx xPx#Px xxPsx Px#pxx pxx#xPx#Px Px#Px#px pxx#xxxPx xxxPx#px#xP xPx#xP xxpxx#P xxxPx#Px Px#xPx P#Pxx xxxpxx#P xPx#Px xxxxxpxx#P Pxx#xPx

5669b 5670a 5670b 5671a 5671b 5672a 5672b 5673a 5673b 5674a 5674b 5675a 5675b 5676a 5676b 5677a 5677b 5678a 5678b 5679a 5679b 5680a 5680b 5681a 5681b 5682a 5682b 5683a 5683b 5684a 5684b 5685a 5685b 5686a 5686b 5687a 5687b 5688a 5688b 5689a 5689b 5690a 5690b 5691a 5691b 5692a 5692b 5693a 5693b 5694a 5694b 5695a 5695b 5696a 5696b 5697a

C F A1 A1 B1 C A1 A1 B1 A1 B1 B1 A1 A1 B1 A1 F D* B1 A1 B1 A1 B1 B1 B1 A1 A1 B1 A1 A1 E D B1 A1 C A1 A1 A1 B1 A1 F D* A1 A1 C A1 E A1 A1 A1 A1 A1 B1 A3 F A1

471 xxxxxP#px Iudeo#Px px#xxxpxx Pxx#Px xxxxxPxx#P xxxPsx PXx#xPx P#xPx xxxxxPxx#P Px#xPx xxxPx#P xxPx#P xPx#Px xpx#xxPx xxxpxx#xP P#xxxxpXx px#Iudeono Px#PS#P xxxxxPx#px xxPx#xxxPx xxxxPx#xP Px#xPx xxPxx#xP xxPx#P xxpX#xxP xpxx#Px xxxxxxxPx#xPx xxPx#P Px#Px xxPx#xPx PXx#px px#PXx xxxxPx#P Px#xPx xxxxxP#px PXxx#Px xxxPXSx P#xxxPx xxxxPx#P Px#xpxx xxpx#Iudeono xxxPx#xxxxPxx#P Px#xPx Pxx#Px xxxP#px xPx#xPx PSx#P xxxxPSx#Px xxxPxx#Px Px#Px pxx#Px Px#xPx xxxpx#xP xxxxPx Iudeo#Px Px#xPx

MOSP-Index 1.qxd

5/11/04

472 5697b 5698a 5698b 5699a 5699b 5700a 5700b 5701a 5701b 5702a 5702b 5703a 5703b 5704a 5704b 5705a 5705b 5706a 5706b 5707a 5707b 5708a 5708b 5709a 5709b 5710a 5710b 5711a 5711b 5712a 5712b 5713a 5713b 5714a 5714b 5715a 5715b 5716a 5716b 5717a 5717b 5718a 5718b 5719a 5719b 5720a 5720b 5721a 5721b 5722a 5722b 5723a 5723b 5724a 5724b 5725a

2:49 PM

Page 472

Index to the scansion of the Heliand B1 A1 C B1 A1 A1 B1 A1 B1 B1 D A1 B1 or E C B1 A1 E B1 A1 B1 A1 D* A1 B1 A1 A1 B1 B1 A1 E* B1 B1 A1 A1 C B1 B1 C A1 A1 B1 A1 B1 F F D*A1 B1 D* F A1A1 CA1 ED A1 A3 C A1

xxPx#px P#xPx xxxP#Px xxxxPx#P Px#Px Px#xxxxpxx xxxxxpxx#xP xPx#xxPx xPx#P xPS#P xpx#Pxx xpx#xxxPx xxxxxxPXx#P xPsx xpxx#P P#xxPx xpxSx#P xPx#P Px#Px xxPx#P Px#Px xpx#xPsx xPx#xPx xxxP#xxpx Px#Px Px#xxxPx xxxPx#P xxxxxxPx#P Px#Px pxx#Px#xpx xxxxxPx#P xxxPx#P Px#Px Px#Px xpXSx xxPx#px xxxxxPxx#px xxP#px Px#Px pxx#Px xxxPx#px xPx#xPx xxxxxPx#P Iudeono#PSx Ioseph#xxPx PXx#xxxxPxx#Px xxxxxxxPx#P Px#xxxxpXSx xxxxxxxPx#Iudeono Px#pxx#Px xxxxxxxpxsx#Px Px#xxxpx#Pxx pxx#xxPx xxxxxPx xxPsx Px#xxxPx

5725b 5726a 5726b 5727a 5727b 5728a 5728b 5729a 5729b 5730a 5730b 5731a 5731b 5732a 5732b 5733a 5733b 5734a 5734b 5735a 5735b 5736a 5736b 5737a 5737b 5738a 5738b 5739a 5739b 5740a 5740b 5741a 5741b 5742a 5742b 5743a 5743b 5744a 5744b 5745a 5745b 5746a 5746b 5747a 5747b 5748a 5748b 5749a 5749b 5750a 5750b 5751a 5751b 5752a 5752b 5753a

B1 A1 C A1 B1 A1 B1 A3 C A1 B1 D A1 A1A1 A3A1 A1 B1 D* B1 D* B1 C A1 A1 B1 A1 B1 C A1 A1 B1 D A1 D* A1 D B1 E C B3 A1 A3 A1 D* F D* B1 A1 F B1 A1 A1 B1 B1 A1 A1

xxxPx#P xPx#xxPx xxxP#Px Px#xpxx xxxxxPsx#P Px#xPx xxxP#xP xxxxxxPx xxxxxP#px Px#xxPx xxxxxpx#P P#PXx Px#Px P#xxxxxxPx#Px xxxxxPx#xPx xP#xxxxPx xxxPx#P Px#Psx xxxxxPx#xP P#xxPSx xxxPx#P xxxpx#Px xxxPx#Px Px#xPx xxpxx#P px#xxPx xxxpx#P xxxPSx Px#PX Px#xPx xxxxxpxx#xpx xxxpx#PSx PXx#Px pxx#Psxx xxP#xPx xpx#Px#P xxxxxPx#px PXx#P xxpx#Px xxxxxpx Px#Px xxxxpxx pxx#xPx pxx#Psx Maria#xxxPxx pxx#Psxx xxPX#px P#xPx PS#Iudeono xxpxx#P pxx#xPx pxx#xPx xxxxxPx#P xxxxPx#P P#xPx px#xPx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 473

Index to the scansion of the Heliand 5753b 5754a 5754b 5755a 5755b 5756a 5756b 5757a 5757b 5758a 5758b 5759a 5759b 5760a 5760b 5761a 5761b 5762a 5762b 5763a 5763b 5764a 5764b 5765a 5765b 5766a 5766b 5767a 5767b 5768a 5768b 5769a 5769b 5770a 5770b 5771a 5771b 5772a 5772b 5773a 5773b 5774a 5774b 5775a 5775b 5776a 5776b 5777a 5777b 5778a 5778b 5779a 5779b 5780a 5780b 5781a

B1 A1 A1 B1 D* A1 B1 C B1 A1 B1 A1 B1 A1 B1 A3 F A1 B1 C C A1 B1 F C A1 A1 A1 B1 B1 A1 A1 B1 C C A1 B1 C A1 E* B1 D* B1 A1 D* A1s A1 B1 A1 A1 B1 A1 A1 C F A1

xxxxPx#P Px#xPx xxxxxxxxxxPx#xPx xPx#px xP#xPx#xpx xpx#xxxPx xxxPx#P xxxpx#Px xxxxPx#P xxpx#xxxPx xxxxxxxxPx#P Px#xPx xxxxPx#P P#xPx xxxxxxPx#P xxxxxpxx xxxpx#Iudeono px#xxxPx xxxxxxxxPx#px xxpx#Px xxxxxpx#Px Px#Px xxPxx#P Iudeono#xPx xxxxxpx#Px px#xxxPx Px#Px px#xxxxPx xxxpXx#P xxPXS Px#Px Px#xPx xxxP#xP xxxxP#px xpx#Px Px#Px xxxxPx#P xxxPsx P#xxxpxx pxx#Px#xpx xPx#px xPx#xPsx xxxpxx#P xpX#xxxxpXx Px#P#xP pxS#px P#xxxxPx xxPx#P P#xxPx Px#Px xxxxPx#xP xP#xxxPx Px#Px xxxP#Px Iudeo#Px px#xxxPx

5781b 5782a 5782b 5783a 5783b 5784a 5784b 5785a 5785b 5786a 5786b 5787a 5787b 5788a 5788b 5789a 5789b 5790a 5790b 5791a 5791b 5792a 5792b 5793a 5793b 5794a 5794b 5795a 5795b 5796a 5796b 5797a 5797b 5798a 5798b 5799a 5799b 5800a 5800b 5801a 5801b 5802a 5802b 5803a 5803b 5804a 5804b 5805a 5805b 5806a 5806b 5807a 5807b 5808a 5808b 5809a

D D* A1 C E F B1 A1 A1 A1 B1 C A1 D A1 A1 D C C B1 C A1 C A1 B1 A1 C A1 C C B1 D* A1 D* B1 A1 A1 A1 F A1 C D A1 D* D E* C A1 B1 D* B1 C A1 A1 B1 A1

473 P#PSx Px#Px#P Pxx#pxx xxpx#Px PSx#P Mariun#pxSx xxPx#px xP#xxPx pXx#xxPx Px#xxPx xxxxxPx#P xxxxPsx Px#Px px#Pxx Px#Px Px#pxx xP#pxxx xxxpx#Px xxpx#Px xxxxPx#P xxxpx#Px xpXx#xPx xxxxxP#Px xPx#Px xxxxPsx#P xPx#xxxpxx xxP#Px xPx#xxPx xxxxpx#Px xpx#Px xxPx#P Px#xpxSx pxx#xpxx px#xpXsx xxxPx#xP xPx#pxx xxxPx#Px xPx#Px Px#Iuðeono xPx#xxpXx xxxxxpx#Px P#Pxx#P Px#xPx xxPx#Psx P#P#xP xPx#P#xxpx xxxxpx#Px xpxxx#xPx xxxxxxPx#xP PS#Pxx#px xxxxPx#xP xxPSx xxxxxxxxxxPx#Px xpX#xxxPx xxPx#P P#xxxPx

MOSP-Index 1.qxd

5/11/04

474 5809b 5810a 5810b 5811a 5811b 5812a 5812b 5813a 5813b 5814a 5814b 5815a 5815b 5816a 5816b 5817a 5817b 5818a 5818b 5819a 5819b 5820a 5820b 5821a 5821b 5822a 5822b 5823a 5823b 5824a 5824b 5825a 5825b 5826a 5826b 5827a 5827b 5828a 5828b 5829a 5829b 5830a 5830b 5831a 5831b 5832a 5832b 5833a 5833b 5834a 5834b 5835a 5835b 5836a 5836b 5837a

2:49 PM

Page 474

Index to the scansion of the Heliand A1 A1 B1 D*A1 A3A1 A1A1 A1 A1A1 A1 C C E A1 C A1 A1 B1 A3 A1 E F A3 A1 F C A1 B1 B1 C A1 B1 A1 C A1 C B1 A1 A1 A1 E B1 C A1 D* C C A1 A1 A1 B1 A1 A3 F A1s A1 D

PXSx Px#xPX xxxxxxPx#P xP#xxxxPxx#Px xxxxxPx#Px xpxx#xxxpxx#xPx P#xxxxPx xP#xpXx#pxx Px#xxPx xxxpx#Px xxxpx#Px PXx#px Px#Px xxxxxPXx P#xPx P#xxPx xxxxPx#px xxxxxxPx xxxxxxPx#Px PXx#P xNazareth-S xxxxpxx xxxPx#Px Iudeo#Px xxxP#Px PxSx xxxPx#P xxxxPx#P xxxxxpx#Px xP#xxxPx xxxxPx#px Px#pxx xxxxxxP#px xxxPx#Px xxxxpx#Px xxPsx#P Px#Px xpx#xxxPx Px#pxx pxSx#P xxPS#px xxPXx xxxxxPx#Px Px#xpxSx xxP#px xxpx#Px xxxxxxPx#Px Px#xxxPx Px#Px xxxPx#P xPx#xPx xxxPx Simon#Petruse PS#px Px#Px px#Pxx

5837b 5838a 5838b 5839a 5839b 5840a 5840b 5841a 5841b 5842a 5842b 5843a 5843b 5844a 5844b 5845a 5845b 5846a 5846b 5847a 5847b 5848a 5848b 5849a 5849b 5850a 5850b 5851a 5851b 5852a 5852b 5853a 5853b 5854a 5854b 5855a 5855b 5856a 5856b 5857a 5857b 5858a 5858b 5859a 5859b 5860a 5860b 5861a 5861b 5862a 5862b 5863a 5863b 5864a 5864b 5865a

C F B1 A1 B1 A1 C A1 B1 A1 C A1 B1 A1 A1 A1 B1 C B1 A1 B1 E D B1 A1 D A1 A1 B1 C B1 C B1 A3 B1 A1 B1 F C C A1 E A1 A3 A1 A1 B1 B1 A1 A1 B1 A1 C D* B1 A3

xxP#Px xxGalileo#P xxxxxPx#px xpx#xxPx xxxxPx#xP Px#Px xxxxP#px Px#Px xxxxxPx#P Pxx#Px xpxSx Pxx#xPx xxxxxxxxPx#P PXSx px#xxPx xpxx#xpxx xxxxxPxx#px xxxpx#Px xxxPxx#xP xPx#xpXx xxxxP#xpx PXx#px xxxP#Pxx xxxPx#px px#xPx px#Pxx Px#Px pxx#Px xxxxxPx#P xxxPsx xxxxPx#P xxPsx xxxPx#px xxxxxxxPx xxxxPx#P Px#Px xxxxxxPx#P xGalilea-Sx xxxxxpx#Px xP#Px xPxxx#Px PXxx#P Px#Px xxxxpxx xxxPx#Px Px#xPx xxxxxxxxPxx#P xPx#px Px#xPx PXx#xPx xxxxxxPx#P xpx#xpxx xxxxP#px Px#PSx xxxxxxPx#P xxxxxxpxx

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 475

Index to the scansion of the Heliand 5865b 5866a 5866b 5867a 5867b 5868a 5868b 5869a 5869b 5870a 5870b 5871a 5871b 5872a 5872b 5873a 5873b 5874a 5874b 5875a 5875b 5876a 5876b 5877a 5877b 5878a 5878b 5879a 5879b 5880a 5880b 5881a 5881b 5882a 5882b 5883a 5883b 5884a 5884b 5885a 5885b 5886a 5886b 5887a 5887b 5888a 5888b 5889a 5889b 5890a 5890b 5891a 5891b 5892a 5892b 5893a

C F B1 A1 B1 A1 C D* B1 A1 C C A1 A1 C A1 C F C B1 B1 A1 F C A1 A1 A1 A3 B1 A1 B1 F A1 D* B1 A1 B1 A1 B1 A1 B1 A1 B1 D* B1 A1 B1 A1 A1 A1 B1 A1 A1 D*A1 B1 A1

xxxxP#px xGalileo#P xxxxxPx#px xpx#xxPx xxP#xxP xPx#xPx xxxxxxxP#px Px#xP#px xxxxxpxx#xP xxpXx#xpxx xxxxP#px xxpx#Px xxxxxPx#Px PS#xPx xxpxXx px#xxxxPx xxxxxxP#pxx Iudeono#Px xxxxxpx#Px xxPx#P xxxPsx#P Px#xPx xxxxxPx#Iudeono xxxxxPSx pxx#Px PS#xPx Px#xPx xxxxP#x xxxPxx#P xpx#xxxPx xxxxPx#px Iudeo#Px P#xxpxx Px#xP#px xxxxxxPx#P xPxx#xxpxx xxxxxPx#px xpxx#xPx xxxxxxxxPx#P xPx#xxxxPx xxxxxxPx#xP P#xPx xxxxxPsx#P xxPx#xxxxPXx xxxPx#P Px#xxPx xxxxxxPx#px Pxx#Px Px#xxxxPx xxPx#xxPx xxxPx#P xPx#xxxPx xxxxxpxx#Px xPx#xxPxx#Px xxxxPx#px Px#Px

5893b 5894a 5894b 5895a 5895b 5896a 5896b 5897a 5897b 5898a 5898b 5899a 5899b 5900a 5900b 5901a 5901b 5902a 5902b 5903a 5903b 5904a 5904b 5905a 5905b 5906a 5906b 5907a 5907b 5908a 5908b 5909a 5909b 5910a 5910b 5911a 5911b 5912a 5912b 5913a 5913b 5914a 5914b 5915a 5915b 5916a 5916b 5917a 5917b 5918a 5918b 5919a 5919b 5920a 5920b 5920’a

B1 D* A1 C F C C F B1 B1 F D B1 C C A1 A1 A1 C A1 C B1 A1 D* B1 F C D B1 A1 C D* A1 A1 C F A1 A1 B1 A1 A1 E B1 F B1 A1A1 A3A1 A1A1 A3A1 B1A1 A3A1 B1A1 B1 A1 C A1

475 xxxxxxPx#P Px#xP#px xxxxxxxPx#Px xxxxpx#Px Iohannes#xxPetrus xxpXSx xP#px Iohannes#xPx xxxxxxpx#xP xxxP#xxP Simon#Petrus P#PXS xxxxP#xP xxP#Px xxxxpx#Px PxSx Px#Px Px#Px xxxxxPsx pxx#xPx xxpx#Px xxxxPx#xpx Pxx#Px Px#Pxx xxxxxxPx#P xP#xxIohannes xxP#Px px#PS#P xxP#xxP xpx#xxPx xxxxxxxxxxP#px xPx#PSx xPx#xPx P#xPx xxxxxP#px Iohannes#xxPetrus xxxxxPx#Px xxPx#xPx xxPXS P#xxpxx PXSx PXx#xxxpx xxxPx#P Maria#xxMagdalena xxxPxS px#xpxx#xPx xxxxxxPx#Px xxPx#xxxxxxxPx#xPx xxxxxPx#xPx xP#xxxP#xPx xxxxxxxPx#Px xPx#xxxxxPxSx xxxxxxPxx#P Px#Px xxxxPSx xPx#xPx

MOSP-Index 1.qxd

5/11/04

476 5920’b 5921a 5921b 5922a 5922b 5923a 5923b 5924a 5924b 5925a 5925b 5926a 5926b 5927a 5927b 5928a 5928b 5929a 5929b 5930a 5930b 5931a 5931b 5932a 5932b 5933a 5933b 5934a 5934b 5935a 5935b 5936a 5936b 5937a 5937b 5938a 5938b 5939a 5939b 5940a 5940b 5941a 5941b 5942a 5942b 5943a 5943b 5944a 5944b 5945a 5945b 5946a 5946b 5947a 5947b 5948a

2:49 PM

Page 476

Index to the scansion of the Heliand A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 A3A1 A3A1 D A1A1 CA1 A1A1 CA1 D A3A1 A1A1 A3A1 A1A1 A3A1 A1A1 B1 A1A1 A3A1 A1 B1 or E D* A1 A1 B1 A1 A1 E A1 A1 A1 A1 C D* A1 A1s B1 A1 B1 A1 B1 A1 A1 D* A1 D* B1 D

xxxxxPx#Px Px#xxxPx#Px xxxxxxxPx#xPx xPx#xPx#Px xxxxxxxxPx#xPx xPx#xxxPx#Px xxxxxxPx#Px P#xxxxPx#xPx xxxxxxxxpxx#xPx Px#xxxxPx#Px xxxxxxxPxx#Px xxxxPx#xPx xxxxxxxpx#Pxx Px#xPxx#Px xxxxxxPxx#Px PS#Px#Px xxxxxPxx#Px xpx#PXxx#P xxxxPx#Px xP#xPx#Px xxxxxPX#Px Px#xxxxxPx#xPx xxxxxxxPx#Px xPx#xxxPx#Px xxxxxpxS#px pxx#xPx#Px xxxxxxPx#xPx Px#xPx xxxx(xx)xxpxXx#px xPx#xxPSx xxxPx#Px Px#Px xxxxPxx#px pxSx Px#xxPx PSx#P Px#Px xP#xxxPx xxxxxxxPx#Px Px#xxxPx xP#px xpx#xxPXx xxxPx#Px PS#px xxPX#P xPx#xPx xxxxxxxxPx#xP pXxx#Px xxxxxPxx#xP xPx#Px xxxxPS#Px PXx#xxxpx#px xxxxxPXSx pxx#PXx xxxPxx#P P#pXSx

5948b 5949a 5949b 5950a 5950b 5951a 5951b 5952a 5952b 5953a 5953b 5954a 5954b 5955a 5955b 5956a 5956b 5957a 5957b 5958a 5958b 5959a 5959b 5960a 5960b 5961a 5961b 5962a 5962b 5963a 5963b 5964a 5964b 5965a 5965b 5966a 5966b 5967a 5967b 5968a 5968b 5969a 5969b 5970a 5970b 5971a 5971b 5972a 5972b 5973a 5973b 5974a 5974b 5975a 5975b 5976a

A1 A1 B1 A1 C D* B1 A1 B1 A1 B1 C B1 F B1 F A1 B1 A1 D* F A1 C C C A1 B1 D* C D* B1 B1 C D* B1 D B1 D* B1 F F A1 C A2 B1 A1 C A3 F D* A1 A1 D* A1A1 B1 E

PXSx Px#xPx xxxxPx#xP xPx#xPx xxxxxpx#Px Px#xPSx xxxxxPx#px Px#xxPx xxxxpXx#px xPx#xxxPx xxxxPx#px xxpx#Px xxpx#xxP xGalileo#P xxxxxPx#P xxxxxxHierusalem xxPxx#Px xxPx#px P#xPx Px#xxxPXx xxxxEmaus xPx#Px xxxxxpx#pxx xxxpx#Px xxxxxpx#Px xpxx#xxxxPx xxxxxPxx#P PXx#px#px xxxxxxpxSx xPx#PXx xxxxxxxxPx#xP xxxxxxxPx#px xxxxxxxxxxxpx#Px xPx#xxPXx(xx) xxPx#px px#pxxx#P xxxxP#xpx xPx#PSx xxxxPx#P(xxx) xxxHierusalem Iudeono#Px Pxxx#Px xpXSx xxPx#px#P xxPxx#P Px#Px xxxP#px xxxxPx xBethania xP#xxPx#P xxPxxx#xPx Pxx#xxxPx xP#xxP#px xxxxxPx#pxx#Px xxxxxPxx#P px#xxP

MOSP-Index 1.qxd

5/11/04

2:49 PM

Page 477

Index to the scansion of the Heliand 5976b 5977a 5977b 5978a 5978b 5979a 5979b 5980a

D* E B1 E B1 B1 A1 C

xxPx#P#px pxSxx#px xxxxP#xpx PXx#P xxxxpX#xpx xxxxPx#px xPx#Px xpx#Px

5980b 5981a 5981b 5982a 5982b 5983a 5983b

C F A1 D* B1 A1 E

477 xxxxxP#px xxHierusalem Pxx#Px Px#pxXx xxPS#px Px#xxxxxPx PXx#P

MOSP-Index 2.qxd

5/11/04

2:50 PM

Page 478

Index to the scansion of the Old Saxon Genesis* 1a 1b 2a 2b 3a 3b 4a 4b 5a 5b 6a 6b 7a 7b 8a 8b 9a 9b 10a 10b 11a 11b 12a 12b 13a 13b 14a 14b 15a 15b 16a 16b 17a 17b 18a 18b 19a 19b 20a 20b 21a 21b *

F A1 B1 B1 D B1 A1 C A1 A1 B1 A1 A3 B1 C A1 D B1 A1 B1 A3 A1 A1 B1 A2 or D* B1 A3 B1 C A1 A1 D B1 A1 D B1 A1 A1 A1A1 C A1 B1

xxxxxEua#xxxAdam pxx#xpxx xxxPxx#P xxxxxxPx#P px#Pxx xxxPx#P pxx#xPx xpXSx xP#xxxxPx xxxxxPx#PX xxxxxxxPx#P Px#Px xxxxPx xxxxpx#xP xxPXx P#xPx pXsXx xxPx#px pxx#xxPx xxxPx#xP xxxxxPx Px#Px pxx#Px xxxxxPx#xxP Px#pxsx xxxxxxPxx#P xxxxxPx xxxxxxxxxPx#px xxxxP#px Px#xxPx Px#xxPx xpx#P#Px xxPx#P pxx#xPx px#P#xxP xxpxx#P Px#xxxpxx Px#Px Px#xpXx#Px xxxpx#Px xpx#xxPx xxxP#xpx

22a 22b 23a 23b 24a 24b 25a 25b 26a 26b 27a 27b 28a 28b 29a 29b 30a 30b 31a 31b 32a 32b 33a 33b 34a 34b 35a 35b 36a 36b 37a 37b 38a 38b 39a 39b 40a 40b 41a 41b 42a 42b

A1 B1 A1 B1 A1 B1 A3 E – – A1 B1 A1 C B1 D A1 A1s A1 E A1 B1 A1 B1 E F A1 A1s or E E A1 A1 B1 A1 B1 B1 A1 A1 A1 C A1 A1 B1

xxpxx#xxPx xxxPx#P xpx#xPx xxxxxxPXx#P PX#Px xxxxxPx#P xxxxPx xxxxxpXSx#P PX . . . ... Pxx#xxxpxx xxxPx#xpX Px#xxPx xxxxxpx#Px xxxPx#px PSxxx Px#Px pXS#px px#xPx xxxP#Px#P PX#xxPx xxP#xxpx xpx#xPx xxxxxxxxPx#P PSx#px xxxxKain#xpx Px#xxxPx pXsx#px PSx#xpX xpX#xxPx xpx#xPx xxxpxx#xP(xx) Px#xxPx xxxP#xxP xxpxx#P Px#pXx Px#xxxxpXx Px#xPx xxxpx#Px Px#Px xPx#xPx xxxxxxPx#P

For symbols and notations, see Symbols and Abbreviations, and section 1.4 in the text.

MOSP-Index 2.qxd

5/11/04

2:50 PM

Page 479

Index to the scansion of the Genesis 43a 43b 44a 44b 45a 45b 46a 46b 47a 47b 48a 48b 49a 49b 50a 50b 51a 51b 52a 52b 53a 53b 54a 54b 55a 55b 56a 56b 57a 57b 58a 58b 59a 59b 60a 60b 61a 61b 62a 62b 63a 63b 64a 64b 65a 65b 66a 66b 67a 67b 68a 68b 69a 69b 70a 70b

A3 B1 C B1 B1 B1 A1 B1 A1 C A1 A1 D A1 D C DA1 A3A1 A1A1 A3A1 A1 C D* B1 E B1 F A1 B1 C A1 B1 A1 B1 C C D* B1 B1 A1 A1 B1 B1 A1 A3 A1 A1 A1 A1A1 A3A1 A3 C A1 B1 A1 A1

xxxxxxpxx#(xx) xxxxxxpXx#P xxxpx#Px xxxxxPx#xpx xxxxxxPx#px xxPx#px Px#Px xxxxxxxxxpXx#xP Px#xPx xxxxxxpx#Px Px#xPx xP#xxxxPx P#Px#px xPx#pxx xP#PxS xpx#Px P#xxxxPxx#Px xxxxxxPx#pxx xP#xxxPXSx xxxxxP#xPx px#xpxx xpXSx xPx#pxSx xxxxxPx#px pXsx#xpx xxpXx#P Kain#xxxPx#Pxx P#xxxxpxx xxxxxxxPX#P xpXSx Px#xPx xxxxxPXx#px(xx) px#xxxPx xxxxPx#P xxxPsx xxxxxxxxxpx#Px P#xxxPXsx xxxxxpxx#xpx xxxxPx#P Pxx#Px PS#Px xxPx#px xxxxPx#xP PX#xPx xxxxPx Px#Px Pxx#xPx xxxxxxxPx#Px px#xxxxPx#Px xxxxxxxxxxxPx#Px xxxpxx xxxxxxxpx#Px xpx#xxxxPx xxxxPx#xpx pxx#PX P#xxP#x(xx)

71a 71b 72a 72b 73a 73b 74a 74b 75a 75b 76a 76b 77a 77b 78a 78b 79a 79b 80a 80b 81a 81b 82a 82b 83a 83b 84a 84b 85a 85b 86a 86b 87a 87b 88a 88b 89a 89b 90a 90b 91a 91b 92a 92b 93a 93b 94a 94b 95a 95b 96a 96b 97a 97b 98a 98b

A1A1 B1 A1 C A1 B1 A1 B1 A1 E A1 C A1A1 A3A1 A1A1 E A1 A1 D* C A1 A1 F A1s B1 C F A1 A1 C A3 C A1 B1 A1 C A1 B1 A2 B1 F B1 A1 B1 A1 B1 E B1 B1 B1 E A1 C A1 E B1

479 Px#Px#Px xxxxPx#P xpxx#xPx xxxxpx#Px Px#xxPx xxxPx#P px#xxxxpXx xxxPx#xP Px#xxxxxPx PSx#P Px#xxxPx xxxxxxP#px xPx#xxxPx#Px xxxxxxxxxxxPx#Px Px#xxPx#Px PXx#P xxPx#Px Px#xPx xP#xxxxPx#px xxxxP#Px Px#xpxx pxx#xxxxPx Adama#xxEuun PS#px xxPx#P xxxxxP#Px xxAdamas#px Px#Px Px#xpxx xxxxxpx#Px xxxxxPx xxxpx#Px P#xxxPx xxPx#P PxSx xxxxpx#Px Px#xxxxpxx xxxPx#P xpxx#PS xxxxxPx#xpx Kain#xxxxPx xxxxxxPx#xP PXxx#xxPx xxxxxxxxxPx#P PXx#xxxpXx xxxxxxxxPx#P PSx#xpx xxxxxP#xP xxxxPx#px xxxPx#P PSx#P P#xxPx xxxPXx xPx#xPx PSx#px xxxxxxxxxxxPx#xpx

MOSP-Index 2.qxd

5/11/04

480 99a 99b 100a 100b 101a 101b 102a 102b 103a 103b 104a 104b 105a 105b 106a 106b 107a 107b 108a 108b 109a 109b 110a 110b 111a 111b 112a 112b 113a 113b 114a 114b 115/6a 115/6b 117a 117b 118a 118b 119a 119b 120a 120b 121a 121b 122a 122b 123a 123b 124a 124b 125a 125b 126a 126b 127a 127b

2:50 PM

Page 480

Index to the scansion of the Genesis A3 A1 A1 A1 D* B1 D C C A1 A1 E D* D A1 A1 B1 A1 B1 F A1 B1 A1 C E C A1 B1 B1 A1 B1 A1 E B1 A1 D A1 E F A1 D* C A1 C A1 B1 D* B1 F B1 D B1 F B1 A1 B1

xxxxxPx pxxSx Px#Px pxx#xPx pxx#PSx xxxxPx#P P#pxx#P xxpx#Px xxxxPxx pxxSx Px#xxPx px#Px#P Px#PSx xP#Pxx Px#Px Px#xPx xPxsx Px#Px xxxpx#xpx xxxxSeð#xpx Px#Px xPx#P pxx#PX xxpx#Px PSx#P xxpx#Px Px#xxxxxPx xxxxPx#P xxxxxxxPx#P Px#Px xpxx#xP Px#Px px#Px#px xxxxPx#P Px#Px xP#Pxx Px#xPx xxpx#Px#P xPx#xxKaina Pxx#Px pxx#PSx xxxpx#Px Px#Px xxxPXx Px#Px xxxxPx#P Px#xpxSx xxxPx#xP px#xKaina xxxxPx#P px#P#xxP xxxPx#P Seðas#xPx xPx#P Px#xPx xxxxPx#P

128a 128b 129a 129b 130a 130b 131a 131b 132a 132b 133a 133b 134a 134b 135a 135b 136a 136b 137a 137b 138a 138b 139a 139b 140a 140b 141a 141b 142a 142b 143a 143b 144a 144b 145a 145b 146a 146b 147a 147b 148a 148b 149a 149b 150a 150b 151a 151b 152a 152b 153a 153b 154a 154b 155a 155b

A1 B1 C A1 E B1 A2 C F B1 A1 B1 A3 A1 A2 or D* B1 B1 A1 A3 B1 A1 C D A1 A1 B1 B1 A1 A1 B1 F A1 C A1 D* E D* A1 B1 A1 A1 A1 C A1 A1 B1 A3 F A1 B1 A1 C D* B1 A1 B1

Px#Px xxP#xP xxxxxP#Px Px#xPx pxSx#xP xxxPx#P P#xxPS xxxP#px Enoch#xxPx xxxPx#P Px#xPx xxxxPXS xxxxxpxx pXx#xPx PXx#xxPsx xxxxxxPx#xpx xxxpxx#xxP pxx#PX xxxxxxPx xxPx#P px#xPx xxxxxxxxpX#Px P#pxx#P pxx#Px Px#xPx xxxxPx#px xxPxS Px#Px px#xPx xxxPx#P xxEnocha#xpx Px#pxx xxxPsx pxx#xPx xP#xPx#P xxpx#Px#px px#xxPsx Px#Px xxPxS Px#xPx xPX#xPx P#xxxxpxx xpx#Px pxx#xPx Px#Px xxxxxxxP#xP xxxxxxPx Sodomo-Sx px#xxPx xxxxxPX#P Px#Px xxP#px pxx#pXSx xxxxxxPXx#P Px#xPx xxxxxPX#P

MOSP-Index 2.qxd

5/11/04

2:50 PM

Page 481

Index to the scansion of the Genesis 156a 156b 157a 157b 158a 158b 159a 159b 160a 160b 161a 161b 162a 162b 163a 163b 164a 164b 165a 165b 166a 166b 167a 167b 168a 168b 169a 169b 170a 170b 171a 171b 172a 172b 173a 173b 174a 174b 175a 175b 176a 176b 177a 177b 178a 178b 179a 179b 180a 180b 181a 181b 182a 182b 183a 183b

A1 C A1 C F B1 F A1 F C A2 or D* C A1 C B1 A1 A2 or D* B1 A1 E A1 C B1 A1 B1 A1 E B1 A1 B1 A1 B1 A1 B1 D* B1 A1 B1 C A1 A3 D A1 B1 D C D* F A1 B1 A1 A1 A1 B1 B1 C

Px#pxx xxxP#px xPxx#Px xxPXx Px#xSodoma xxxxPx#xP xxxxMambra Pxx#Px xxxxAbrahama xxxpx#Px px#xxPsx xxxxxxPXx P#xPx xxxxxpx#Px xPx#P Px#xPx xxPx#xP#px xxxpx#xP P#xxxPx xxpx#Px#P P#xxpxx xxP#Px xxxPx#P Px#Px xxPx#P PX#P#x pxSx#px xxxPx#P P#xxxPx xxxPx#xP Px#xPx xxPx#P Px#Px xxxxxxPx#P xPx#xxxPxx xxxxPx#xP P#xxPx xxxPx#P xxxpxSx Px#Px xxxxxPx px#PSx Px#Px xxxxxxPx#P(xx) px#Px#P xxpx#Px Px#xxxPx#px xxxxxSodoma-S px#xxPx xxxxxPx#xP px#xxPx xxxxPx#Px Px#xxPx xxxxPx#px xxP#xxP xxP#Px

184a 184b 185a 185b 186a 186b 187a 187b 188a 188b 189a 189b 190a 190b 191a 191b 192a 192b 193a 193b 194a 194b 195a 195b 196a 196b 197a 197b 198a 198b 199a 199b 200a 200b 201a 201b 202a 202b 203a 203b 204a 204b 205a 205b 206a 206b 207a 207b 208a 208b 209a 209b 210a 210b 211a 211b

D C A1 C A1 A1 A1 A1 E B1 F B1 A2 or D* B1 B1 D A1 B1 A1 C B1 A1 B1 A1 A1 A1 A1 B1 A1 A1 A3 E E C C B1 B1 D A1 A1 A1 B1 A1 C A3 D A1 A1 A1 C A1 B1 B1 A1 F A1

481 px#PSx xxxxPXx P#xPx xxxxxpXSx Px#xPx px#xpxx Px#xPx Px#pxx PSxx#P xxPx#px Abraham#xxpXx xxxPx#P Px#Psx xxxxxPX#P xxPx#xpx(xx) P#pXSx Px#xPx xxxxPx#P xpX#xxxPx xxP#px xxxxPXS Px#Px xxxPx#xP PX#P#x xxxxPx#Px pxx#xxPx Px#xxPx xxxPx#xP xPx#xPx Px#Px xxxxxxxxPx PSxx#P Psxx#px xxxxP#px xxPXx xxxPx#P xxxxpxx#xP P#pXSx xxxPx#Px pXxx#Px Pxxx#Px xxxxP#xP PX#xxxPx xpx#Px xxxxxPx px#PSx xxxPx#Px(xx) pXxx#Px Pxx#pxx xxpx#Px Px#xPx xxxxxxxxpx#xpx xxxxxPx#P Px#Px Abraham#xxpXx Px#Px

MOSP-Index 2.qxd

5/11/04

482 212a 212b 213a 213b 214a 214b 215a 215b 216a 216b 217a 217b 218a 218b 219a 219b 220a 220b 221a 221b 222a 222b 223a 223b 224a 224b 225a 225b 226a 226b 227a 227b 228a 228b 229a 229b 230a 230b 231a 231b 232a 232b 233a 233b 234a 234b 235/6a 235/6b 237a 237b 238a 238b 239a 239b 240a 240b

2:50 PM

Page 482

Index to the scansion of the Genesis D A1 A1 A1 B1 A1 E A1 A1 C A3 D A1 C A1 B1 A1 A1 E C A2 B1 F C F D A1 E A3 B1 A1 C A1 B1 A3 D A1 C A1 B1 A1 B1 A1 A1 A3 E A1 B1 F A1 A1 B1 A3 D A3 E

P#Pxx Px#Px xP#xxPx(xx) Px#P#x xxxPx#P Px#Px PSx#px Px#xxPx Px#xPx xxxxxP#px xxxPx P#pXSx Px#xPx xxP#Px Px#xxPx xxxpxxx#P(xx) Px#xxxxPx Px#Px PsXx#px xxxxxxxpx#Px xP#xxxPS xxxxxPx#P Px#xxSodoma xxxP#px Abraham#xxpXx pxSSx Pxx#xPx px#Px#xP xxxxxPx(xx) xxxxPx#xP P#xxPx xxxpx#Px Px#xxxxxxPx xxxxxPx#xP xxxxxxxxxxxPx P#pXSx Px#xpxx xxpx#px xpxx#xxPx xxxpx#xP Px#Px xxPx#px Px#xPx xxxxxxxP#xPx xxxpxx PSx#P Px#xxxxPx xxxxxxxP#xpx xxxxSodoma-S Px#Px Px#xxpxx xxxxPx#xP xxxxxPx px#PSx xxxpxx(xx) PSxx#P

241a 241b 242a 242b 243a 243b 244a 244b 245a 245b 246a 246b 247a 247b 248a 248b 249a 249b 250a 250b 251a 251b 252a 252b 253a 253b 254a 254b 255a 255b 256a 256b 257a 257b 258a 258b 259a 259b 260a 260b 261a 261b 262a 262b 263a 263b 264a 264b 265a 265b 266a 266b 267a 267b 268a 268b

B1 A1 B1 A1 F A1 A1 D* C A1 D D A1 D* A2 or D* E F B1 A1 C A3 C F A1 A1 B1 E B1 B1 A1 A1 B1 A1 C A1 A1 A1 C A3 E F C A1 B1 A1 C F A1 B1 B1 F C A1 B1 B1 A1

xxPx#P Px#Px xxxxxxxxxxxpXx#P pxx#Px xxxxAbraham#P Px#Px Px#Px xxP#xPx#xpx xP#Px P#xPx xP#pxxx xxpx#Pxx pXx#xxxPx xP#xP#px Px#xxPsx px#Pxx#P Pxx#xSodoma xxPx#xP PX#xxPx xxxxxxxP#px xxxxPx xxpXxx xxSodoma-S Pxx#Pxx Px#Px xxxxPx#px pXSx#xpx xxxxxPxx#P xxxxpxx#xP Pxx#Px pXS#Px xxPXx#xP Pxx#Px xxxP#Px xPx#xPx xP#xxxPx px#xPx xxxP#px xxxxxxPx pXSx#P Loth#xxPx xxP#px pXx#xxxxpXx xxxxpxx#xP Px#xPx xxpx#Px xxAbrahamas pxxSx xPx#P xxpxx#P xxGiordanas#px xPSx xpx#xxPx xxxxPX#P xxPx#P Px#xPx

MOSP-Index 2.qxd

5/11/04

2:50 PM

Page 483

Index to the scansion of the Genesis 269a 269b 270a 270b 271a 271b 272a 272b 273a 273b 274a 274b 275a 275b 276a 276b 277a 277b 278a 278b 279a 279b 280a 280b 281a 281b 282a 282b 283a 283b 284a 284b 285a 285b 286a 286b 287a 287b 288a 288b 289a 289b 290a 290b 291a 291b 292a 292b 293a 293b 294a 294b 295a 295b 296a 296b

A1 B1 A3 A1 A1 C A1 B1 A1 D D B1 A3 D C B1 A1 B1 A1 C A1 C A2 or D* B1 D* C A1 B1 D* A1 D B1 A1 E D* A1 B1 E C B1 A1 B1 F F C A1 A1 B1 A3 B1 D* B1 D B1 B1 C

xxxPx#pXX xxxxxxpxx#px xxxxxPX Pxx#Px Px#xxPx xxxpx#Px xpxx#xxPx xxxxxxxPx#P P#xxPx xxpx#Pxx pXsXx xxxxPx#xP xxxxxxxPx P#Pxx xxxxP#Px xxPx#P xxPx#xpxx xxxxPx#P Px#xxxPx xxP#Px xpx#xxPx xxxxxpx#pxx xxPx#xxxPsx xxxxPXS pxx#pXSx xxpx#Px Pxx#Px xxxPx#P P#xPx#px PXSx px#Pxx xxPx#xpx Px#xPx P#Px#P pxx#P#xP Pxx#pxx xxxxpxx#xP Psx#P xxpxSx xxxxxPxx#px xpxx#xPx xxpxx#P xxSodoma-sx xPx#xLoða xxP#px Px#Px xxxxPx#Px xxxxPx#xpx xxxxxpxx xxxxxPx#px Px#xxxxPXx xxxxxP#xP px#pXsxx xxxxxxpxx#xP xxxPx#P xxxxxxxxxpx#Px

297a 297b 298a 298b 299a 299b 300a 300b 301a 301b 302a 302b 303a 303b 304a 304b 305a 305b 306a 306b 307a 307b 308a 308b 309a 309b 310a 310b 311a 311b 312a 312b 313a 313b 314a 314b 315a 315b 316a 316b 317a 317b 318a 318b 319a 319b 320a 320b 321a 321b 322a 322b 323/4a 323/4b 325a 325b

A1 C A1 B1 A1 D* A1 E A1 A1 A3 C A1 C A1 C A3 A1 D* A1 D A1 F B1 F B1 A1 C C A1 D* B1 A1 B1 A1 B1 A1 D* D D A1 B1 A1 D A1 A1 A2 E D* A1 F B1 A2 C A1 C

483 xxxPx#Px xxxPXx P#xxPx xxxxxPx#P Px#xpxx Px#Pxx Px#xxxPx pXsXx#px Px#xxxxPx Px#Px xxxxxxPx xxxpx#Px Px#xxxxxPx xxxP#px Px#xxpxx xxxxxP#Px xxxxxxPx Px#Px xpxx#P#px Pxx#Px px#Pxx Px#Px Pxx#xSodomo xxPx#P Loth#xxxxPx xxxPx#xP pXxx#Px xxP#px xxxxP#px pxx#xPx P#xxPxx xxxpXx#xpx Px#xPx xxxpxx#xpx Px#xPx xPxx#P Pxx#Px Px#P#xP P#pxxxsx P#P#px P#xxPx xxxxxPx#P Px#xxPx px#PXx P#xxPx pxx#pxx Px#PS xP#P#xP xPx#xPSx P#xxPx Sodoma-Sx xxxxP#xxP xxxPx#xPS xxxxpx#Px Px#xPx xxxxxpXSx

MOSP-Index 2.qxd

5/11/04

484 326a 326b 327a 327b 328a 328b 329a 329b 330a 330b 331a 331b

2:50 PM

Page 484

Index to the scansion of the Genesis F A1 A3 A1 A1 B1 A1 B1 A1 B1 D C

xSodomo-S Px#xPx xxxxxxPx P#xPx PX#xxPx xxxP#xP P#xxPx xxxxxxxPx#P pxx#Px xxxxP#xP px#pXsx xxxxxxPxx

332a 332b 333a 333b 334a 334b 335a 335b 336a 336b 337a 337b

A1 F A1 A1 B1 B1 A3 B1 A1 B1 D* A1

Px#Px xxLoðas#P xP#xxxxPx Px#Px xxxxpxx#xP xxxxP#xP xxxxPx xxPx#P Px#xPx xxPXS Px#xPXsx xxxxxPx#Px

MOSP-Author index.qxd

5/11/04

2:47 PM

Page 485

Index of authors Behaghel, Otto and Burkhard Taeger 1, 2, 89, 353 Berron, Gottfried 317 Bischoff, Bernhard 1 Bliss, A. J. 356 Boor, Helmut de 4, 5 Bostock, J. Knight 4, 5, 279 Brunner, Karl 14, 16 Cable, Thomas 36 Campbell, A. 14, 16, 19 Cathey, James E. 2 Doane, A. N. 1, 353 Duncan, Edwin 161, 163 Fulk, R. D. 6, 36, 158, 214 Gallée, Johan Hendrik 12, 13, 15, 25, 26, 213 Gantert, Klaus 1, 2, 5, 6 Genzmer, Felix 2 Green, D. H. 4 Haubrichs, Wolfgang 5 Heusler, Andreas 3, 4, 6, 10, 30, 69, 156, 165, 166, 170, 229, 255, 256, 260, 262, 263, 271, 317 Hinderschiedt, Ingeborg 170, 184 Hirt, Herman 95, 99, 114, 226, 229 Hofmann, Dietrich 4, 5, 6, 11, 26, 27, 28, 30, 61, 63, 65, 73, 74, 76, 80, 82, 83, 85, 89, 90, 94, 95, 96, 97, 99, 105, 106, 108, 109, 112, 113, 115, 116, 117, 125, 126, 129, 136, 145, 146, 149, 156, 157, 159, 161, 163, 165, 166, 170, 194, 195, 197, 202, 204, 206, 207, 259, 263, 266, 270, 296, 297, 299, 300, 303, 304, 305, 306, 307, 308, 309, 311, 312, 316, 317, 319, 320, 322, 346, 347, 348, 349, 350 Holthausen, Ferdinand 12, 13, 24, 25, 26, 213, 363 Howell, Robert B. 17, 18 Hutcheson, B. R. 32, 266 Jones, Charles 17

Kartschoke, Dieter 4 Kauffmann, Friedrich 26, 27, 30, 61, 67, 68, 73, 76, 80, 88, 90, 94, 95, 96, 97, 99, 109, 111, 113, 124, 125, 126, 129, 136, 145, 149, 156, 157, 159, 161, 163, 165, 195, 199, 200, 201, 204, 223, 346, 347, 348, 350 Kendall, Calvin B. 178 Klaeber, Frederick 6 Krahe, Hans and Wolfgang Meid 22, 213 Krapp, George Philip 349 Krapp, George Philip and Elliott van Kirk Dobbie 349 Kuhn, Hans 115, 178, 281, 282 Lehmann, Winfred P. 4, 5, 11, 22, 59, 235, 255, 259, 260, 262, 286, 288, 289, 290, 291, 293, 294 Luick, Karl 14, 16 Lutz, Angelika 17 Martin, Ernst 73, 88, 90, 94, 95, 109, 124, 166, 263 Mayer, Chr. Aug. 291 Mitchell, Bruce 283, 284, 286 Momma, H. 178, 281, 282 Murphy, G. Ronald 2, 3, 155, 343 Murray, Robert W. and Theo Vennemann 22, 23 Neuner, Erich 269 Petersen, William L. 2 Priebsch, Robert 1, 2 Prokosch, Eduard 22, 213 Rathofer, Johannes 3, 343 Rauch, Irmengard 11, 13, 24, 289 Rieger, Max 180, 208, 235, 255, 256, 261, 262, 270, 271, 317 Russom, Geoffrey 11, 27, 32, 33, 36, 37, 40, 41, 65, 72, 74, 88, 89, 201, 207, 243, 305, 308 Scott, Mariana 2 Sehrt, Edward H. 283, 349

MOSP-Author index.qxd

486

5/11/04

2:47 PM

Page 486

Index of authors

Sievers, Eduard 1, 11, 25, 30, 36, 61, 65, 66, 73, 88, 90, 94, 95, 99, 112, 113, 124, 153, 159, 165, 173, 196, 209, 223, 235, 255, 256, 296, 317, 347, 349 Sowinski, Bernhard 2, 3 Stapel, Wilhelm 2 Stübiger, Martin 262 Suzuki, Seiichi 6, 7, 10, 12, 16, 17, 22, 24, 27, 31, 32, 36, 46, 55, 56, 57, 58, 62, 63, 65, 69, 70, 71, 79, 85, 86, 90, 93, 95, 97, 100, 102, 110, 112, 113, 123, 130, 132, 134, 140, 141, 142, 143, 145, 149, 151, 152, 153, 161, 167, 172, 173, 174, 177, 178, 195, 196, 198, 199, 200, 201, 202, 204, 208,

209, 210, 211, 212, 213, 214, 222, 223, 226, 230, 232, 234, 235, 254, 256, 260, 263, 268, 269, 272, 273, 274, 277, 278, 279, 287, 293, 295, 302, 317, 323, 324, 325, 326, 328, 337, 339, 341, 343, 355, 356, 363 Taeger, Burkhard 1, 3 van Coetsem, Frans 23, 330 Vennemann, Theo 21, 22, 287 von See, Klaus 236 Weringha, J. J. fon (⫽ Weringh, J. J. van) 2 Zanni, Roland 5

MOSP-Subject Index.qxd

5/11/04

2:53 PM

Page 487

Index of subjects Alliteration, 10, 235–94 and anacrusis, 166, 239–43 in Beowulf, 236, 238, 239, 248, 251–2, 253–4, 262, 263, 266, 268, 269, 287, 290–1, 292 canonical pattern (aax) of, for heavy verses, 136, 221, 266, 270, 273, 274, 275, 277, 278, 338, 339, 344, 363 consonant, 287 consonant ⫹ vowel, 288–94 in Beowulf, 290–1 in Edda, 290 in Genesis, 364–5 in Hildebrandslied, 290 motivation for, 293–4 in Genesis, 364–5 of heavy verses, 221–2, 263–78, 338–9 lexical basis of, 255–86 (see also Word classes) and lift formation, 279, 293–4, 339–40 metrical basis of, 235–55 and metrical types, 237–9 by nature, 112, 255, 260, 261, 262 (see also Lift formation) phonological basis of, 287–94 by position, 235, 255, 260, 263 (see also Lift formation) reorganisation of, 254–5, 271–4, 275, 293–4, 339–40 and resolution, 207, 243–55 Rule for, 287, 292 single vs. double, 235, 236 syntactic basis of, 255–86 triple alliteration, 237, 295 vowel, 287, 291–2, 294, 340 Anacrusis, 10–11, 160–75, 175, 186, 329, 335–6 and alliteration, 166, 239–43 in Beowulf, 97, 160, 161, 166, 167, 169, 171–2, 239, 329 in the b-verse, 163, 166 derivation of, 176–7 as a full metrical position, 166–7, 169–70, 344 in Genesis, 360–2

in hypermetric verse, 328, 329 realised by independent words, 165–6, 174 reorganisation of, 160, 172–5, 178, 329, 335, 336, 344 as a site for accommodating sentence particles, 166, 174, 282–3, 285 in subtype A1s, 92 in type A1, 93, 161–3, 282–3, 299, 308, 311, 312–13, 329 in type A1A1, 299 in type A2, 161 in type D, 163–5, 167 in type D*, 163–4 in type E, 165, 173–4 in type E*, 165 and verse distribution, 167–71, 180 Analogical extension, 93, 152, 154, 160, 174, 178, 211, 239, 272, 273, 339, 342. See also Category extension Andreas, 349 Beowulf. See also under relevant subjects as a precursor of the Heliand, 6–7 Book epic, 3–7 Bow style, 3–4, 5 Breaking, 16–22, 330 C manuscript, 1, 25, 88, 89, 146, 206, 207, 289, 300, 351 Category extension, 56, 70, 101 Christ, 349 Clause-initial vs. non-clause-initial, 280–6. See also man in pronominal use Cliticisation, 177, 178, 335 encliticisation, 141, 142, 177, 335 procliticisation, 171–2, 173, 174, 177, 335 Coda Law, 21 Compound, 5, 25, 59, 60, 181, 305, 315. See also under individual metrical types and subtypes; Quasi compound (De)categorisation between basic and increased metrical types, 152, 172–3, 174, 178, 196, 201, 236, 254, 331, 335, 340, 344

MOSP-Subject Index.qxd

5/11/04

2:53 PM

488

Page 488

Index of subjects

(De)categorisation (continued) between heavy and normal drops, 56–8, 60, 70–2, 102–3, 136, 151–2, 181–2, 183, 272, 294, 331, 339, 340, 344 between lift and drop, 293, 331, 337, 338, 339, 340, 344 Defooting, 11, 12–13, 23, 102, 151, 330. See also Foot; Restoration of syncopated vowels Demotion, 176. See also Expansion; Promotion Disyllabic ending, 26, 146, 187–90, 213, 307 in the final drop of type A3, 80, 117 in the first drop of type A1, 36–40, 47, 48–50, 332 in the second drop of subtype D1, 126 in the second drop of type A1, 73–5, 144 in the second drop of type C, 124 in the second drop of type E, 188–90 Drop. See Heavy drop; Normal drop Elene, 349 Expansion, 176, 177. See also Demotion; Promotion Foot, 12, 13, 24, 32, 33, 49, 50, 53, 56, 57, 58, 60, 102, 103, 111, 121, 132, 146, 147, 148, 151, 152, 188, 189, 190, 208, 209, 211, 212, 215, 255, 256, 293, 330, 331, 332, 335, 336, 337, 339. See also Defooting lexical vs. nonlexical, 12, 54, 58, 104, 200, 205, 206, 333 monosyllabic vs. disyllabic, 24, 121, 194, 209, 255, 336, 337 unfooted, 12, 47, 48, 50, 53, 56, 59, 60, 102, 104, 189, 293, 330, 332 Foreign names, 27, 345–52 Dauid, 346 Emaus, 346 Erodes, 346–7 Eua, 347 Iacob, 347 Iesus, 347 Iohannes, 347 Iudas, 347–8 Iudeo, 349–52 Maria, 27 Mattheus, 346 Peter, 348 Rumu, 348 Satanas, 348 Simon, 348 Genesis, 1, 353–65. See also under relevant subjects

Gradation, 341–3. See also Parameter based on the first drop of type A1, 46–61, 71, 102–3, 154, 332 based on the first drop of type B1, 80–7 based on the first drop of type C, 113–14 based on the first drop of type D, 131–4, 333 based on the first drop of type D*, 144, 334 based on the first drop of type E, 149–55, 334 based on the second drop of type B1, 104–5, 110–11, 333 based on the second drop of type D, 133–4, 333 based on the second drop of type D*, 137, 333–4 based on the second drop of type E, 154, 155, 334 based on the second lift of type C, 117–23, 333, 342 based on the second lift of type D, 125–6, 333 Heavy drop, 8, 27, 33, 70–1, 176, 181, 234, 332. See also Normal drop prototype of, 10, 75, 112, 147, 148, 190, 193, 330–1 of type A2, 68, 71–2, 176 of type E, 153 Heavy verse, 132, 221, 295, 318, 338–9 with alliterative pattern aax for the a-verse, 264–6 with alliterative pattern axa for the a-verse, 267–8, 269 with alliterative pattern axx for the a-verse, 221–2, 266–7 alliterative patterns of, 221–2, 263–78, 338–9 in Beowulf, 135–6, 266, 268–9, 271, 273, 363 in Genesis, 363–4 of subtype A1s, 269, 272, 274–5, 339 of type A1, 269, 272 of type A2, 217–20, 221, 222, 233 of type D, 128, 129, 136–7, 274, 275–7 of type D*, 137, 139, 143, 217–20, 221, 222, 269, 272, 277–8 of type E, 149–50, 151, 269, 273, 276–7, 339 of type E*, 156, 269, 273, 277–8, 339 Heliand, literary background of, 1–7. See also C manuscript; M manuscript High Vowel Deletion, 12, 212 Hildebrandslied, 5, 290 Hook style. See Bow style

MOSP-Subject Index.qxd

5/11/04

2:53 PM

Page 489

Index of subjects Hypermetric verse, 8, 295–329, 336, 344. See also Heavy verse; Verse, normal alliteration in, 295, 316, 318 alternative scansion as subtype D2b, 306–7 alternative scansion as subtype D*2b, 304–6 alternative scansion as type A1, 307–13, 319–20, 329 anacrusis in, 328, 329 asymmetry between the a-verse and the b-verse, 319–20, 322–3, 328–9 in Beowulf, 324–6, 327–9, 336 defining features of, 295, 303, 319–20 prototype of, 301, 305, 315–16, 318, 319, 323, 325, 326, 327 reorganisation of, 323–9, 336 structure of, 314–16, 326 type A1A1, 296–9, 301, 302, 317, 324–5, 327–8 type A1⫹C, 299–300 type A3A1, 307–10, 317, 324–5, 327, 328–9, 336 type B1A1, 300–1, 325 type B1C, 301 type B1D, 302 type B3A1, 311–12 type CA1, 312–13 type DA1, 302 type DD*, 302 type D*A1, 302–3, 325 type EA1, 303 type ED, 303

489

Linking word, 296, 297, 302, 314–15, 326–7. See also Linking lift M manuscript, 1, 25, 88, 146, 206, 351 man in pronominal use, 259, 279–86 appearing in the verse-initial drop, 279–86 in Beowulf, 280–1 Marked vs. unmarked, 8, 9, 10, 48, 49, 50, 53, 55, 59, 60, 61, 62, 66, 86, 90, 92, 94, 95, 104, 121, 122, 132, 133, 140, 147, 152, 153, 154, 173, 175, 176, 181, 187, 190, 196, 202, 215, 220, 223, 236, 254, 255, 263, 281, 287, 302, 305, 306, 316, 318, 328, 332, 333, 337, 338, 340, 348, 349 Metrical hierarchy, 11, 341. See also Metrical unit; Prosodic hierarchy Metrical position. See Drop; Lift Metrical strength. See Prominence Metrical type, 8, 27. See also under individual types and subtypes basic, 9–10, 176 in Beowulf, 8–10 derived, 9–10, 176, 179 the system of, 9–10, 175–82 in Genesis, 359 Metrical unit. See Heavy drop; Lift; Line; Normal drop; Verse Mora, 16, 23–4, 32, 62, 132, 189, 212–13, 214, 293, 336, 337. See also Foot; Stress; Syllable bimoraic, 24, 32, 62, 103, 132, 188, 189, 214, 293, 336, 339 Muspilli, 5

in prevocalic position, 25–6 Kaluza’s Law, 211–17, 222, 337–8. See also Resolution; Suspension of resolution in Beowulf, 211, 212, 216, 222 Kuhn’s Laws, 178, 281–2, 284, 286 Lift, 8, 27, 33, 254–5, 256, 260, 293, 339–40. See also Lift formation prototype of, 10, 193, 330–1 Lift formation. See also Alliteration, by nature; Alliteration, by position; Lift and alliteration, 279, 293–4, 339–40 and foot construction, 293, 339 by nature, 256, 261, 274 by position, 256, 257, 262 Line, 4, 7–8, 11 Linearity-based Metrical Strength Scale, 10, 88, 147, 181, 182, 184, 208, 209, 232, 282, 298, 316 Linking lift, 313–14, 316, 318–19. See also Linking word

Normal drop, 8, 27, 33, 49, 180–1, 182, 256, 293, 332. See also Heavy drop prototype of, 10, 33, 49, 57, 58, 59, 70, 102, 103, 132–3, 153, 154, 190, 193, 330–1 in relation to the a-verse/b-verse distinction, 184–7, 309 verse-initial vs. non-verse-initial, 182–7, 286 Old English metre, 7–11, 25 Parameter, 7. See also Disyllabic ending; Foot; Prefix; Prominence; Stress; Syllable length; Syllables, number of for suspension of resolution, 203–11 for variation of the normal drop, 182–7 for variation of type A1, 47–53, 183, 332 for variation of type B1, 104, 333 for variation of type C, 120–2, 333 for variation of type D, 127, 333 for variation of type D*, 137, 144, 333–4

MOSP-Subject Index.qxd

490

5/11/04

2:53 PM

Page 490

Index of subjects

Parameter (continued) for variation of type E, 149, 151, 155, 334 Preference condition. See Parameter Prefix. See also Word as occupant of the first drop of type A1, 36, 41, 45, 47, 51, 332 as occupant of the first drop of type B1, 80–2, 116 as occupant of the first drop of type C, 113–17 as occupant of the first drop of type D*, 142–3 Principle of balance and harmony, 3, 7, 216, 230–1, 243, 272, 273, 316, 343, 344 Principle of four positions per verse, 8, 25, 78, 158, 206, 347 Principle of left dominance, 113, 230, 235, 255, 260, 263, 267, 268, 269, 270, 273, 283, 284, 286, 297, 318, 338 Principle of maximal contrast, 230, 273, 343, 344 Prominence, 8, 23, 27, 49, 53, 65, 69, 70, 75, 82, 86, 91, 103, 113, 115, 116, 117, 120, 121, 122, 133, 136, 139, 143, 147, 148, 153, 154, 171, 182, 184, 187, 188, 189, 193, 194, 202, 203, 204, 205, 207, 208, 209, 210, 215, 216, 221, 222, 223, 226, 228, 229, 230, 232, 233, 237, 238, 243, 254, 256, 257, 260, 261, 262, 269, 286, 293, 294, 302, 316, 330, 336, 337, 338, 339, 340, 342, 344, 351 as the basis of parameters, 53, 122, 334–5, 340–1 (see also Parameter) contour, 130, 147, 148, 209, 230 syntagmatic arrangement of, 91, 92, 115, 116, 124, 130, 208, 209, 222, 228, 351 (see also Suspension of resolution) Promotion, 176. See also Demotion; Expansion Prosodic hierarchy, 12, 33, 208, 337, 341. See also Foot; Metrical hierarchy; Mora; Syllable Prototype, 7, 57, 60, 63, 103, 104, 111, 121, 132, 147, 152, 161, 162, 164, 166, 174, 254, 332 as a conceptual basis of metrical organisation, 7, 99, 260, 332 of the heavy drop, 10, 75, 112, 147, 148, 190, 193, 330–1 of the hypermetric verse, 301, 305, 315–16, 318, 319, 323, 325, 326, 327 of the lift, 10, 193, 330–1 of the normal drop, 10, 33, 49, 57, 58, 59, 70, 102, 103, 132–3, 153, 154, 190, 193, 330–1

Quasi compound, 69–70, 91, 92, 190–3, 305. See also under individual metrical types and subtypes; Compound quað he phrase extrametrical status of, 27–8 Reorganisation, 341–4 of alliteration, 254–5, 271–4, 275, 293–4, 339–40 of anacrusis, 160, 172–5, 178, 329, 335, 336, 344 of hypermetric verses, 323–9, 336 of resolution, 228–9, 230–5, 254–5, 337–8, 342 of suspension of resolution, 211, 216, 337–8, 342 of type A1 and type A2, 53–61, 70–3, 102–3, 152, 172, 181, 196, 214, 272, 331–2 of type B1, 101–4, 174 of type C, 122 of type D, 136, 172–3, 181, 201, 331, 332 of type D*, 160, 178, 179, 335 of type E, 151–5, 239 of type E*, 159–60, 179 Resolution, 10, 32, 91, 116, 132, 194–235, 336–8, 344, 351. See also Kaluza’s Law; Suspension of resolution and alliteration, 207, 243–55 in Beowulf, 198, 199, 201, 202, 203, 211, 216, 223, 228–9, 230, 232, 234, 248, 251–2, 253–4 and distinction between the first and second lifts, 223–35, 248, 251 on the first drop of type E, 146, 147, 201–3, 204, 208, 209, 210, 212, 337 on the first lift of type A1, 223, 224 on the first lift of type C, 116, 198–9, 209, 222, 223, 224, 226–30, 231 on the first lift of type D, 130, 199, 223, 224, 253–4, 351 on the first lift of type E, 223, 224, 232, 342 in Genesis, 362–3 and metrical types, 223–35 and the prosodic hierarchy, 132 reorganisation of, 228–9, 230–5, 254–5, 337–8, 342 on the second lift of subtype D1, 199, 204, 210, 212, 223, 224, 337, 342, 346, 351 on the second lift of subtype D2a, 199, 223, 224, 230–1, 342 on the second lift of subtype D*1, 200, 224, 230, 231, 232, 342 on the second lift of type A1, 223, 224, 253, 337

MOSP-Subject Index.qxd

5/11/04

2:53 PM

Page 491

Index of subjects on the second lift of type A2, 223, 224, 232–4, 342 on the second lift of type B1, 223, 224, 234–5, 343 on the second lift of type C, 198–9, 208, 209, 216, 224 on the second lift of type D*, 200, 223, 224 on the second lift of type E, 223, 224 Restoration of syncopated vowels, 11, 12–13, 23, 102, 330 Retraction, 16 Sentence particle, 166, 174, 178, 281–3, 285, 286, 317. See also Kuhn’s Laws Smoothing, 16–22 Stress, 10, 11, 12, 20, 23, 24, 26, 48, 53, 62, 71, 104, 121, 131, 132, 133, 134, 136, 180, 181, 182, 194, 197, 203–4, 231, 232, 234, 256, 262, 273, 281, 286, 330, 331, 333, 335, 336, 339, 340, 342, 343, 344, 346, 347, 349, 353, 354, 357. See also Foot; Parameter; Weakening of stress lexical vs. non-lexical, 24, 29, 35, 36, 48, 60, 61, 66, 67, 70, 71, 72, 75, 79, 91, 101, 102, 104, 111, 112, 120, 123, 125, 126, 129, 130, 131, 132, 133, 135, 136, 143, 146, 147, 148, 149, 151, 153, 154, 157, 174, 180, 183–4, 187, 188, 190, 193, 202, 205, 206, 231, 269, 272, 295, 333, 334, 342, 348, 354, 355, 359 as a metrical organiser, as opposed to syllable length, 53, 61, 71, 180, 183–4, 231, 340–2 primary vs. secondary vs. weak, 9, 11, 13, 21, 22, 24, 48, 65, 85–6, 89, 108, 116, 117, 120, 121, 122, 123, 125, 126, 128, 130, 131, 132, 133, 135, 137, 139, 146, 151, 156, 162, 172, 188, 193, 197, 199, 202, 203, 204, 205, 208, 209, 216, 217, 220, 221, 222, 230, 231, 269, 272, 273, 286, 297, 314, 316, 317, 331, 333, 334, 335, 338, 339, 342, 346, 347, 352, 355, 363, 364 Subtype A1s, 61–6, 274–5, 348 alternative scansion as subtype D2a, 64, 128–9, 274–5 alternative scansion as type B1, 64, 90–3 alternative scansion as type E, 63–4, 65 with anacrusis, 92 in Beowulf, 62–3, 64, 195, 196, 272, 274, 275 compounds in, 61, 194–5 in Genesis, 354

491

quasi compounds in, 190–1 Subtype D1, 125–6, 134. See also Type D Subtype D2a, 126–9, 130, 131, 133, 134. See also Type D alternative scansion as subtype A1s, 64, 128–9, 274–5 Subtype D2b, 129–30, 131, 133–4. See also Type D alternative scansion as hypermetric, 306–7 alternative scansion as type B1, 83, 106–7, 257 alternative scansion as type E, 129, 149, 275–7 Subtype D*1, 137–9. See also Type D* Subtype D*2a, 137–9. See also Type D* alternative scansion as type A2, 136, 217–23 alternative scansion as type C, 259, 260 Subtype D*2b, 137–9. See also Type D* alternative scansion as hypermetric, 304–6 alternative scansion as type B1, 108–9, 136, 258–9, 260 alternative scansion as type E*, 137, 156, 277–8 Suspension of resolution, 66, 194–211, 337–8. See also Kaluza’s Law; Resolution in Beowulf, 198, 201, 202, 210–11, 216, 222, 230 on the first drop of type E, 201–3, 204, 208, 210, 337, 347 on the first lift of type A1, 206 on the first lift of type B1, 206–7 on the first lift of type E, 207 on the lift of type A3, 206 mechanism of, 208 parameters for, 203–11 reorganisation of, 211, 216, 337–8, 342 Rule for, 205, 222 on the second lift of subtype A1s, 194–7, 210, 212, 337 on the second lift of subtype D1, 199–200, 204, 210, 212, 337 on the second lift of type C, 121–2, 197–9, 204, 208, 212 Svarabhakti vowels in Old English, 13–22 in Old Saxon, 11, 13–22, 23, 214–15, 330, 337–8 Syllabification in Old Saxon, 24–5 and West Germanic Gemination, 22–3 Syllable, 12, 17, 20–2, 22–3, 24, 26, 33, 47, 49, 62, 121, 132, 194, 208, 234, 237, 287, 288, 292, 293, 294, 327, 334, 336–7, 339, 340. See also Parameter Syllable Contact Law, 22–3

MOSP-Subject Index.qxd

492

5/11/04

2:53 PM

Page 492

Index of subjects

Syllable length, 23–4, 47, 48, 49, 50, 53, 59, 132, 187, 212, 213, 214, 215, 216, 222, 332, 337–8, 348. See also Parameter; Syllable as a metrical organiser, as opposed to stress, 58, 59, 61, 71, 183, 340–1, 342 of the first drop of subtype D1, 125, 126, 131–2, 134–5, 333 of the first drop of subtype D2a, 128, 131–3, 134–5, 333 of the first drop of type A1, 41, 47, 48, 53, 58, 59, 60, 71, 332 of the first drop of type E, 145, 153, 334 of the second drop of type D*, 137, 220, 333–4 of the second lift of subtype D1, 125, 126, 134–5, 333 of the second lift of type C, 117, 121–2, 132, 228–30, 333 Syllables, number of. See also Parameter in the first drop of type A1, 45, 47, 51, 53, 55, 58, 59–60, 68, 70–1, 183, 332, 353 in the first drop of type A1A1, 297–8, 301, 327 in the first drop of type A2, 67–9, 217–20 in the first drop of type B1, 77–8, 80–7, 93, 99, 183 in the first drop of type C, 77–8, 113, 116–17, 183 in the first drop of type D*, 141, 144, 217–20, 306, 334, 357 in the first drop of type E*, 159 in the initial part of type A3, 77–8, in the initial part of type A3A1, 308–9, 319, 327–8, 329 in the initial part of type B3A1, 311 in the initial part of type CA1, 312–13 in the normal drop, 180–1, 182–7 in the second drop of type A1A1, 298, 327 in the second drop of type B1, 87–9, 155, 183, 300–1 in the second drop of type D*, 305 in the second drop of type E, 155, 303, 334, 348 in the third drop of type E*, 156 Syntactic constituency, 128, 129, 131, 132, 134, 137, 151, 156, 217, 221, 264–71 as a metrical organiser, 273–8 te ⫹ gerund used as an independent verse, 115–16 Type A1, 133, 175, 176, 180, 187, 315–16. See also Subtype A1s alternative scansion as hypermetric, 307–13, 319–20

alternative scansion as type A3, 257–8, 260 with anacrusis, 93, 161–3, 282–3, 299, 308, 311, 312–13, 329 in Beowulf, 9, 31–2, 36, 41, 46, 53, 55 compounds in, 29, 30, 46, 72, 195 the first drop of, 29–66, 148, 151, 181, 182–4, 184, 272, 286, 297–8, 315, 331 in Genesis, 353–4 gradation of, 46–61, 71, 102–3, 154, 332 reorganisation of, 53–61, 70–3, 102–3, 152, 172, 181, 196, 214, 272, 331–2 the second drop of, 66–75 Type A2, 66–73, 133, 175, 176, 179 alternative scansion as subtype D*2a, 136, 217–23 with anacrusis, 161 compounds in, 66, 70, 72, 73, 233 derivation of, 176 in Genesis, 354 quasi compounds in, 69–70, 73, 191, 233 reorganisation of, 53–61, 70–3, 102–3, 152, 172, 181, 196, 214, 272, 331–2 Type A2a in Beowulf, 9, 30, 56, 70–2, 73, 195, 196 in Genesis, 353 Type A2b in Beowulf, 9, 70, 179 in Genesis, 359 Type A3, 76–80, 175, 176, 179, 309 alternative scansion as type A1, 257–8, 260 in Beowulf, 9, 78, 79, 179 compounds in, 79 derivation of, 112, 176 the final drop of, 78–80, 331 in Genesis, 354 quasi compounds in, 79 Type B1, 80–111, 122–3, 175, 176, 180, 187 alternative scansion as subtype A1s, 64, 90–3 alternative scansion as subtype D2b, 83, 106–7, 257 alternative scansion as subtype D*2b, 108–9, 136, 258–9, 260 alternative scansion as type C, 106, 123–4 alternative scansion as type E, 88, 94–105, 173–4 alternative scansion as type E*, 108–9, 157 in Beowulf, 9, 85–6, 90, 93–4, 95, 97, 101–2, 109–10, 355 compounds in, 89–94, 95–7, 101 the first drop of, 80–7, 182–4, 184, 300 in Genesis, 354–5 gradation of, 80–7, 104–5, 110–11, 333 quasi compounds in, 91, 92, 191 reorganisation of, 101–4, 174

MOSP-Subject Index.qxd

5/11/04

2:53 PM

Page 493

Index of subjects the second drop of, 87–111, 155, 181–2, 183, 184, 301, 331 Type B3, 111–13, 175, 179, 311 in Beowulf, 9, 112, 179 derivation of, 112, 113, 176 in Genesis, 355 Type C, 113–24, 175, 176, 180 alternative scansion as subtype D*2a, 259, 260 alternative scansion as type B1, 106, 123–4 in Beowulf, 9, 122, 123, 229, 355 compounds in, 117, 123, 229–30 the first drop of, 113–17, 182–4, 184, 312–13 in Genesis, 355–6 gradation of, 113–14, 117–23, 333, 342 quasi compounds in, 191–2 reorganisation of, 122 the second drop of, 123–4, 182, 331 the second lift of, 117–23, 132 Type D, 124, 125–36, 175, 176, 180. See also Subtype D1; Subtype D2a; Subtype D2b with anacrusis, 163–5, 167 in Beowulf, 9, 129–30, 134–6 compounds in, 125, 126, 129, 130, 135, 199, 200 in Genesis, 356 gradation of, 125–6, 128, 131–6, 333 proportion of, to type D*, 140, 172 quasi compounds in, 130, 192 reorganisation of, 136, 172–3, 181, 201, 331, 332 Type D*, 136–45, 175, 176, 179–80. See also Subtype D*1; Subtype D*2a; Subtype D*2b alternative scansion as type A1⫹C, 299–300 with anacrusis, 163–4 in Beowulf (corresponding to the three subtypes of types D1, D2a, and D2b), 9, 140–2, 143, 144, 172, 179, 238, 357 compounds in, 137, 143 derivation of, 176 the first drop of, 141–4, 182, 184, 217–20, 286, 331, 344, 351 in Genesis, 356–8 gradation of, 137–9, 144, 333–4 proportion of, to type D, 140, 172 quasi compounds in, 143 reorganisation of, 160, 178, 179, 335 Type E, 145–55, 175, 176, 180 alternative scansion as subtype A1s, 63–4, 65 alternative scansion as subtype D2b, 129, 149, 275–7

493

alternative scansion as type B1, 88, 94–105, 173–4 with anacrusis, 165, 173–4 in Beowulf, 9, 145, 149, 151–2, 155, 238–9, 358 compounds in, 148–9 the first drop of, 145–9, 187–90, 286, 303, 347, 348, 350 in Genesis, 358–9 gradation of, 149–55, 334 quasi compounds in, 192 reorganisation of, 151–5, 239 the second drop of, 155, 184, 187–90 Type E*, 156–60, 175, 176, 179–80 alternative scansion as subtype D*2b, 137, 156, 277–8 alternative scansion as type B1, 108–9, 157 with anacrusis, 165 analogous configuration of, in Beowulf, 158, 159, 179 compounds in, 156 derivation of, 176 the first drop of, 156, 159, 184 in Genesis, 359 reorganisation of, 159–60, 179 the second drop of, 156 the third drop of, 156–7 Verse, 8. See also Heavy verse; Hypermetric verse normal, 8, 295, 344 opposition between the a-verse and the b-verse, 179, 180, 184, 186–7, 344 Vowel epenthesis. See Svarabhakti vowels Vowel system, Old Saxon, 288–9 Weakening of stress, 11–23, 57, 136, 175, 181, 288, 293, 294, 330, 335, 339, 340, 344 Wessobrunn Prayer, 5 West Germanic Gemination, 11, 22–3, 330 Word. See also Heavy verse; Prefix as occupant of the final drop of type A3, 79 as occupant of the first drop of type A1, 36, 41, 45, 47, 51, 269 as occupant of the first drop of type D, 128 as occupant of the first drop of type D*, 269 as occupant of the first drop of type E, 149, 269, 331 as occupant of the first drop of type E*, 156, 269 as occupant of the second drop of type A1, 68–9 as occupant of the second drop of type A2, 66

MOSP-Subject Index.qxd

494

5/11/04

2:53 PM

Page 494

Index of subjects

Word (continued) as occupant of the second drop of type B1, 105–10 as occupant of the second drop of type C, 123–4 as occupant of the second drop of type D, 129 as occupant of the second drop of type D*, 137, 139 as occupant of the third drop of type D*, 137, 139 Word classes. See also Alliteration, by nature; Alliteration, by position; Lift

formation, by nature; Lift formation, by position class 1 (substantives), 92, 106, 107, 110, 157, 255, 256, 260, 261, 262, 293, 314, 317–18, 331, 338, 339, 364 used in the verse-initial drop, 279–86 class 2 (finite verbs and adverbs), 92, 107, 110, 136, 255, 256, 257, 259–60, 260, 261, 262, 302, 318 class 3 (function words), 107, 110, 255, 256, 259, 260, 261, 262, 283, 331

MOSP-Verse Index.qxd

5/11/04

2:54 PM

Page 495

Index of verses cited for discussion or exemplification Andreas 560a: 349 Beowulf 5a: 10 5b: 10 6b: 9 7a: 10 7b: 10 8a: 56 8b: 9 11a: 9, 161 11b: 9 13a: 56 14a: 56 22a: 9 31a: 9 65a: 9 66a: 55 80a: 109 90a: 268 104b: 158 105b: 55 136a: 55 147a: 149, 268 202a: 90 252a: 161 272b: 110 292b: 25 316a: 263 326a: 9 345b: 9 376a: 272 380a: 281 387a: 130 409a: 161 438a: 31 451b: 25 455b: 110 469b: 110 473a: 141 501b: 100 517a: 268 525a: 109 545a: 149, 268

608a: 55, 56 658a: 149 667b: 158 744a: 199 758a: 262 908b: 272 911a: 203 938a: 9 995b: 158 998b: 158 1009b: 63 1033a: 9 1108a: 11 1111a: 55 1128a: 31 1136a: 158 1166a: 302, 325 1168a: 325, 328 1168b: 325 1175a: 280 1182b: 110 1187a: 158 1198a: 31 1307a: 9 1395a: 149 1422a: 266 1424b: 159 1459b: 158 1485a: 268 1532a: 9 1537a: 262 1616b: 110 1649a: 56 1667a: 161 1681a: 158 1682a: 9 1696a: 109 1719a: 9 1728a: 9 1729a: 281 1763a: 109 1782b: 158 1814a: 55 1819b: 25 1828b: 63, 196

MOSP-Verse Index.qxd

5/11/04

496 1859a: 56 1864a: 109 1876b: 110 1918a: 158 1941b: 96 2193a: 31 2252a: 161 2281a: 281 2297a: 324 2309b: 123 2313a: 268 2355b: 280 2387b: 158 2409a: 161 2441a: 95 2455a: 161 2457b: 63 2555a: 281 2574b: 110 2757b: 158 2774a: 130 2805a: 25 2870a: 109 2894a: 158 2987a: 30, 268 2996a: 281 3037b: 158 3056a: 280, 283 3065a: 281 3175a: 280 3181a: 281 Christ 637a: 349 Elene 216a: 349 Genesis 9a: 363 18a: 363 23b: 355 25b: 360 30b: 354 31b: 356, 361 35b: 354, 363 49a: 363 58b: 355 66b: 360 79a: 360 80a: 357 82b: 354 93a: 353 102a: 363 105b: 360 115a: 363

2:54 PM

Page 496

Index of verses cited 125a: 363 135a: 357 139a: 363 154b: 355 165b: 361 173a: 357 178a: 363 221a: 363 233b: 360 244b: 359 246a: 363 256b: 355 274a: 363 287b: 363 294a: 357 300b: 363 315b: 359 Heliand 1a: 26 1b: 26 2a: 235 6a: 156, 267 7a: 266 9a: 264 15a: 264 16a: 267 18b: 124 19a: 263 20b: 82 24a: 76 24b: 39 25a: 30, 54 27a: 40 31b: 206 36b: 156, 159 38a: 259 39a: 95, 145 40b: 161 45a: 35, 54 46b: 296 51a: 82 52a: 267 52b: 173 55a: 259 60b: 347 61b: 350 62a: 196, 203 65b: 46 68a: 187 76b: 82 82a: 147 82b: 125 94a: 197 95a: 267, 282 100b: 198

MOSP-Verse Index.qxd

5/11/04

2:54 PM

Page 497

Index of verses cited 103b: 276 106a: 33, 54, 331 112b: 277 117b: 105 121a: 143 144b: 73 155b: 137 167a: 149 174b: 275 189a: 113 195a: 64, 90 196a: 41 196b: 275 197a: 108 205a: 35 207a: 79 209a: 36 216a: 161 221a: 89, 331 222b: 125 232a: 105 235a: 112, 236, 263 236b: 347 254b: 156, 278 261a: 264 266a: 304 268a: 157, 267 279a: 95, 145, 303 282a: 259 289a: 163 296a: 161 302a: 179 303a: 289 303b: 289 304a: 178 313a: 117 314b: 235 316b: 33 319a: 144 326a: 206, 347 326b: 275 340a: 258, 259 342a: 144 345a: 279 347a: 304 358a: 41, 54 359b: 82 369b: 275 372a: 264 376b: 95, 97 395a: 276 408a: 280 409a: 70 410b: 191 412a: 128, 264, 271 414a: 306

418a: 27 421b: 161 424a: 143 440a: 264 450a: 64, 90 464b: 199 467b: 82 468a: 159 473a: 217 484b: 82 490b: 296 492b: 278 493a: 267, 282 499b: 28 502b: 109 503a: 267, 282 507a: 259 510a: 276 512b: 109 518a: 266 519b: 195 533a: 137 537a: 201, 203 552b: 82 557a: 264, 296 558a: 264 564a: 259, 260 572a: 39 573a: 111, 236 575a: 145 591a: 110 593a: 267, 282 594a: 259 594b: 25 595a: 267 598a: 264 599a: 129 600a: 264 600b: 307 601b: 319 604a: 301 605a: 297 610a: 145, 146, 188, 189, 201 611a: 50, 297 619a: 146 626a: 107, 129, 264 639b: 276 641a: 161, 166 659a: 200 661a: 264, 276 662b: 109 671a: 264 674a: 66 685a: 347 696a: 137 696b: 351

497

MOSP-Verse Index.qxd

5/11/04

2:54 PM

498 700a: 266, 276 703a: 195, 203 712a: 41 717a: 258 730a: 117 732a: 165 735a: 157, 159 743a: 304 752a: 80 754a: 267 760a: 200 763a: 264 769b: 276 776a: 264 785a: 46, 163 792a: 61 794a: 266, 275 798a: 266, 267 801a: 258 838a: 266, 275 847a: 257 848a: 69 855a: 264 865a: 267 874b: 109 877a: 80, 350 880a: 80, 350 884a: 80, 350 890b: 109 894a: 261 895b: 96 897b: 323 899a: 296 899b: 311 901a: 303 903a: 304 917a: 187 926b: 276 928b: 124, 198 929a: 95, 145, 146, 189, 201 936a: 144 945a: 266 946a: 306 952a: 262 954a: 300 957a: 129 960b: 275 961a: 264, 304 965a: 347 972a: 50 973b: 276 976a: 115 978a: 156, 278 982b: 275 984a: 264 988a: 50

Page 498

Index of verses cited 989a: 302 990a: 264 991a: 264 992a: 264 993a: 264 1005a: 143, 264, 304 1007a: 137 1010a: 129, 266, 275 1011b: 64 1014a: 35 1023a: 115 1024a: 191 1036a: 347 1041a: 126, 199 1056a: 30 1057b: 123 1059a: 264 1061a: 65 1061b: 109 1062b: 123 1077b: 195 1080a: 257 1085a: 262, 277 1090a: 261 1096a: 300 1098a: 201 1099a: 257 1105a: 259 1107a: 264, 304 1108a: 348 1110a: 304 1114a: 258 1114b: 348 1117a: 266, 276 1133a: 97, 149 1134b: 277 1135b: 107, 275 1139a: 80 1143b: 206 1144a: 304, 307 1159a: 267 1163a: 200, 264, 274 1169b: 161 1180a: 137, 217, 267 1215a: 257 1216b: 35 1221a: 80, 117 1223a: 156, 277 1229a: 80, 117 1239a: 261 1240a: 259 1240b: 41 1242b: 94, 95 1255a: 156, 159 1265b: 347 1271a: 264

MOSP-Verse Index.qxd

5/11/04

2:54 PM

Page 499

Index of verses cited 1282a: 264 1287b: 275 1298a: 80 1301a: 297 1315a: 237 1344a: 267 1348a: 26 1295a: 130 1300b: 312 1308b: 317 1311a: 296 1312b: 317 1314b: 317 1315a: 295 1318b: 302 1320b: 317 1323a: 149, 331 1324a: 258, 259, 261 1335a: 266, 275 1348a: 264, 305 1355a: 39 1373b: 106, 123 1374b: 197 1384a: 264 1394a: 113 1412a: 264 1417a: 257 1418b: 25 1429a: 305 1434a: 80 1440a: 264, 267 1444a: 259 1446a: 125 1448b: 149 1451a: 304 1460a: 163 1467b: 159 1469a: 257 1471a: 267 1473a: 267 1479a: 279, 283 1481a: 257 1484a: 283 1494a: 267, 282 1494b: 87 1495a: 89 1516a: 258, 259 1521a: 279 1522a: 110 1525a: 258 1526b: 145, 303 1533a: 111, 206, 236 1535b: 279 1536a: 258 1541b: 311 1552b: 109

1554a: 307 1554b: 312 1555a: 277 1557a: 66, 233, 264 1561a: 300 1571b: 94, 95 1581a: 264 1592a: 143 1596a: 107, 275 1600a: 206 1601b: 261 1602a: 109 1603a: 206 1607a: 109 1610a: 258 1623a: 95 1634a: 267 1637a: 297 1638a: 69 1643b: 65 1668a: 348 1676a: 156, 159 1681a: 302 1683a: 302 1684b: 336 1685b: 297, 318 1687a: 299 1695a: 29 1698a: 191 1707a: 267 1710a: 282 1718b: 319 1730a: 304 1732b: 198 1742a: 63, 95, 201 1746b: 110 1752a: 258 1763b: 107 1773a: 129 1774a: 149, 266, 267, 276 1791a: 41 1796a: 302 1802b: 195 1810a: 264 1826a: 305 1833a: 297 1865a: 264 1867b: 116 1871a: 117 1879b: 279 1889b: 89 1890a: 88 1893a: 198 1921a: 264 1931a: 267 1935b: 261

499

MOSP-Verse Index.qxd

5/11/04

500 1939a: 33, 54 1950b: 25 1973a: 305 1973b: 312 2000b: 275 2001b: 206 2007b: 137 2012a: 258 2014b: 206 2019b: 261 2024a: 267 2031a: 146 2036b: 277 2037a: 32, 54 2038a: 267 2049a: 258 2051a: 136, 258 2062a: 106, 123, 331 2084a: 74, 187 2085a: 159 2087a: 258 2100a: 275 2101a: 258 2104b: 27 2109a: 68 2115a: 200 2122a: 264 2133a: 267, 270, 271 2138a: 267, 269, 339 2143a: 197 2145a: 129, 264 2158a: 46 2161b: 277 2170a: 267, 269, 339 2171a: 266, 267 2176a: 264 2181a: 143, 331 2183b: 109 2198a: 261 2202b: 106 2206a: 159, 267 2215a: 297, 328 2218a: 64, 267, 274 2223a: 144, 278 2231a: 301 2234a: 268, 269, 339 2238b: 128 2243a: 142, 258 2244b: 199 2250b: 275 2251a: 264 2270b: 261 2272a: 261 2274a: 137 2274b: 33 2279a: 258

2:54 PM

Page 500

Index of verses cited 2282b: 82 2284a: 258 2285a: 143, 268, 269, 272, 339 2290a: 302 2291b: 275 2292a: 261 2298a: 264 2318a: 166 2321a: 351 2325a: 267 2333a: 159, 165 2337a: 126 2343a: 72 2345b: 88, 109 2348a: 137, 264 2349a: 40, 54 2354a: 233 2355a: 264 2366a: 125 2367a: 111, 236 2371a: 264, 270 2382a: 61, 195 2400a: 137 2407b: 262 2428a: 25 2435a: 33 2438a: 279, 284 2443b: 199, 203 2444a: 258, 283 2458a: 95, 97, 331 2467b: 277, 300 2471a: 80, 187 2482a: 196 2482b: 303 2492a: 30, 32, 54, 331 2502a: 261 2515a: 143 2516b: 235 2520a: 267 2529a: 156, 159 2543a: 259 2547a: 110 2553a: 258 2556a: 265 2561a: 258 2576a: 265 2581a: 265 2585b: 137 2586a: 348 2594a: 264 2596a: 303 2597a: 303 2601a: 265 2615b: 157 2619b: 207 2630a: 149

MOSP-Verse Index.qxd

5/11/04

2:54 PM

Page 501

Index of verses cited 2632b: 275 2635b: 146, 187 2639a: 261 2644a: 221 2645a: 258 2648b: 64, 93, 274 2652a: 159 2664a: 351 2689b: 126, 148, 188, 189, 200 2699b: 129 2700b: 146, 201 2701a: 198 2722a: 126, 200 2727a: 198 2728b: 351 2735a: 303 2737b: 90 2755a: 107 2756a: 111, 236 2762b: 275 2768a: 259 2780b: 206 2782a: 279 2783b: 94, 95 2804a: 145, 146, 188, 189, 201, 350 2805a: 156, 268, 282 2813a: 257 2820b: 323 2821a: 265, 270 2825b: 317 2832a: 265 2836a: 64, 274 2847a: 265 2850a: 124 2850b: 95 2853a: 142 2854b: 124 2856a: 304 2858a: 296 2858b: 296 2870a: 264, 266, 268, 269, 274, 339 2879a: 258 2882a: 157, 268 2882b: 276 2892a: 233 2900a: 163 2908a: 265 2908b: 64, 274 2914a: 202 2917a: 265 2934a: 261 2935a: 35 2937a: 265 2948a: 300 2975a: 267 2975b: 143

2991a: 265 2993a: 146 2994a: 259 2997a: 261 3005a: 265 3006b: 159 3014b: 137 3020b: 235 3021a: 26 3027b: 261 3029b: 199 3031a: 258 3036a: 159 3036b: 319 3037a: 297 3044a: 200 3046a: 143, 265, 278, 305 3054a: 206, 348 3059a: 199, 265 3061a: 187 3062a: 237, 295 3063a: 300 3065a: 265 3066b: 303 3067b: 296, 297, 318 3068a: 300 3069a: 348 3069b: 279 3070a: 265 3072b: 312 3073a: 261 3074b: 276 3075a: 26 3079a: 145 3081a: 163 3082a: 265 3085a: 266, 275 3093b: 348 3097a: 87, 300 3099a: 265 3107a: 276 3108a: 275 3108b: 348 3117a: 257 3119b: 113 3122a: 195 3127a: 303 3132a: 268 3135a: 123 3140a: 258, 279 3144a: 258 3144b: 276 3147a: 267, 278 3149a: 40 3172a: 265 3174b: 149

501

MOSP-Verse Index.qxd

5/11/04

502 3186a: 125 3189a: 161 3196a: 348 3197b: 106 3201a: 348 3210a: 348 3220b: 109 3228b: 277, 305 3230b: 109 3236b: 235 3241a: 265 3248b: 36 3251a: 237 3257b: 347 3261a: 65 3264a: 258 3269a: 105 3270b: 217 3277a: 200 3278b: 261 3290b: 50 3292a: 35 3295b: 277 3298a: 265, 303 3299a: 282 3306a: 348 3321a: 258 3323a: 268 3344a: 143, 304 3346a: 261 3350b: 74 3362b: 276 3363a: 259 3364b: 64, 274 3386b: 109 3405a: 280 3408a: 257 3414a: 143, 331 3425a: 258, 283 3429a: 258, 283 3432a: 259, 283 3435a: 258 3438a: 33 3441a: 143 3442b: 27 3450a: 265 3455a: 206 3471b: 64, 274 3482a: 259 3483a: 268 3492a: 117 3497a: 297 3499b: 307 3502a: 317 3518a: 280 3524a: 278

2:54 PM

Page 502

Index of verses cited 3551b: 196 3557a: 347 3562a: 303 3563a: 265 3581b: 157 3587a: 173 3595a: 347 3611b: 319 3612b: 276 3633a: 141 3639a: 217 3644a: 265 3653a: 265 3646a: 304 3646b: 109 3665a: 157, 165, 268 3669a: 265 3673b: 125 3682b: 346 3691b: 235 3692a: 46 3693a: 111, 236 3694b: 149 3697b: 201 3710a: 268 3716a: 347 3721b: 29 3745b: 32 3747a: 72 3759a: 264 3766a: 79 3772a: 65 3780a: 76 3783a: 268, 271, 350 3794b: 89 3802a: 305 3805a: 79, 339 3822b: 277 3838a: 29, 54 3842a: 72 3882a: 108 3883a: 259 3895a: 258 3895b: 217 3897a: 191 3902b: 149, 276 3928a: 351 3929a: 258 3960a: 261 3962a: 112, 236 3962b: 235 3974a: 261 3981a: 267, 276 3994a: 143, 265, 305 3998b: 277 4002a: 268

MOSP-Verse Index.qxd

5/11/04

2:54 PM

Page 503

Index of verses cited 4009b: 109 4011a: 264 4032a: 35 4035b: 276 4038a: 111, 236 4062a: 64, 128, 267, 274 4074a: 76 4084b: 278 4089a: 265 4102b: 129, 275 4103b: 277 4107b: 283 4115a: 265 4122a: 165 4128b: 26 4131b: 278 4136a: 25 4142a: 348 4157a: 198 4158a: 258 4165a: 137, 259 4170a: 350, 351 4177b: 46 4195b: 26 4199a: 276 4211a: 303 4215a: 143, 331 4217a: 39 4220a: 258 4226a: 79 4231a: 268 4241b: 276 4244b: 35 4245a: 278 4250a: 117 4256a: 265 4262a: 280 4264b: 235 4275a: 265 4288b: 40 4291b: 303 4316b: 125 4319a: 259 4325a: 259 4329b: 130, 276 4330a: 66 4342a: 106, 107 4349a: 280 4353a: 30 4354a: 163 4368b: 276 4371a: 74 4374a: 143 4379a: 265 4382a: 259 4411a: 303

4415a: 302 4418b: 311 4421b: 126 4426a: 258 4458a: 106, 123 4478a: 347 4483a: 301 4492a: 69 4495a: 265 4508a: 348 4547b: 278 4562b: 350 4568a: 297, 302 4593b: 148, 188, 189 4598a: 348 4600a: 262 4615a: 348 4618a: 124 4622b: 276 4624a: 348 4629a: 348 4633a: 265 4635b: 124 4658a: 146 4663a: 143 4703a: 200, 203 4704a: 235 4705a: 262 4708b: 145 4720b: 347 4741a: 124 4751a: 305 4753b: 217 4765b: 68 4778a: 107, 275 4789b: 276 4799a: 267, 278 4810b: 348 4813b: 282, 284, 285 4820a: 261 4824b: 276 4826a: 268, 350 4836a: 259 4861a: 258 4866b: 348 4867a: 262, 277 4879b: 276 4885b: 32 4890a: 280 4898a: 110 4909a: 72, 265 4915a: 145, 146 4916a: 265 4925b: 350 4928b: 109 4935a: 201

503

MOSP-Verse Index.qxd

5/11/04

504 4936a: 258 4946a: 351 4949a: 173 4957a: 130 4967a: 258 4972a: 305 4972b: 306 4973a: 259 4974a: 297 4992a: 265 5000b: 277 5029a: 268 5033a: 26 5042a: 143, 272 5045b: 88 5087a: 265 5089a: 265 5094a: 265 5104a: 68, 124, 296 5107a: 351 5112b: 296 5113a: 296 5124a: 161 5140a: 265 5145a: 348 5148a: 268 5149a: 73 5156a: 258 5160a: 348 5169a: 264 5170a: 237, 275 5194b: 61 5195a: 141 5197a: 80 5214a: 278, 300 5231a: 265 5237a: 280 5249a: 280 5253a: 348 5259a: 350 5264b: 41 5270a: 347 5289b: 35 5292a: 258 5298a: 300 5327a: 137, 277 5376b: 109 5380a: 130, 257 5395a: 265 5419a: 266 5420a: 304 5429a: 348 5443a: 199 5443b: 30 5452a: 237 5458a: 259

2:54 PM

Page 504

Index of verses cited 5466a: 163 5473a: 300 5490b: 276 5498a: 261 5499a: 259 5509a: 267 5511a: 265 5528a: 258 5542b: 110 5544b: 207 5548a: 258 5551a: 265 5552a: 263 5555b: 319 5561a: 265 5584a: 64, 267, 274 5590a: 302 5605a: 206 5605b: 64, 128, 274 5607a: 27 5622a: 159 5623a: 266 5633b: 106 5654a: 305 5654b: 306 5655a: 265 5660a: 159 5663b: 156, 278 5664b: 278 5671a: 36 5678a: 305 5684a: 258 5687a: 40 5690a: 304 5695b: 106 5712a: 268 5720a: 303 5722a: 265 5723a: 303 5727b: 95, 331 5730b: 206 5738b: 206 5743a: 265 5745a: 111, 236 5754b: 165 5755b: 277, 304 5773a: 268 5775b: 137, 277 5782a: 265 5793b: 96 5802a: 129, 265, 307 5803b: 276 5804a: 268 5806a: 143, 265, 278, 305 5807b: 165 5809a: 258

MOSP-Verse Index.qxd

5/11/04

2:54 PM

Page 505

Index of verses cited 5811a: 303 5830a: 117 5857a: 115 5858a: 146, 187 5869a: 265 5879a: 79 5892a: 303 5892b: 235 5894a: 265 5898b: 348 5899a: 108 5905a: 74 5907a: 130, 266 5907b: 107

5917a: 297 5918a: 300 5919a: 300 5920’b: 323 5929a: 129, 306, 307 5937a: 206 5946a: 143 5958b: 346 5962a: 266 5966a: 265 5970a: 266 5973a: 259 5973b: 50 5974b: 137

505

MOSP-Verse Index.qxd

5/11/04

2:54 PM

Page 506