HARVARD UNIVERSITY THE GRADUATE SCHOOL OP ARTS AND SCIENCES
THESIS ACCEPTANCE CERTIFICATE
The undersigned, appointed b...
87 downloads
4554 Views
18MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
HARVARD UNIVERSITY THE GRADUATE SCHOOL OP ARTS AND SCIENCES
THESIS ACCEPTANCE CERTIFICATE
The undersigned, appointed by the Division Department of C e l t i c Languages and L i t e r a t u r e s Committee
have examined a thesis entitled The Ogham Inscriptions of Scotland: An Edited Corpus
presented by Katherine Stuart Forsyth candidate for the degree of Doctor of Philosophy and hereby certify that it is worthy of acceptance. Signat£™ .T. Typed name .Jqb.n. T,. JCp.ch.. Signature ... Typed name
Patrick.K,. ..F.or.d...
Signature .... Typed name
Date ....April. 9......1.95.6....
The Ogham Inscriptions of Scotland: An Edited Corpus
A thesis presented by
Katherine Stuart Forsyth to The Department of Celtic Languages and Literatures in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
in the subject of
Celtic Languages and Literatures
Harvard University Cambridge, Massachusetts
April 1996
UMI Number:
9631492
UMI Microform 9631492 Copyright 1996, by UMI Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.
UMI
300 North Zeeb Road Ann Arbor, MI 48103
® 1996 by Katherine Stuart Forsyth All rights reserved
ABSTRACT
The Ogham Inscriptions of Scotland: An Edited Corpus
An illustrated catalogue of all inscriptions in the ogham script from Scotland. Thirtyseven are extant, a further three are documented but lost. They range in date from at least the sixth century AD to the tenth. Each catalogue entry contains a description of the inscribed object, its archaeological and historical context, a detailed description of the inscription, comments on the form of the script, notes towards an interpretation of the text, previous readings, and a select bibliography. It is argued that the division into 'Pictish' and 'Irish-looking' on the basis of form of script is unhelpful, and that the script of Scottish oghams is comparable to that of contemporary oghams in Ireland. The possibility of a 'palaeography' of ogham is raised. There are very few Irish oghams from the early 'post-classical' period (later seventh and eighth century), and Scottish oghams provide important evidence for this intermediate stage in the development of the script. Irish and Scottish oghamists were working within a single tradition, innovations, such as additional letters, were passed back and forth. In some instances the orthography used in Scotland diverges from that of contemporary Ireland, e.g. perpetuation of older orthographical conventions (simplex/geminate consonant distinction), innovative use of 'H'. A key difference is that post-classical ogham in Ireland had only marginal status, whereas in post-seventh century Scotland, ogham was a prestige script used on grand
public monuments such as cross-slabs. The wide range of contexts in which ogham was used there, especially informal and domestic contexts, suggests knowledge of the script was not restricted. Scottish oghams present difficulties of interpretation because many are fragmentary and weathered, the value of some supplementary characters is not certain, orthographical conventions are imperfectly understood, the language of the texts (Goidelic/Brittonic) is not always known, word-division is rarely indicated, often the kind of text to be expected is not known. More texts can be interpreted than previously thought, practical problems go a long way to explaining why others resist interpretation. Jackson's theory of a second non-Indo-European Pictish language is not upheld.
349 words
CONTENTS
TABLE OF CONTENTS
i
ACKNOWLEDGEMENTS
x
DEDICATION
xiii
PREFACE
xiv
LIST OF ILLUSTRATIONS
xvi
ABBREVIATIONS
xxvii
authorities and general (xxvii); standard abbreviations for the pre-1974 counties of Scotland (xxx); other abbreviations and conventions (xxxi)
INTRODUCTION
xxxii
THE OGHAM SCRIPT DESCRIBING OGHAM TRANSLITERATING OGHAM
xxxii iii
Transliteration of 'H' (viii); Transliteration of 'V (viii) TRANSCRIBING AN OGHAM INSCRIPTION
ix
THE PALAEOGRAPHY OF OGHAM
xi
THE TERM 'SCHOLASTIC
xviii
THE FORFEDA
xviii
A NOTE ON WORD-DIVISION
xxi
OGHAM IN SCOTLAND THE STUDY OF SCOTTISH OGHAM
Ix
THE CORPUS
lxiv
THE TEXTS
lxix
THE INTRODUCTION OF OGHAM TO SCOTLAND
lxxiii
SCHEDULE OF READINGS
lxxvi
CONTENTS
ii
GUIDE TO ENTRIES CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND ABERNETHY
lxxx I 2
discovery (2); site (2); description of object (4); description of carving (5); description of ogham inscription (7); form of script (8); interpretation of text (9); discussion (9); bibliography (10); previous readings (10)
ACKERGILL
11
discovery (11); site (11); description of object (13); description of carving (13); description of carving (symbols) (14); description of ogham inscription (15); form of script (17); interpretation of text (17); discussion (20); bibliography (22); previous readings (22)
ALTYRE
23
discovery (23); locality (25); description of object (27); description of carving decoration (28); description of carving - ogham inscription (29); form of script (35); interpretation of text (37); discussion (39); bibliography (40); previous readings (40)
AUQUHOLLIE
41
discovery (41); site (42); description of object (43); description of carving (43); description of ogham inscription (44); form of script (47); interpretation of text (48); linguistic discussion (50); general discussion (53); bibliography (54); previous readings (54)
BAC
MHIC
CONNAIN
discovery and current location (55); site (55); description of object (58); description of carving (59); form of script (61); interpretation of text (61); discussion (65); bibliography (67); previous readings (68)
55
iii BIRSAY
69
site (69); Early Christian sculpture from the Brough of Birsay (71); runic inscriptions from the Brough of Birsay (72); select bibliography (site only) (74); ogham inscriptions from the Brough of Birsay (74); discovery (74); BIRSAY 1 (76); description of object (76); carving technique (77); description of ogham inscription (77); form of script (80); interpretation of text (80); bibliography (81); previous readings (81); BIRSAY 2 (81); description of object (81); description of carving (82); rbrm of script (84); interpretation of text (85); bibliography (85); previous readings (85); BIRSAY 3 (85); description of object (86); description of carving (86); form of script (88); interpretation of text (89); bibliography (91); previous readings (91); general discussion of the Birsay oghams (91)
BLACKWATERFOOT
1-2
93
site (93); discovery (94); description of carving - OGHAM INSCRIPTION I (95); interpretation of text (96); description of carving - OGHAM INSCRIPTION 2 (97); interpretation of text (98); form of script (99); discussion (100); bibliography (100); previous readings (101)
BRANDSBUTT
102
discovery (102); site (102); description of object (103); description of carving technique (104); description of carving - decoration (105); description of carving - ogham inscription (109); form of script (112); interpretation of text (112); discussion (115); bibliography (115); previous readings (116)
BRESSAY
117
discovery (117); site (118); description of object (118); description of carving decoration (119); description of ogham inscription (122); form of script (126); interpretation of text (129); discussion (136); bibliography (137); previous readings (138)
B ROD IE discovery (139); locality (140); description of object (141); description of carving - decoration (142); description of ogham inscription (145); form of
139
iv
script (155); interpretation of text (156); discussion (157); bibliography (158); previous readings (158)
BUCKQUOY
160
discovery (160); site (160); description of object (164); carving technique (164); description of ogham inscription (167); fonn of script (174); interpretation of text (175); discussion (177); parallels (181); historical context (183); conclusion (185); bibliography (186); previous readings (186)
BURRIAN
187
discovery (187); site (187); description of object (189); description of carving (190); description of decoration (190); ogham inscription (192); fonn of script (198); interpretation of text (199); discussion (203); bibliography (204); previous readings (205)
CUNNINGSBURGH
1-3
discovery (206); site (207); runic inscriptions (208); bibliography
206 (209);
ogham-inscribed fragments-CUNNINGSBURGH 1 (209); description of object (209); description of carving (210); description of ogham (211); fonn of script (212); interpretation (212); bibliography (212); previous readings (212); oghaminscribed fragments - CUNNINGSBURGH 2 (213); description of object (213); description of carving (213); form of script (216); interpretation of text (217); bibliography (218); previous readings (219); ogham-inscribed fragments CUNNINGSBURGH 3 (219); description of object (219); description of carving (219); fonn of script (222); interpretation (223); bibliography (225); previous readings (225); the Cunningsburgh oghams - discussion (225)
DUNADD discovery (227); site (227); further indications of the use of literacy at dunadd (228); roman alphabet inscription (229); rock-carvings (230); description of ogham carving (230); form of script (236); interpretation (236); discussion (239); bibliography (242)
227
D UPPLIN
243
discovery (243); locality (244); description of object (246); description of carving - decoration (247); iconography (249); the ogham inscription (249); bibliography (252); the roman alphabet inscription (252); reading and interpretation of text (253); historical context - Custantin filius Fircus (257)
FORMASTON
261
discovery (261); site (261); description of object (262); description of carving decoration (263); description of ogham inscription (267); form of script (272); interpretation of text (278); discussion (285); bibliography (286); previous readings (287)
G IG HA
1 (KILCHATTAN)
288
current location (288); history (288); site (289); description of stone (290); description of carving (291); form of script (295); interpretation of text (296); discussion (297); bibliography (298); previous readings (298)
GOLSPIE
299
discovery (299); locality (300); description of object (301); description of carving - decoration (302); description of carving - ogham inscription (307); form of script (312); interpretation of text (313); discussion (318); bibliography (319); previous readings (320)
GURNESS
321
discovery (321); site (321); description of object (323); carving technique (324); description of ogham inscription (324); form script (328); interpretation of text (329); discussion (331); bibliography (332); previous readings (332)
INCHY RA discovery (333); locality (335); description of object (336); description of carving - symbols (337); relationship of symbols and order of phases (340); relationship of symbols (to ogham inscriptions) (342); description of ogham inscriptions (344); form of script (352); interpretation of text (354); discussion (358); bibliography (359); concordance (359); previous readings (359)
333
vi LATHERON
360
discovery (360); site (360); description of object (361); description of carving decoration (362); description of carving - ogham inscription (364); form of script (367); interpretation of text (368); discussion (372); bibliography (373); previous readings (373)
LOCHGOILHEAD
374
current location (374); site (374); description of stone (374); description of carving (376); roman alphabet inscription (376); description of ogham inscription (377); form of script (379); interpretation of text (379); discussion (382); bibliography (384); previous readings (384)
LOGIE
ELPHINSTONE
385
discovery (385); site (385); historical context (386); description of object (390); carving technique (390); description of carving - symbols (392); description of ogham inscription (393); form of script (396); interpretation of text (398); discussion (399); bibliography (400); previous readings (401)
LUNNASTING
402
discovery (402); site (402); description of object (402); description of carving decoration (403); description of ogham inscription (403); form of script (408); interpretation of text (413); discussion (417); bibliography (419); previous readings (419)
NEWTON
420
discovery (420); locality (422); description of object (424); carving technique (424); description of ogham inscription (426); form of script (431); interpretation of text (432); the non-ogham inscription (437); the Newton symbol stone (439); discussion (440); bibliography (441); previous readings of ogham (442)
POLTALLOCH current location (443); discovery (443); site (444); locality (447); description of object (448); description of ogham inscription (448); form of script (449);
443
CONTENTS
Vll
interpretation (450); discussion (453); bibliography (454); previous readings (455)
POOL
456 discovery (456); site (456); description of object (458); carving technique (459); description of ogham inscription (460); fonn of script (461); interpretation of text (463); discussion (465); bibliography (466); previous readings (466)
ST. N I N I A N ' S
ISLE
467
discovery (467); site (467); roman alphabet inscription (471); description of ogham-inscribed object (471); description of ogham inscription (472); fonn of script (474); interpretation of text (476); discussion (478); bibliography (479); previous readings (479)
S COON IE
480
discovery (480); site (480); description of object (480); description of carving decoration (481); description of carving - ogham inscription (483); fonn of script (486); interpretation of text (486); discussion (491); bibliography (493); previous readings (494)
WHITENESS
495
discovery (495); site (495); description of object (496); description of carving decoration (497); description of ogham inscription (498); fonn of script (500); interpretation of text (501); discussion (501); bibliography (502); previous readings (502)
APPENDIX DUBIA
ABERNETHY
503
2
discovery (504); description of object (504); description of carving (504); discussion (505); bibliography (506); previous readings (506)
504
viii
CONTENTS BROUGH
OF
BIRSAY
507
discovery (507); description of object (507); description of carving (507); discussion (508); bibliography (508)
BURRIAN
509
discovery (509); description of object (509); description of carving (509); bibliography (510)
DRUMOYNE
511
FOSHIGARRY
512
discovery (512); description of object (512); description of carving (512); bibliography (513)
GIGHA
2
514
current location (514); description of stone (514); description of carving (514); discussion (514); bibliography (515)
G U RN ESS
516
discovery (516); description of objects (516); bibliography (517)
IONA
518
L OCHNAW
519
discovery (519); description of object (519); description of carving - crosses (519); description of carving - 'ogham' (520); discussion (523); bibliography (525)
POOL
526 discovery (526); description of object (526); description of carving (526); bibliography (527)
TOLLARD
HOUSE
528
CONTENTS
ix
VAIVOE
529
OTHERS
530
BIBLIOGRAPHY OF WORKS CITED
531
Envoi
NB. Bound in two volumes with continuous pagination:
Volume I i-lxxxvi, 1-226 Prefatory Matter; Introduction; Guide to Entries; Catalogue from Abemethy to Cunningsburgh
Volume 2 227-553 Table of Contents (Vol. II), Catalogue from Dunadd to Whiteness, Appendix: Dubia, Bibliography
X
ACKNOWLEDGEMENTS
Many individuals and institutions have provided generous support to me, financial, logistical, intellectual, or emotional, during my eight years in tlie doctoral program of tlie Department of Celtic Languages and Literatures, Harvard University. It is impossible to name all those who have offered help and advice, but I am particularly indebted to tlie providers of major financial assistance, and I wish to record my gratitude to them at the outset.
Without a scholarship from tlie St. Andrews Society of tlie State of New York I could never have come to Harvard in tlie first place, and without tlie generous financial aid of tlie University itself, I could not have stayed. My first few months back in Scotland were partly financed by Harvard in the form of a Jens Aubrey Westengard Travelling Scholarship, a Radcliffe Grant for Graduate Women, and a Travelling Scholarship from the Dept. of Celtic Languages and Literatures.
The difficulty of juggling paid
employment and dissertation work in tlie following years made me all tlie more appreciative of tlie chance to devote my energies to full-time study during my tenure of tlie Sir John Rh£s Memorial Studentship in Celtic Studies, in tlie University of Oxford, and subsequently as a Julia Mann Junior Research Fellow at St. Hilda's College, Oxford. I wish also to record my gratitude for tlie smaller grants I received for specific purposes; tlie grant from tlie St. Andrew's Society of Washington D.C. which enabled me to purchase the computer on which all the work for this project was done, the grant from tlie President's Fund of tlie Edinburgh Association of University Women which helped defray the cost of photographs and other illustrations and of printing and binding tlie dissertation, and the grant from tlie Society of Antiquaries of Scotland which enabled me to travel to Orkney to conduct fieldwork.
I have lived in four countries while working on this project and have been fortunate in the support I have received from various institutions, some of which have even become for a while my home. The Department of Celtic in die University of Edinburgh allowed me to affiliate as a post-graduate worker and
ACKNOWLEDGEMENTS
xi
thus provided a lone scholar with logistical support and a stimulating intellectual and social environment. My colleagues in the Special Collections Department of the Library of the University of Edinburgh made earning my daily bread so much less of a chore. In Oxford I have indeed been fortunate, Jesus College and St. Hilda's College have given much and asked little in return. It has been a privilege and a pleasure to be a member of both institutions.
The staff of the NMR in Edinburgh were most helpful while I was conducting archival research there and I am particularly grateful to Graham Ritchie and David Easton for their assistance. The field-work for this study, which was conducted in Ireland and throughout Scotland, would have been impossible without the friendly co-operation of land-owners, church key-holders, and museum attendants. I wish to record my particular gratitude to the following curators who allowed me access to ogham inscriptions in their collections or otherwise assisted me: Colleen Batey (Kelvingrove Museum and Art Gallery), Anne Bnindle (Tankeraess House Museum), Mike King (Perth Art Gallery and Museum), Bernard Meehan (Library of Trinity College, Dublin), Raghnall 6 Floinn (National Museum of Ireland), Siobhan Ratchford and Alfred Truckell (The Observatory, Dumfries), Susan Youngs (British Museum), and especially Michael Spearman (NMS). Christopher Morris enabled me to examine the Birsay 3 ogham and I wish to thank him, John Hunter, and Anna Ritchie, for providing me with photographs and information about the oghams they excavated. Other practical assistance came from Marianna Lines, Niall Robertson, Tom Gray, and other members of the Pictish Arts Society.
The Department of Celtic Languages and Literatures has been very tolerant of a student who was, for five of her eight years, in absentia and I wish to record in particular my gratitude to my Chairman, Patrick Ford. I am ever more aware of my debt to those who taught me at Harvard - John Carey, John Koch, and Nerys Patterson - and to my fellow graduate students in the Departments of Celtic and of History. As a supervisor John Koch provided humour, support, and, above all, freedom. It was he who first encouraged
ACKNOWLEDGEMENTS
xii
me to pursue this topic and for this, as for so much else, I am very grateful to him.
Perhaps the most pleasurable aspect of working on this project is the many friendships it has spawned. It would be impossible to name all those who have sustained me through their friendship, and fostered my general intellectual development over the last eight years but I am acutely conscious of my debt to them. I have endeavoured to acknowledge all specific suggestions where appropriate in the text, but in addition [ wish to record here my profound gratitude to the following friends and colleagues who have given particularly generously of their time and expertise to offer advice and encouragement: Dauvit Broun, Thomas Charles-Edwards, Thomas Clancy, Stephen Driscoll, Joe Eska, Ian Fisher, Anthony Harvey, Isabel Henderson, John Higgitt, Jim Livesey, Damian McManus, Fionnbarr Moore, Elisabeth Okasha, Oliver Padel, Ross Samson, Richard Sharpe, Simon Taylor, Charles Thomas, Alex Woolf, and Patrick Wormald.
I have left till last my greatest debt, to my parents Moira and Alastair Forsyth. Their unerring support has been financial, logistical, intellectual, and emotional. This work could have been neither begun nor concluded without them, and it gives me the greatest pleasure to be able, at last, to dedicate it to them, with love.
K. S. F.
St. Hilda's College, Oxford March 1996
xiii
To my parents
xiv
PREFACE
My aim in this study has been, not to write a history of ogham in Scotland, but rather to provide the materials from which such a history might be written. Included are all inscriptions in the ogham alphabet from within the boundaries of modem Scotland, regardless of language of text or supposed ethnicity of carver. This marks a major departure from previous studies which have separated out supposedly Tictish' inscriptions from 'Irish-looking'. With the exception of Blackwaterfoot and Gigha, I have examined all inscriptions personally. In addition, I have visited the find-spot or original location of most of the oghams from Mainland Scotland and Orkney.
The main part of the corpus consists of individual entries, one for each site which has produced an ogham inscription. As well as providing a description of die inscribed object, a detailed account of its inscription, comments on the form of the script used, and notes towards an interpretation of the text, I have attempted to provide a broad historical and archaeological context for each ogham. I feel this too, is an innovation, one in keeping with the new topographically-aware Insular epigraphy championed by Charles Thomas. Those currently active in the field are increasingly coming to see inscribed stones, not merely as vehicles for text, but as monuments in a social and historical landscape. To view the ogham inscriptions of Scotland from this perspective engenders a radical reappraisal of their importance for our understanding of the language and culture of early medieval north Britain.
The last scholar to address this topic was Oliver Padel, whose 1972 University of Edinburgh MLitt. thesis, Inscriptions ofPialand, has remained the final word on the subject for almost a quarter of a century. The present readings were made independently of PadeFs, but were checked against his (and those of previous scholars). I found his readings uniformly accurate and sensitive, and only rarely did I find myself in disagreement. As a master's thesis, Inscriptions ofPictland was necessarily restricted in scope. Padel provided a thorough examination of each inscription, but, quite naturally, did not stray from the
PREFACE
xv
interpretive framework set out for ogham by his supervisor Kenneth Jackson. Padel did not question, for instance, the traditional model of the Dalriadic import of ogham into Scotland and the separate development of the script there, both assumptions which I would argue are no longer tenable.
Included in the present study are eight oghatns not covered by Padel. Two, Gigha and Poltalloch, which he excluded as not 'Pictish',1 and six which have come to light since he wrote: Blackwaterfoot 1-2, Birsay 3, Dupplin, Lochgoilhead, and Pool. More important, however, than new finds, are die twin revolutions in ogham scholarship and Pictish studies which have occurred over the last decade. Work done by McManus and Harvey from the mid-1980s, and more recently by Sims-Williams, has transformed our understanding of ogham. The script has been divested of its cryptic and druidic overtones and, far from being a crude and clumsy cipher, has come to be appreciated as a sophisticated indigenous response to the challenge of roman alphabet literacy. In Scotland, archaeologists, notably Leslie Alcock and his pupils, have led the way forward to a de-mystification of the Picts, who are now increasingly seen as a Celtic people, typical among their barbarians neighbours on the northern and western fringes of early medieval Europe.
My aim has been to bring to bear on Scottish ogham these new approaches, to look at the inscriptions again from a fresh perspective, and thereby to challenge the traditional orthodoxy on the Picts, Pictish, and 'Pictish' ogham. The technical complexity of the material means the work is necessarily at an interim stage. Any conclusions remain tentative. In the Introduction which follows I make some brief general observations and point to a number of ways in which future research might develop. I have not set out to produce a 'Grammar of Pictish' or to write another chapter in the history of ogham. I hope, however, that the present work has laid the ground-work which will bring such goals a little closer.
1 He included two oghams, Weeting (Norfolk) and Kirkmichael (Isle of Man), which I have not treated since they arc not in Scotland. I would argue, in any case, argue that the label 'Pictish' is inappropriate for them. I have also excluded the six roman alphabet inscriptions from Pictland mentioned by Padel.
xvi
ILLUSTRATIONS
Illustrations (unpaginated) are grouped together at tlie end of tlie relevant entry.
Unless otherwise
specified, all maps, drawings, and photographs are by the author. I am grateful to Frances Hood, John Hunter, Christopher Morris, Anna Ritchie, and especially to Michael Spearman (NMS) for providing photographs and, where appropriate, giving permission for their use. Unless otherise stated as schematic, drawings of inscriptions are from photographs and are to scale. The numbers in such diagrams correspond to numbered sections in tlie discussion of each inscription.
INTRODUCTION
Transliteration key to the ogham alphabet [after Thomas 1994 fig. 3.1] The 'Ogham-band' The Palaeography of Ogham (explanatory diagrams): Slope, profile, spacing, special characters, etc.
LOCATION MAPS: The ogham inscriptions of D41 Riada The ogham inscriptions of Orkney Tlie ogham inscriptions of the eastern Mainland (south) The ogham inscriptions of tlie North-East Caithness and Sutherland. Location of sites mentioned in text [after Batey 1991 fig. 1] The ogham inscriptions of Shetland
CATALOGUE
Location Map. The ogham inscriptions of Scotland
ILLUSTRATIONS
xvii
ABERNETHY * Conjectural map of the early medieval shire of Abernethy [Driscoll 1991 fig. 5.3] * Ogham-inscribed fragment [Butler pi.II]; (key to numbering) * Side view of recumbent grave-marker, Meigle 11 [ECMS fig.345A] * Possible reconstruction of monument
ACKERGILL * Plan of cemetery [Close-Brooks 1984 fig.5.6] * Ogham-inscribed Symbol Stone [ECMS fig. 25] * Schematic representation of the ogham (key to numbering)
ALTYRE * Location map. Moray and Easter Ross showing extent of areas of prehistoric and proto-historic settlement [Jones et al. 1993:48 fig.3.1] * Pictish Moray [after Shepherd 1993:77 fig. 4.2] * Ogham-inscribed cross and detail of inscription [Calder & Jackson 1957 figs. 14, 15] - Ogham inscription, based on impression by Marianna Lines (plant matter on cotton) [reduced]; (key to numbering)
AUQUHOLLIE * Ogham-inscribed pillar (detail) * Ogham-inscribed pillar * Ogham inscription (key to numbering)
BAC MHIC CONNAIN -
Site Location maps. Above: Bac Mhic Connain and the wheelhouse sites of Vallay Sound [Hallén
ILLUSTRATIONS
xviii
1994:190 Illus. 1]. Below: As above showing relief [Scott 1948:69 fig.7J. •
Ogham-inscribed knife handle [Hallén 1994:220 Illus. 12.5]
•
Plans of the structures at Bac Mhic Connain [Hallén 1994:191 Illus.2]
•
Ogham-inscribed knife handle [® Trustees of the National Museums of Scotland, Neg. P9557]
-
Ogham inscription (key to numbering)
BIRSAY •
Location Map. Orkney with inset Brough of Birsay and Point of Buckquoy [after Morris 1990:70 fig.3]
•
Plan of main Pictish and Norse settlement [Curie 1982 ill. 3]
•
Sculptured slab with Pictish symbols and figures [Curie 1982 ill. 4]
•
Cross-incised stones: Large [Ritchie 1986:12], small [Curie 1982 ill.45 (detail - no.607)] Birsay 1
- Rubbing of ogham inscription [Photo. Department of the Environment (Historic Scotland) No. A 2894-2, June 1971 (Crown Copyright)] • Ogham inscription [Padel 1972:56] • Schematic representation of ogham (key to numbering)
Birsay 2
• Rubbing of ogham inscription [Photo. Department of the Environment (Historic Scotland) No. A 2894-1, June 1971 (Crown Copyright)] • Schematic representation of ogham (key to numbering)
Birsay 3
* Ogham-inscribed slab, upper and lower feces [photo, supplied by Prof. C D . Morris, Dept. of Archaeology, University of Glasgow, Reg. BB80] • Ogham-inscribed slab,frontand back feces [photo, supplied by Prof. C. D. Morris, Dept. of Archaeology, University of Glasgow, Reg. BB80] • Ogham inscription • Schematic representation of ogham (key to numbering)
•
Rune-inscribed fragments, Birsay II (Barnes OR 8) [photo. Historic Scotland, Neg. A 995-1
xix
ILLUSTRATIONS (1938), Crown Copyright] •
Rune-inscribed fragment, Birsay III (Barnes OR 9) [photo. Historic Scotland, Neg. A 995-3 (1938), Crown Copyright]
•
Rune-inscribed seal's tooth, Birsay IV (Barnes OR 11) [Curie 1982:ill.37]
BLACKWATERFOOT •
King's Cave - Exterior of King's Cave [Balfour 1910 pl.XXXIII. I]; Plan of interior showing area of Balfour's excavation [Balfour 1910.214]
• Interior of cave showing position of ogham inscriptions [®RCAHMS Neg. B69005] Blackwaterfoot 1 - Ogham inscription [Jackson 1971 pi. IX]; Schematic representation of ogham inscription (key to numbering) Blackwaterfoot 2 • Ogham inscription [®RCAHMS Neg. B69011]; • Ogham inscription; Schematic representation (key to numbering)
BRANDSBUTT •
Ogham-inscribed Class I Pictish Symbol Stone [®University of Aberdeen, Anthropologcial Museum]
• Ogham-inscribed Class I Pictish Symbol Stone • Ogham inscription (key to numbering)
BRESSAY •
Ogham-inscribed cross-slab [Close-Brooks & Stevenson 1982:35]
• Ogham-inscribed cross-slab [n.a. 1865, pl.XII, based on Stuart 1856, pl.XCIIII-XCV] •
Cross-slab from Papil, Shetland [^Trustees of the National Museums of Scotland, Reg. IB 46]
• Ogham inscription on left edge [^Trustees of the National Museums of Scotland, Reg. IB 119] (key to numbering)
ILLUSTRATIONS
XX
• Ogham inscription on right edge [^Trustees of the National Museums of Scotland, Reg. IB 119] (key to numbering)
BRODIE • Ogham-inscribed cross-slab (reverse face) [Foster 1996 fig.48 (Tom Gray)] • Ogham-inscribed cross-slab showing position of inscriptions [ECMS fig. 136] • Front right: schematic representation of ogham inscription (key to numbering) • Back left: Detail of ogham (key to numbering) [After Foster 1996 fig.48] •
Back right: Detail of ogham (key to numbering) [After Foster 1996 fig.48]
BUCKQUOY •
'Figure-of-eight* house in which inscribed whorl was found [Ritchie 1977:177 fig. 3]
-
Ogham-inscribed spindle-whorl [Crown Copyright: Institute of Geological Sciences, NERC].
-
Ogham inscription (key to numbering)
BURRIAN • Location map [MacGregor 1974:64 fig.l] •
Broch site [RCAHMS 1946 fig.88]
•
Ogham-inscribed slab [^Trustees of the National Museums of Scotland, Neg. P9554]
•
Ogham-inscribed cross-slab [MacGregor 1974:97 fig.21
-
Ogham inscription (key to numbering)
-
Detail of cross
CUNNINGSBURGH • Location Map [Turner 1994:316 Illus.lJ Cunningsburgh 1 • Ogham-inscribed fragment [^Trustees of the National Museums of Scotland,
ILLUSTRATIONS
xxi Reg. IB 114] * Ogham-inscribed fragment, front and back; Schematic representation of the ogham (key to numbering)
Cunningsburgh 2 • Ogham-inscribed fragment [^Trustees of the National Museums of Scotland, Reg. IB 115] * Ogham-inscribed fragment; Schematic representation of the ogham, face and arris (key to numbering) Cunningsburgh 3 ' Ogham-inscribed fragment * Schematic representation of the ogham (key to numbering) * Macalister's reconstruction [Macalister 1940.216 pl.VIb]; Alternative reconstruction * Runic inscriptions. Above: Cunningsburgh 1 [ECMS fig. 14] Below: Cunningsburgh 3 [ECMS fig. 15]
DUNADD -
Plan of fort [Campbell & Lane 1993.53]. Rock carvings, including ogham inscription, at location a
* Incised rock surface with detail of boar [RCAHMS 1988:158] * Incised boar [RCAHMS 1988:157] * Incised rock surface with detail of ogham inscription [RCAHMS 1988:158 and RCAHMS 1988:159] * Ogham inscription, upper and lower lines (key to numbering) * Inscribed pebble (enlarged) [Okasha 1985 pl.VIII] * Detail of inscription
ILLUSTRATIONS
xxii
DUPPLIN * Location of the Dupplin cross in relation to Forteviot. D, Dupplin Cross; H, Haly Hill; S, scarp of the former course of the Water of May; I, site of Invermay cross. Hatching, areas of cropmarks [Alcock & Alcock 1993 illus. 12] * Free-standing cross [Alcock & Alcock 1992: illus. 13-14]. * Romilly Allen's drawing of the Dupplin Cross [ECMS fig.334]. • The base of the Dupplin Cross showing traces of ogham lettering [® The Trustees of the National Museums of Scotland |. -
Inscribed panel on the west lace of the shaft of the Dupplin cross [® The Trustees of the National Museums of Scotland].
• Text panel (cast) with lettering marked with charcoal - lines 1-2. • Text panel (cast) with lettering marked with charcoal - lines 1-3, 7. • Sculptured stones from Forteviot (not to scale). Fragment of cross-arm [Alcock & Alcock 1992 illus.5]; Front, side, and rear view of cross-slab Forteviot 1. Height 0.6m [Alcock & Alcock 1992 illus.4 (Tom Gray)]; Arch [Alcock & Alcock 1992 illus.6]
FORMASTON * Fragment of ogham-inscribed cross-slab [^Trustees of the National Museums of Scotland, Reg. IB 111] •
Reconstruction of original monument [Macalister 1940:213 pl.v]
- Cross-slab from Kinnord, Aberdeenshire. - Ogham inscription, inner and outer lines (key to numbering)
GIGHA • Ogham-inscribed pillar [RCAHMS 1971 pi. 16]
ILLUSTRATIONS
xxiii
* Comparison of previous readings: (a) Macalister 1902; (b) Diack 1926; (c) Jackson (RCAHMS) 1971; (d) Key to numbering of letters
GOLSPIE * Ogham-inscribed cross-slab [Close-Brooks 1989:cover, 9] * Ogham-inscribed cross-slab. Side view showing ogham [ECMS 48A] -
Ogham inscription (key to numbering)
GURNESS * Plan of the Post-Broch period structures [Hedges 1987.11:64 fig.2.11] -
Plan of die 'Shamrock* and 'Annexe* [Hedges 1987.11:66 fig.2.12]
-
Carved stones recovered from die site. Above: Pictish symbol stone. Below: Cross-incised slab, unknown date. [Hedges 1987.11:125 fig.2.51, 126 fig.2.52]
* Ogham-inscribed knife handle [Hedges 1987.11:96 figs. 2.22, 118 fig.2.44, details] * Ogham inscription (key to numbering)
INCHYRA -
Symbol and ogham-incised slab [RCAHMS 1994a:95a]
-
Symbol and ogham-incised slab showing position of inscriptions [Stevenson 1959 pi.Ill]
-
'Unfinished' symbols and inscription A (i) [after Stevenson 1959 pi.IV] (key to numbering)
* Detail of damaged edge showing inscription A (ii) - key to numbering [photograph ® Perth Art Gallery and Museum] * Inscription B: Top: Edge and top of slab showing position of lettering [Stevenson 1959 pi. Ill]; Middle: Detail of inscription (key to numbering) [RCAHMS 1994a:93]; Bottom - Detail of lettering on edge [Henderson 1967 pi.24] * Detail of broad end of slab showing symbols and inscription C (key to numbering) [Stevenson
ILLUSTRATIONS
XXIV
1959 pl.IIII
LATHERON • Ogham inscribed cross-slab [Curie 1940 pi.xxii] • Ogham-inscribed cross-slab • Ogham inscription (key to numbering)
LOCHGOILHEAD Faces b and c [photograph RCAHMS Argyll.VII.194 B,C] Faces a-c [drawing RCAHMS Argyll.VII. 194 A(l)] Rubbing of face b (reduced to 60% actual size) Selection of roman letters (actual size) Rubbing of edge c (reduced to 60% actual size) Schematic representation of ogham inscription (key to numbering) Ogham inscription (actual size)
LOGIE ELPHINSTONE • Class I Pictish Symbol Stones at Logie Elphinstone: Logie Elphinstone 1 [ECMS fig. 188], Logie Elphinstone 3 [ECMS fig. 191] •
Impression by Marianna Lines (plant matter on cotton) [Lines 1989]
• Detail of ogham inscription (key to numbering)
LUNNASTING •
Ogham-inscribed slab [Goudie 1878 fig.2 (=1901:36); ECMS fig.12]
• Ogham text with key to numbering
ILLUSTRATIONS
XXV
NEWTON •
Location Map. OS First Edition, Sheet 44 (detail), showing Newton House, Bridgend of Shevack, Wood of Pitmachie, Bridge of Pitmachie, and Old Rayne. Scale - 6W to the mile.
• Class I Pictish Symbol Stone [ECMS fig. 193] •
NMS Cast of ogham-inscribed pillar [® Trustees of the National Museums of Scotland, Reg. IB 108]; (key to numbering)
•
Ogham inscription (key to numbering)
•
Detail of roman alphabet inscription [Simpson 1940 pi. 60]
-
Roman alphabet inscription
POLTALLOCH •
Ogham-inscribed fragment [®Tmstees of the National Museums of Scotland, Neg. P9555]; (key to numbering)
POOL •
Location map [after Hunter 1990:176 ill. 10.1]
•
Symbol-inscribed stone, J. R. F. Burt [Nicoll 1995:179 fig.31]
•
Ogham-inscribed slab (detail) [photograph supplied by J.R. Hunter, Dept. of Archaeological Sciences, University of Bradford]
-
Ogham inscription (key to numbering)
ST. NINIAN'S ISLE • Location Map [Close-Brooks 1981:3] •
Ogham-inscribed slab [^Trustees of the National Museums of Scotland, Reg. IB 112]; (key to numbering)
-
Inscribed sword chape (front and back) [Close-Brooks 1981 fig. 15]
-
Inscribed sword chape: Detail of inscription [Jackson 1973b fig.29], Form of script used [Brown
ILLUSTRATIONS 1993:246]
SCOONIE • Ogham-inscribed cross-slab [ECMS fig.360] • Ogham inscription with key to numbering - Sites associated with the cult of Ethernan
WHITENESS • Ogham-inscribed fragment • Stevenson's reconstruction of interlace pattern [1981 fig.l] • Two possible reconstructions • Schematic representation of ogham inscription (key to numbering) • Ogham-inscribed fragment [Stevenson 1947 pl.22 (4)] • Fragment of cross-slab NMS IB 248 [Tait 1938 fig.5]
APPENDIX: DUBIA
Location Map. The ogham inscriptions of Scotland: Dubia
ABERNETHY 2 • Sketch by the Earl of Southesk [1895:249]
BIRSAY • Allegedly ogham-inscribed pebble [Curie 1982:ill.45 detail]
ILLUSTRATIONS
xxvii
BURRIAN * Spindle-whorl inscribed with allegedly ogham-like markings [MacGregor 1974:91 fig 18, detail]
DRUMOYNE * Pillar with incised linear markings [Photo. A. MacLennan, Kelvingrove An Gallery & Museum] * Linear carvings
FOSHIGARRY * Bone disc inscribed with allegedly ogham-like markings [Hallgn 1994:218 Illus,5|
GIGHA 2 * Lower portion of pillar bearing ogham-like markings [Photo. Frances Hood]
GURNESS * Bone casket-lid allegedly inscribed with ogham [Hedges 1987.11:106 fig.2.32, detail]
LOCHNAW * Pillar fragment * Detail of carving with key to numbering
POOL * Bone pin with ogham-like markings [after Holder 1990:72]
ENVOI 'Hey, you've missed out theiW J.D. Moir Pictish Arts Society Newsletter 5, Spring 1990, p.5
xxviii
ABBREVIATIONS
AUTHORITIES AND GENERAL - See also bibliography
ATig
'Annals of Tigernach\ ed. Whitley Stokes, Revue celtique 16 (1895) 374-419; 17 (1896) 9-33, 119-263, 337-420; 18 (1897) 9-59, 150-97, 267-303.
AU
The Annals of Ulster (to A.D. 1131), ed. Sean Mac Airt & Gearoid Mac Niocaill, Dublin 1983.
CGH
Corpus Genealogiarum Hiberniae, Vol.1, Michael A. O'Brien, Dublin Institute for Advanced Studies, 1962.
CIIC
Corpus Inscriptionem Insularum Celticarum, 2 Vols. (1945, 1949), R. A. S. Macalister, Dublin: Stationery Office [numbering, by inscription, continuous through both volumes].
DIL
Dictionary of the Irish Language Based Mainly on Old and Middle Irish Materials: Compact Edition, Royal Irish Academy, Dublin: 1983.
DNB
The Dictionary of National Biography
ECMS
Early Christian Monuments of Scotland, James Romilly Allen & Joseph Anderson, Edinburgh: Society of Antiquaries of Scotland, 1903 [unless otherwise specified, references are to Pan III].
HE
Bedefs Ecclesiastical History of the English People, ed. and transl. Bertram Colgrave &
ABBREVIATIONS
XXIX
R. A. B. Mynors, Oxford: Clarendon, 1969 [cited by book and chapter number].
NGR
National Grid Reference
NMR
National Monuments Record of Scotland, RCAHMS, John Sinclair House, Bernard Terrace, Edinburgh
NMS
National Museums of Scotland (and bodies to which it is the successor institution, e.g.
National Museum of Antiquities of Scotland, Edinburgh)
NSA
The New Statistical Account of Scotland, 15 Vols., Edinburgh, 1845.
OD
Ordnance Datum
OS
Ordnance Survey
OSA
The Statistical Account of Scotland, 21 Vols., Edinburgh, 1791-9.
PSAS
Proceedings of the Society of Antiquaries of Scotland, Edinburgh.
RCAHMS
The Royal Commission on Ancient and Historic Monuments of Scotland.
VC
Adomndn's LifeofColumba, ed. and transl. Alan O. Anderson & Marjorie O. Anderson, Edinburgh: Thomas Nelson, 1961 [cited by book and chapter number, all translations Anderson 1991 J.
ABBREVIATIONS
XXX
STANDARD ABBREVIATIONS FOR THE PRE-1974 COUNTIES OF SCOTLAND
ABD Aberdeenshire
RNF
Renfrewshire
ANG Angus
ROS
Ross and Cromarty
ARG Argyllshire
ROX
Roxburghshire
AYR Ayrshire
SHE
Shetland
BNF
Banffshire
STL
Stirlingshire
BTE
Bute
SUT
Sutherland
BWK Berwickshire
WIG
Wigtownshire
CAI
WLO
West Lothian
Caithness
CLA Clackmannanshire DMF Dumfriesshire DNB Dunbartonshire ELO
East Lothian
FIF
Fife
INV
Inverness-shire
KCD Kincardineshire KNR Kinross-shire LAN
Lanarkshire
MLO Midlothian MOR Morayshire NAI
Nairnshire
ORK Orkney PEB
Peebles-shire
PER
Perthshire
ABBREVIATIONS
xxxi
OTHER ABBREVIATIONS AND CONVENTIONS
Languages
OW.
Old Welsh
Olr.
Old Irish
MW.
Middle Welsh
EOIr.
Early Old Irish
OC.
Old Cornish
Modlr. Modern Irish
OB.
Old Breton
ON.
Old Norse
WCB.
Welsh/Comish/Breton
L.
Latin
Symbols
<
developed from
>
developed into
*
unrecorded hypothetical form
Others
s.a.
sub anno (in Annals, under the year)
nom.
nominative
s.v.
sub verbo (under die word or heading)
gen.
genitive
s.n.
sub nomine (under a specified name)
sg./pl.
singular/plural
For conventions used in ogham transcription, see Introduction: Transcribing an ogham inscription.
xxx ii
INTRODUCTION
THE OGHAM SCRIPT
DESCRIBING OGHAM The Ogham character is an alphabetic form of writing in which ... letters ... are denoted by combinations of parallel strokes in number from one to five, set in varied positions with respect to a central stem-line' [Macalister 1945 :iv) The external form and internal structure of the ogham alphabet has been amply covered by McManus [1991:1-43), and there is no need to repeat his discussion here. What follows is a brief explanation of the methodology employed in the catalogue for describing ogham inscriptions with a note of special terminology adopted. Learned Irish texts, such as Auraicept na nÉces 'The Scholar's Primer' [Calder 1917 l.lff/2260ff; Ahlqvist 1992], De dúilibfeda naforfid (a short tract on the supplementary letters) [Calder 1917 1.5416-5463J, and In Lebor Ogaim 'The Book of Ogam' (a.k.a. 'The Ogam Tract') [Calder 1917 1.5465ff; McManus 1991:138-40J), reveal a developed vocabulary for referring to various aspects of ogham [McManus 1991:3]. However, in the interests of accessibility, I have, in general, preferred to translate the Irish terms. Thus 'stem' or 'stem-line' is used for the hypothetical base-line, not drulm ('ridge', 'edge', 'back'), and each component part of a letter is labelled a stroke, notflesc'twig', ('score' is more common in earlier literature but in order to be able to discuss the script independently of epigraphic manifestations of it, I have favoured the more neutral 'stroke', and kept 'score' for incised lines regardless of whether or not they are letter-strokes). I have, however, kept aicme (pi. aicmi, 'family', 'class', 'group') to refer to the four 'quinary groups' into which the original alphabet was divided Traditionally, each group was known by the name of its first letter, thus: Aicme Beithe (BLVSN), Aicme húatha (HDTCQ), Aicme Muine (MGGWSIR), Aicme Ailme (AOUEI). I have compromised with b-aicme, h-aicme, etc. By extension, I use 'b-strokes' to refer to letter strokes on the 'b-surface', i.e. to the right of the
xxxiii
INTRODUCTION
stem, and 'h-strokes' and 'h-surface' to the left [cf. Macalister 1945:xv|. I have not used Irish táebomnai 'consonants', ocfeda 'vowels' (or more generally 'letters'), but I have kept foifeda (sg. forfid) for the assortment of supplementary characters subsequently added to the original alphabet (see below). SimsWilliams uses a short-hand for identifying letters by their position within the aicme, thus h4 is 'the fourth member of the h-aicme, i.e. C \ m3 is 'the third member of the m-aicme, i.e. ?GW' [1993]. This system is very useful for referring to individual letters, for which purpose it is adopted below, but is too cumbersome for transcribing full texts.
Like its model the roman alphabet, ogham is read along the stem from left to right. Thus when disposed vertically, as was always standard epigraphic practice, ogham reads upwards with the reader's head inclined to the left (there is one definite case in Scotland of a top-down vertical text: still, of course, reading left-toright, but with die reader's head turned to the right). Even in the manuscripts, where ogham was habitually written on a horizontal stem, texts are described as if arranged vertically, like the orthodox inscriptions [McManus 1991:3]. I have followed this convention, regardless of the actual alignment of the stem, referring to h-strokes as being to the left (i.e. 'above' a horizontal stem) and b-strokes to the right ('below' a horizontal stem).
Similarly, the directions 'forward' and 'backward', are relative to the
direction of reading.
Native writers were not concerned with the palaeography of ogham (i.e. the style in which the characters were written) any more than they debated the form of uncials, half-uncials, or Insular minuscules employed in their manuscripts. Thus in order to discuss the finer points of the script it is necessary to generate a few new terms. In referring to different parts of a letter-stroke, I follow Macalister in using the terms 'proximal tip' and 'distal tip', to refer to the points on the stroke nearest to and furthest from the stem [1945:xv]. Long strokes which cross the stem have both right and left proximal and distal tips. Strokes which intersect with the stem at 90° are said to have no slope. The closer the angle to 45°, the greater
INTRODUCTION
xxxiv
or sharper the slope is said to be (angles lower than 45° appear to have been avoided). The direction of slope is said to be •forward* if the stroke's proximal tip is further •forward* relative to the direction of reading than its distal. If they slope at all, b-strokes habitually slope forward, and h-strokes backward. If they slope, m-strokes habitually slope forward, i.e. with their left distal tips 'forward' of the point at which they cross the stem. In some later inscriptions it was common practice to enhance legibility by linking the component strokes of a single letter with a horizontal stroke through their distal tips. This is a 'bind-stroke', and such letters are said to be 'bound'.
In order to enable discussion of the relative length of strokes of different aicmi, I have generated the term 'ogham-band' to refer to the width of the area occupied by the strokes. This is dictated in each case by the distance from the stem of the distal tips of the longest letter-strokes, usually the b- and h-strokes, but occasionally the m-strokes (the middle of the band is, of course, the stem). If a b-stroke occupies half the ogham-band, a vowel-stroke of equal length would occupy the middle two quarters of the ogham-band, one either side of the stem. The length of vowel-strokes in particular varies greatly between texts, the longest occupy the full width of the ogham-band, shorter ones might occupy less than an eighth of it. An m-stroke occupying the full width of the ogham-band is of course longer than a full-width vowel-stroke because the former is oblique and the latter perpendicular.
TRANSLITERATING OGHAM 'It is difficult to believe ... that so grotesque a word as hccwew ever had an actual place in the language of a people civilised enought to practise writing' [Macalister 1902:163] Nineteenth century authors went to the trouble of creating ogham-letter typefaces with which they were able to refer to particular ogham characters directly and unambiguously. Today, however, the inadequacies of commercial fonts, and, perhaps more importantly, the convenience of readers (!) requires that in discussion, ogham characters should be transliterated into letters of the roman alphabet. This is not as easy as it
THE OGHAM ALPHABET. TRANSLITERATION KEY [After Thomas 1994 fig.3.1]
OGHAM-BAND
Vowel-strokes occupying Va band
Vowel-strokes occupying XA band
Vowel-strokes occupying lA band
Vowel-strokes occupying full band
INTRODUCTION
XXXV
sounds. It is clear that in the later period (eighth/ninth century and later) the ogham characters were used in Ireland largely as a cipher for contemporary manuscript spelling. That the correspondence was on a graphemic level (letter to letter) is shown by the way in which it was possible to write Latin, substituting ogham letters for roman, as seen, for instance, in a margin of the St. Gall Priscian [Nigra 1872:15] and the Bodliean Annals of Innisfallen [Macalister 1910]. According to the key preserved in the manuscripts, the standard transliteration in this period was: BLFSN
HDTCQ
MGNQZR
AOUEI
Earlier scholarship took the manuscript tradition as its starting point and all previous transliterations of the Scottish oghams have been based on this key.
More recent work on the structure of the alphabet,
specifically the names of the letters, by McManus [1986, 1988, 1991] and subsequently by Sims-Williams [ 1993], has shown, however, that the above key does not accurately reflect the original values of the ogham symbols. Furthermore, McManus has been able to show that the framers of the ogham alphabet had in mind, not a letter-to-letter correspondence with Latin and the roman alphabet, but rather a letter-to-sound correspondence with the phonemic inventory of their own Irish language. In other words, that 'the values of the Ogam characters are not to be regarded as those of the Latin alphabet arranged topsy-turvy and camouflaged in a primitive Morse code, but rather as the sounds of Primitive Irish as perceived by the inventors' [McManus 1991:31]. These earlier values have been reconstructed with reasonable certainty as follows (italicized letters tentative): BLVSN
ffDTCQ
MG^JR
AOUEI
Since, in the earlier period, the correspondence was letter-sound, it would seem sensible to avoid transliteration with the roman alphabet (which introduces another layer of letter-sound correspondences), and instead, to go straight to a reconstructed phonetic/phonemic transcription enclosed within oblique strokes, //. Using the roman alphabet might, in any case, support the erroneous impression of ogham as cipher for die roman letters, rather than an alphabet in it own right. A phonetic/phonemic transcription
INTRODUCTION
xxxvi
would have the added advantage of dissimilation, reinforcing Harvey's message the Insular orthography is not to be taken at face-value [1989, 1990a, 1990b, 1991].
The draw-backs of such an approach, however, outweigh its advantages. Writing is not speaking, and all spelling is conventional, ogham included. To attempt a phonemic, let alone, a phonetic transcription would be to imply far greater precision than is attainable on present knowledge. It would be impossibly complex too, for pronunciation was dialectally variable and ever-evolving. The Scottish ogham inscriptions cover a range of dates from perhaps the sixth century to the tenth. To reflect the developing language, a new transcription scheme would be required virtually for each inscription (and in any case we have insufficient knowledge of the historical phonology of the early medieval languages of Scotland even to attempt it). So instead we must accept a system of one-to-one correspondences, aware that to do so is not to assign soundvalues, but merely to produce a conventional transliteration which requires further interpretation. That these transliterated oghams in roman letters are to be distinguished from the equivalent manuscript roman letters is indicated here by the use of capitals for the former, italicized lower case for the latter.
If all the texts in Scotland were written in a Goidelic language1 there would be little problem in transliterating them according to the revised McManus/Sims-Williams key, providing, of course, due allowance was made for the date of the inscription (thus for the appropriate stage in, for instance the delabialization of Q /kw/; V /w/ > F /f/, etc.). Certain features of the Scottish inscriptions, such as the apparent adherence to the old convention of using geminate symbols to indicated non-leiiited/nonspirantized consonants, indicate a degree of continuity with the old ogham orthography. Yet other features, such as the novel use of H, indicate that Scottish oghamists were prepared to innovate. Doubtless, this
1 One of the infelicities of established terminology is that the language common to early Ireland and the Western Highlands and Islands of Scotland is called by linguists Old Irish rather than Old Gaelic. When talking of general cultural or linguistic affinity I have used Irish' and 'Gaelic' loosely, avoiding the former in general discussion (when to use it would appear to exclude Goidclicspeaking Scotland), and favouring it in linguistic discussion (when to use 'Gaelic' in contravention of established custom would be potentially confusing). Specifically of languages, 'Old Irish' is intended to embrace the Goidelic language spoken on both sides of the North Channel.
INTRODUCTION
xxxvii
in part reflects the later date of the Scottish texts, the bulk of which date to after the main period of orthodox ogham inscriptions in Ireland. At least as important, however, was the fact that ogham was, for the first time, being used to represent a non-Goidelic language.
Although ogham was framed with the sounds of Primitive Irish in mind, since its starting point was the roman alphabet, there were inevitable orthographic deficiencies from the outset, notably the lack of sufficient symbols to differentiate the full inventory of Irish vowels. The incongruence was all the greater, however, when an attempt was made to use the twenty letters of the ogham alphabet to represent the sounds of a Brittonic language, the lack of phonemic correspondence generating both superfluous letters and uncatered-for sounds. To take just one obvious example, a Brittonic-speaking oghamist would have no need of ogham Q, yet would lack a symbol for the sound /p/. If the ogham alphabet was adopted in Scotland during the Primitive Irish period, then the value of the character Q at that time would have been /kw/. McManus has explained that its letter name, Old Irish Cert, was originally *kwerkw-, cognate with Latin quercus (ultimately from the root *peikw), cf. Welsh perth 'bush' [1991:37, cf. McManus 1986:15, 29]; the Pictish cognate is reflected in Scottish place-names in pert (Perth, Logiepert ABD) [Watson 1926:356-7]. If the borrowing was sufficiently early, it is possible that this fourth member of the h-aicme was assigned the value /p/, either because that is what the borrowers heard as the initial of *kweikw-, or because they were able to recognize the cognate word in their own language. If however, the alphabet was adopted at a date later than the sixth century, Primitive Irish /kw/ would have fallen together with /k/, for which there were now two symbols, Q and C. The correspondence between Irish mac and Welsh map was not lost on the Irish compiler of Cormac's Glossary, c.900: %Dind map Lethain i tlrib Bretan Corn ./. Dun MaicLiathain, ar is mac indnl is map isin Bremais ...' [O'Donovan 1868:111-2]. It may also have been obvious to Scottish oghamists a few centuries earlier. McManus has shown how the creation of the ogham alphabet 'was accompanied by a careful analysis of the sounds of... [the target] language1 [1991:31]. One wonders if the adoption of the alphabet by non-Goidels might be accompanied by a similarly careful
INTRODUCTION
xxxviii
analysis of the mis-match between donor alphabet and target language ? The creative input into Ogam was', he says,'quite considerable* [39]. If the oghamists of Scotland were both as linguistically aware and creative as the original framers of the alphabet, they may well have been able to take the redundant symbol 'Q' and assign to it the new and useful value /p/.
So what, then, was the value of the fourth member of the h-aicme in Scottish ogham inscriptions, /kw/, /k/, or /p/ ? The texts are surely all too late for /kw/, and /k/ is already catered for by C. Tellingly, die Q-character, appears in Scotland fcnly in the formula word MAQQ/MEQQ. Since we dius have no independent check on die value, we cannot know if in Pictland MAQQ might actually have meant /map/. Obviously, to adopt the transcription Q = /p/ would have a striking effect on our readings, rendering them suddenly a lot more 'Brittonic-looking'. Similarly with die third letter of the b-aicme, V /w/, which by the Old Irish period represented /f/, manuscript/. If this is transcribed instead by its Pictish cognate /w/, manuscript uu- (cf. Welsh gw-), the effect is equally transformative. Some Scottish oghams, however, were clearly carved in an Irish-speaking environment, where, presumably Q still meant /k/, and V meant HI. From this it can be seen diat assigning sound values to ogham graphemes is not straightforward. In transliterating the various characters, choosing one option over another can turn a Goidelic-looking text, into a Brittonic-looking one at a stroke. These difficulties must be born constantly in mind, but in an attempt to avoid prejudging die issue, I have adopted, as far as possible, a neutral transliteration, i.e. V not F or W, Q, H, and left assigning possible sound-values to die linguistic discussion of each text. I have also adopted, as far as possible, the principle of one-fbr-one transliteration (exceptions are Gw and S'). In the case of forfeda two or more letters standing for a single character are underlined, e.g. >$$$(r BRSince the letters V and H occur frequently in the inscriptions, however, dieir possible values are discussed below:
INTRODUCTION Transliteration
xxxix
of 7 / ' (See Sims-Williams 1993:162-170)
The first letter of the h-aicme does not appear in classical ogham inscriptions and we do not know its original value. The equation with h, well attested on all media from 'at least the ninth century* [SimsWilliam 1993:162], 'is no more than a cosmetic solution to the problem created by the loss of the original initial consonant' [McManus 1991:36-7]. When the older ogham convention of indicating spirants with simplex as opposed to geminate consonants had broken down, and ogham orthography was merely a cipher for contemporary manuscript spelling, H came to be used with C and T as equivalents of the manuscript digraphs ch and th [for orthographic solutions to the problem of representing the spirants Ixl and /6/ see Harvey 1989, 1990; also Sims-Williams 1992:45-51]. In the period following the Second Spirantization, before the orthographic conventions had settled down, there is clear evidence that h could be used to represent Ixl* Three roman alphabet inscriptions, two from Wales, the otlier from Ireland, preserve this usage; broho- for /broxo-/ (capitals CIIC 349, 401) and Menueh (<*Minawikas) Inchagoill, Co. Galway (CIIC 1), the latter has been described as 'the oldest surviving Irish "text" written in the Latin alphabet' [McManus 1991:61]. H in Scottish oghams does not appear following C and T. Perhaps instead it is a continuation of the Irish convention of H = Ixl translated into ogham terms. The letter sometimes occurs singly, sometimes doubled. McManus has commented on the 'apparently meaningless' and 'capricious interchange of single and double consonants' in both ogham and early manuscript traditions [1991:124-6 §6.31]. When the contrast occurs, as at Lumiasting within the same inscription, or even the same word, however, one wonders if perhaps in at least some cases a linguistic distinction is being drawn. The lenition/non-lenition (spirantized/non-spirantized) opposition can scarcely apply to H, however, since the reconstructed root value is a spirant. From the admittedly limited sample, single/double could reflect an opposition palatal/non-palatal.
Transliteration
of 'V
(See Sims-Williams 1993:140-43)
Sims-Williams concedes that 'no transcription ... is wholly satisfactory in all circumstances' [1993:142].
INTRODUCTION
xl
The Primitive Irish value of the third member of the \xiicme, fern (<*werna) 'alder-tree', is /w/ (transliterated V) [McManus 1991:37-8; Sims-Williams 1993:140-143]. By the Old Irish period this sound had developed into /f/, hence the manuscript transliteration F. By the eighth century, the Welsh reflex of /w/ had become gw. There was already a character for this sound, the third member of the m-aicme, gétaU but in Irish that sound had long since fallen together with its non-labialized counterpart /g/. That the later manuscripts accord m3 the unhistorical value NG indicates that its original value had been forgotten (the character never appears on classical oghams and only very rarely in practical use after that), and this would certainly have happened before the need for it became apparent in Welsh. There are no Welsh ogham inscriptions as late as this in any case. Names such as Uurad (Irish Ferad) and Uurguist (Irish Forms) imply Pictish retained original /w/ [Jackson 1955:163]. Thus the transliteration of fern in Scotland is problematic. In a Gaelic context /w/ is appropriate up to about the seventh century and thereafter HI. In a Pictish context the value /w/ should perhaps be preferred regardless of date. The transliteration V is retained here, however, to preclude prejudging the language of a text and, where the text is supposed to be Pictish, to maintain a distinction between ogham V = /w/ = uu and ogham U = /u/ = u.
TRANSCRIBING AN OGHAM INSCRIPTION
'few undertakings afford so great a scope for human fallibility as the transcription of Ogham legends' [Macalister 1897:2] It is all very well to lay out a system for transliterating ogham characters into roman letters, but how does one apply it to an inscription which is damaged and has letters in doubt ? A developed system of conventions has evolved in Insular roman alphabet epigraphy for presenting readings of damaged inscriptions which express degrees of legibility and nuances of doubt [see for instance Okasha 1993:59]. It is not, however, appropriate for transcribing ogham. In the transcription system adopted here, capitals
INTRODUCTION
xli
are used for letters which are reasonably secure, lower case for letters over which there is an element of doubt. The problem with attempting to go beyond this is that often, because of the nature of the script, there are just too many possible alternatives. Each roman letter is a distinctive shape such that even if only a small portion survives it is often possible to identify the character in question (as often with S or B or A), or to narrow it down sufficiently that context makes it clear (Q/O, E/F, etc.). Ogham letters, in contrast, are mostly slight variations on the same basic shape. If an ogham letter is damaged, even if a single stroke is missing, it becomes an entirely different letter. Furthermore, unlike the roman alphabet, it is not always obvious that a damaged ogham character is incomplete. Similarly, if a fragmentary ogham is incorrectly oriented a character read upside down becomes an entirely different character, and, unless b-and h-strokes slope, it can be impossible to recognize that an inscription is being held the wrong way up.
Another common difficulty is assessing the grouping of strokes. In a loosely spaced row of strokes it is not always clear where letter boundaries lie, thus a group of four h-strokes might be a single group - C or two pairs - DD. For the same reason it can be difficult to judge how many letters are missing in a damaged section. Even if it is possible to assess the number of strokes, which for shorter stretches it often is, it is rarely obvious if a ten-stroke-space held two five stroke letters, or a four and a five well spaced, or three threes and a one, etcetc. If an inscribed roman L is doubtful it is sufficient to indicate this by italicizing it since readers are sufficiently familiar with the roman letter forms to know that possible alternatives to bear in mind are I and T (or whatever). Since few readers are intimately familiar with the ogham alphabet it is not sufficient to italicize an R in the expectation that they will know enough to bear I in mind as a possible alternative.
In order to spell out the feasible alternatives to doubtful letters, and to do so clearly, Padel adopted and developed the system of presenting variant readings vertically, in decreasing order of probability, the most
INTRODUCTION
xlii
probable at the top. This system has the advantage of being relatively uncluttered, easy to read, and clearly highlights the most likely reading. Its drawback, in addition to taking up perhaps several lines of the page, is that it is extremely awkward to type-set accurately. Though possible on a conventional type-writer, it has proved too difficult to replicate using a word-processor. Instead a linear system has been adopted, which, though containing exactly the same information is, unfortunately, not so easy to read. In this system the variant readings are set out linearly between round brackets, divided by oblique strokes in decreasing order of probability, e.g. (a/h/m) means 'probably A, possibly H, or even M \ A blank means 'no letter* not a space, e.g. (/a) 'nothing, or just possibly an A*. A single unidentified letter is indicated with a point, e.g. (r/. v) 'R or some-letter-followed-by-V\ It is sometimes possible to identify the aicme to which a character belongs but not its position in it In such cases, this is expressed as, for instance b+, meaning 'B, or any subsequent member of the b-aicme, i.e. BLVSN', or if, say, at least three strokes are visible, v + , meaning 'V, or any subsequent member of the b-aicme, i.e. V, S, or N'. The system is crude and not altogether satisfactory. It cannot reflect with total accuracy the relative likelihood of the full range of probabilities, but the transliterations are already at the limit of legibility. As a means of condensing a long-hand description of each stroke into a short-hand reading it is serviceable. The reader who needs to know more should, in any case, refer to the stroke-by-stroke descriptions [see Guide to Entries for further information on conventions used in transcription].
THE PALAEOGRAPHY OF OGHAM
'Unlike its Latin counterpart the Ogam script has no datable or chronologically significant palaeography* [McManus 1991:78]
While the forfeda have always been recognized as a later addition to the original inventory of twenty letters, the possibility that the ogham script might have a palaeography of its own, with all the
THE PALAEOGRAPHY OF OGHAM
EXPLANATORY DIAGRAMS
STROKES
SLOPE
None
Slight
Sharp
SLOPE Backward (Forward)
Forward (Backward)
Forward (Backward)
SPACING
All strokes evenly spaced
Some spacing of letters
Spacing equivalent to one missing stroke
PROFILE Average
Elongated
Squat
FORFEDA x-forfid
(variants: > <
> <)
Q~forfid
'flapped' character
angled vowels
A
'feather-mark'
word-dividing dots
hammer-head A
0
U
NON-PHONETIC MARKERS bind-strokes
etc.
INTRODUCTION
xliii
opportunities for dating which that might imply, has been consistently denied. While it is true that 'the rectilinear scores of Ogam do not offer much scope for variety' [McManus 1991:78J, a close inspection reveals subtle variations in such factors as slope, spacing, and relative length of stroke. I am not talking here of differences in carving technique, the execution of inscriptions is a separate issue, but rather of distinctions in the ideal form of the script regardless of medium or scale.
The variations among the Scottish ogham scripts are marked, partly because of the great diversity of contexts of the inscriptions, but also because of the extended period over which the script was in use in Scotland (sixth to tenth century). The variation among the orthodox inscriptions, with their narrower date range (predominantly fifth century and first half of sixth [McManus 1991:97]), is considerably less. To Brash the oghams of Ireland and Wales appeared 'as if they are the work of one hand' [Brash, in Trail 1890:350], but close inspection reveals a number of the same variations as in Scotland, if to a lesser degree. Since the possibility of a palaeography of Irish ogham was always denied, no close attention has been paid to the precise form of script used in Ireland. Even so, it was noticed, for instance, that a small group of oghams in Co. Cork were written in a distinctive script with elongated letters comprising slender strokes tightly-packed ['minute scratchy scores' Macalister 1907]. From casual inspection of several dozen orthodox Irish oghams, my impression is that detailed study of the form of scripts used would reveal other patterns of difference, more subtle than the similar if more dramatic variations easily detected in Scotland. Unfortunately, Macalister was not sensitive to possible variations in slope and spacing, and his diagrams often appear to have been 'tidied up' [for the minimizing of letter-slope contrast his diagram of the ogham stone from Bridell, Pembrokeshire (CIIC 426) with Thomas's photograph, 1994 fig.6.4; for minimizing vowel-stroke length compare his drawing of Eglwys Cymmin, Carmarthenshire (CIIC 362) with NashWilliams 1950 plate III]. Nonetheless, his photographs reveal, for instance, occasional examples of sloping b- and h-strokes (Castletimon, Arklow, Co. Wicklow CIIC 47; Colbinstown V, Co. Kildare CIIC 23). As I argue below, in Scotland the typology has clear dating implications. In Ireland the variations may be more on the level of 'hand', i.e. the same ideal form realized slightly differently by different oghamists.
INTRODUCTION
xliv
Such patterns may nonetheless be useful for recognizing the work of a single oghamist, or for identifying regional schools (as appears to be the implication of the elongated Cork letter-style), even if any dating significance eludes us.
Leaving the forfeda to one side, there appear to be diree areas of variation in ordinary letter strokes: spacing, relative length, and slope. The majority of orthodox (i.e. pre-sevendi century) oghams have all their strokes equidistantly placed, irrespective of letter or word-boundaries. Some, however, leave a slightly wider gap between the last stroke of one letter and the first stroke of the next, especially if they are bodi members of the same aicme and thus liable to be conflated. The size of this gap can vary from only a little more than the average space between the component strokes of a single letter, to a gap as large as if an intermediate stroke has been omitted (i.e. twice the width of the average space between component strokes of a single letter), rarely, die gap may be even larger. This increasing grouping of letter-strokes reaches its apotheosis in die bound-letters of the latest oghams. It does not appear, however, as if extra spacing was ever employed to indicate word-division, though from at least die ninth century a number of other devices ('feather-marks', single dots, dot pairs) were used in both Scotland and Ireland to indicate word-boundaries (see below).
It appears that, as originally conceived, the fundamental distinction in the ogham alphabet was between long consonant strokes and short vowel notches. Modern typographical representations of the alphabet, such as those of Thurneysen [ 1946:10] and McManus [ 1991:2 (top of page)], have fostered the erroneous impression that vowels are made of up circular dots. In fact, even in the earliest examples, diey generally consist of linear strokes formed in the same way as the consonants, diough of considerably shorter length. Close inspection of the orthodox oghams reveals a degree of variation in the lengdi of vowel strokes and these can be as short as an eighth of die ogham-band, or as long as a quarter, or even a tliird of it. Further study is required before it can be judged whether this is due to scribal whim, geology (shorter strokes being
INTRODUCTION
xlv
easier on sharper arrises), or does in fact reflect a chronological progression. It underlines, however, that the base fonn of vowel-letter components was conceived of as linear not circular.
It may have been the influence of manuscript oghams which promoted the use of longer strokes for vowels, but, to the extent to which dots are more easily lost or overlooked, strokes constitute an improvement. From the seventh century longer-stroke vowels were the norm and from the ninth vowel-strokes habitually occupy the full width of the ogham-band (e.g. Ballyspellan CIIC 27, St. Gall Priscian). While vowelstrokes were very short, it is possible to have m-strokes which are more or less perpendicular to the stem, without confusion. This is found on many Irish inscriptions, and also in Scotland (Bac Mhic Connain). If vowels are to be written with longer strokes then it is necessary to find some other way of differentiating the two aicmi. In practice, this was done by having vowel-strokes which were perpendicular, and mstrokes which were oblique across the stem. Since so many orthodox inscriptions have oblique m-strokes and very short vowels, it is clear that the introduction of slope was not a response to lengthening vowel strokes, it may, in fact, have opened the way for it. Further study of the earliest examples is required before it can be ascertained if perpendicular m-strokes are the wr-fbnn and oblique ones an innovation, or whether the m-aicme was conceived of as sloping from the outset and perpendicular ones were merely a scribal variant.
Over time, there was a move from a form of the script which had long-stroked b-, h-, and m-consonants on the one hand and short-stroked vowels on the other (die typical fonn in die orthodox Irish inscriptions) to a fonn in which the primary distinction was between perpendicular b-, h-, and vowel-strokes, against oblique m-strokes (as appears to have become standard in post-seventh century Ireland). In some Scottish oghams, however, the primary vowel/consonant opposition was maintained through the introduction of slope to the b- and h-aicmi. The new contrast was based not on length but on gradient, i.e. not short vowel-strokes : longer consonant-strokes, but perpendicular vowel-strokes : oblique consonant-strokes.
INTRODUCTION
xlvi
To the limited extent to which it is possible to establish a relative chronology of Scottish ogham inscriptions, it may be seen that early inscriptions have perpendicular b- and h-strokes and late ones have all consonants consistently well-sloped. There is an intermediate group of inscriptions with b- and fastrokes sloping a little (Brandsbutt), or an apparently random mixture of sloping and perpendicular b- or h-aicme letters (Latheron), though component strokes tend to be parallel. In some cases slope is employed sparingly to differentiate successive letters, or where space is short (Scoonie). That oghams with short vowels (e.g. Gumess) tend not to have sloping b- and h-strokes appears to confirm that, just as increasing vowel-stroke length has chronological implications, degree and consistency of slope increases over time.
Post-seventh century Irish examples of ogham in 'practical* use, for instance the Killaloe and 'Colman bocht' slabs (CIIC 54 and 749), the Kilgulbin bowl (CIIC 1086), the Ballyspellan brooch (CIIC 27), and the ogham marginalia, exhibit a style of script in which vowel-strokes are full-length but b- and faconsonants are perpendicular. Though a comprehensive study has yet to be made, impressionistically, it appears that in this later period sloping b- and h-consonants are restricted to Scotland. This is potentially important evidence of a distinctively Scottish strand in the ogham tradition.
If slope on b- and fa-
consonants was an inherent tendency in the script, it appears to have been given free reign in Scotland and ultimately systematized in the canonical form of the script. In Ireland, by contrast, it appears never to have attained canonical status and in fact, to have been suppressed at least by the ninth century. Most Scottish oghams with long vowel-strokes have sloping b- and h-consonants, but the few which do not, for instance Blackwaterfoot 1, might, in fact, reflect Irish influence. Clearly further work is needed before slope of b- and h-stroke can be taken as diagnostic of date or of strength of Irish influence, but the phenomenon holds out the prospect that such analysis is possible. For now, the consistency of the direction of slope in Scottish oghams (h-strokes always backwards, b-strokes always forwards) is of the greatest assistance in determining the direction of reading of fragmentary texts. Padel maintained that the consistency of direction in sloping Scottish ogham was 'accepted as dogma, rather than proved' [1972:19 n.l], but, his
INTRODUCTION
xlvii
main exception, Buckquoy, has been shown to read in the direction opposite to what he thought, and thus, in fact, to confirm the standard direction of slope [see entry below]. Padel's sole remaining exception, Abernethy, is also thrown in doubt [seen entry] and the very consistency of all the others, in their great diversity of form and context, may be taken as proving the issue beyond reasonable doubt.
The lengthening of vowel-strokes, the increasing differentiation of the aicmi by slope, and the increasing grouping of strokes by spacing and ultimately by bind-strokes, represent a steady improvement in legibility. This reflects, on a systemic level, the concern for clarity seen on the level of individual inscriptions, e.g. successive letters of the same aicme given additional spacing (passim), or sloped at slightly different angles (Newton et al.); the initial stroke of a group protruding slightly (as at Golspie); a little dot added after carving to separate strokes carved too close together (as at Lunnasting 16).
All previous scholars of ogham in Scotland have drawn a fundamental distinction between 'Irish-looking' inscriptions, with short vowel-notches (e.g. Gigha, Poltalloch, Newton, Auquhollie) and so-called 'Pictish* oghams, with long vowel-strokes (such vowels even being described in some quarters as 'Pictish vowels* [e.g. Padel 1972:19]). However, an unbiased look at the entire corpus of ogham on both sides of the Irish Sea reveals geography or ethnicity to be irrelevant: vowel-stroke length is purely a function of date. Since the Scottish examples are predominantly later than the Welsh or Irish, then in Scotland, naturally, longervowels predominate, but inscriptions which are early exhibit vowel-notches. Conversely, in Ireland, the bulk of the inscriptions have short vowel-strokes, as one would expect of such early examples, but from the seventh century onwards both epigraphic and manuscript oghams have long-stroke vowels.
Another supposedly 'Picrish* feature is the presence of a written stem. The form of the monument on which survive the overwhelming majority of our extant oghams provides a convenient arris to perform the role of base-line. In such cases we would not expect a written stem, but even when local geology yields
INTRODUCTION
xlviii
smooth arris-less boulders, early oghams were written across the smooth face with no stem (e.g. Ballintaggart CIIC 155-63, Buraham 175, Great Saltee Island 43 and numerous other examples). There are over half-a-dozen Scottish examples of stem-less oghams, including the arris-written Auquhollie, Poltalloch, Gigha, Dunadd, and Bac Mhic Connain, and a single example of a stem-less ogham across a flat face, Inchyra 3. That stem-written oghams predominate in Scotland is, again, a reflection simply of die predominantly later date of the Scottish examples. At the very end of the classical ogham period a handful of Irish oghams are written on a stem, for instance, Maumanorig and Aultagh (CIIC 193 and 72), and by the eighth and ninth century written-stems are standard in Irish ogham, just as they are in Scotland (cf. Kilgulbin, Ballyspellan, St. Gall, etc.).
The presence or absence of a stem seems like a fundamental distinction, but how important was it ? Since the base-line was always theoretically present, how big a conceptual step is it to write it in ? The Newton stone may imply 'not very'. The majority of its letters are written on the arris, but the inscription continues across the face for a short section along a stem. Though it remains a possibility that this continuation is a later addition, the letters of the two parts are identical but for the stem. The script of Pool is not at all manuscript-like and the nature of the slab on which it is written may be the sole explanation needed for the inclusion of a stem. In Ireland, no stem-written oghams are very early, and no stem-less ones are late, thus the presence or not of a stem provides a chronological horizon of sorts. In Scotland, however, the picture is more complex. Several of the stemless oghams must be later than Pool which has a stem but dates from as early as the sixth century. Inchyra has one stem-less ogham across a flat face (C), but this is almost certainly later than the two other oghams on this slab, both of which are written with a stem and neither of which can be much earlier than the mid-seventh century, if that. That the use of a stem can be a somewhat arbitrary indicator is shown by Golspie where an otherwise rather manuscript-looking ogham is written without a stem on the arris of a monument which can be no earlier than the eighth century and may be as late as the ninth.
INTRODUCTION
xlix
THE TERM •SCHOLASTIC 'While scholastic oghams use the same alphabet, with slight variations, and have their own characteristics they would appear to be a degeneration, and have no such serious purpose as did the classic ogham stones. They would seem to be displays of learning.1 [Bums 1981:40] Over and above the entirely different physical contexts in which the two types survive, the general appearance of pre-seventh century ogham script is radically different from that of post-eighth as a result of a cluster of features identified by McManus as the presence of a written stem, longer vowel-strokes, forfeda, the small 'feather-mark' word/direction indicator (->}, and new orthographical conventions (such as the use of H as equivalent of manuscript h) [1991:129]. It appears to have been Ferguson who coined the term 'scholastic* to refer to these post-eighth century oghams [Kermode 1907:72], It has been universally adopted and applied, usually disparagingly, to later oghams in both Ireland and Scotland. 'Scholastic', however, is, as McManus has complained, 'a potentially misleading term in that it might be understood to imply that the orthodox inscriptions were not the product of a scholarly tradition, a view which ... is untenable' [1991:129]. Thanks to McManus himself [1986] and to the work of Harvey [1987a/b] and Stevenson [1989] there has been a revolution in the way in which 'classical* ogham is viewed. Gone is the 'childishly impractical' or even barbaric script invented by druids for ritual and cryptic purposes [see McManus 1991:7 for discussion of such views]; in its place a linguistically sophisticated alphabet derived in part from a knowledge of late Latin grammatical theory, eminently practical for the purpose of short epigraphic texts. Ogham was 'learned ' from the start (and is not, by definition, all writing ?). At the other end of the tradition, Sims-Williams's work on the forfeda required post-classical ogham to be taken more seriously (a goal to which it is hoped the present work will also contribute). For all that McManus, despite his reservations, continued to use the term 'scholastic1, it does seem preferable to devise a new way of distinguishing between the two broad phases of the tradition. The form and aspect of the later scripts is clearly derived from contemporary (roman alphabet) manuscript practice, suggesting 'book-hand' as a possible label. This, however, presents too much scope for confusion
INTRODUCTION
1
and doesn't solve the problem of what to call the earlier phase fepigraphic', what then of epigraphic examples of 'book-hand' ogham script ?). Perhaps better is the more neutral 'post-classical', in opposition to the orthodox, or 'classical', oghams of the pre-seventh century.
In Ireland, where virtually all oghams date either before the mid-seventh century or after the late eighth, there is little difficulty in categorizing oghams as either classical or post-classical. In Scotland, however, where the majority of extant oghams date to precisely this intermediate period, drawing a hard line is far from straightforward. The end of the Irish tradition of erecting ogham-inscribed pillars is a social and cultural question quite separate from the continuing development of the script. Since text depends on context, changes in verbal formulae are similarly epiphenomenal. The adoption of manuscript spelling is more telling, but only in Ireland was it simultaneous with palaeographical innovation. In Scotland it appears that, at least in some cases, a post-classical ('scholastic') script was used to write an orthography which was essentially a continuation of the old ogham spellings.
Orthography and script must be
considered separately, and for now we are concerned solely with script.
As described above, there was a long series of modifications to the original form of the ogham script, the first (the introduction of the x -forfid) occurring already in the classical period. Where then is the dividing line to be drawn ? The presence or absence of a written stem is usually taken as the primary diagnostic feature, but what of Golspie which lacks a stem but is in all other respects (spacing, aspect, forfeda) 'scholastic'. Or what of the very simple script of oghams, like Pool or Brandsbutt, which, were it not for the presence of the stem, would be indistinguishable from typical 'classical' examples; or 'stem-written' oghams which lack forfeda exhibited by some stem-less oghams. How meaningful is a line which divides Gurness from Dunadd, and brackets it together with Bressay ? We must face the fact that what we have is not a dichotomy but a continuum. The proliferation of forfeda, lengthening of vowel-strokes, sloping of b- and h-strokes, and spacing of letters and ultimately words, were multi-stage processes begun in the
INTRODUCTION
li
classical period and continued in the post-classical. Only in Ireland is there the illusion of a clean break. A number of the Scottish oghams fall in a grey area between 'classical' and 'post-classical* and it is a matter of subjective judgement which of these are derived from a manuscript-based aesthetic and which are not.
Having acknowledged that there are perhaps as many differences among different forms of post-classical ogham script as exist between post-classical and classical (as befits the greater lengdi of the 'post-classical* period), it would be helpful to be able to refer to these different formal varieties. In the catalogue F have recognized die gross difference between stemless and stem-written ogham by labelling them 'Type T and 'Type IF respectively. Type I may be further divided into Type la, which lacks any forfeda, and Type Ib, which exhibit the first fotfid only. Type II is divided into three classes: Type Ila, which is simply Type la written on a stem (i.e. lacking forfeda); Type lib with commonly occurring forfeda (including angled vowels); and Type lie which has one or more of die following features: bind-strokes, word-division, unique or me forfeda. This is a provisional classification for descriptive purposes only. It is not without dating significance - Type lie is exclusively late, all the earliest oghams are Type la - but is not intended to imply a chronological progression. The possibly sixth century date of the earliest datable Type Ila, Pool, is considerably earlier than the latest Type I, the eighth of ninth century Golspie. Simple forms undoubtedly continued in use after the introduction of more developed varieties, especially in informal contexts. Since forfeda indicate attention to orthographical detail die lack of them in non-formal or nonclerical contexts may reflect simply die less learned environment (similarly with geminate consonant symbols which are more common in formal texts and those from an ecclesiastical milieu).
The development of the script from the orthodox inscriptions of die fifth and sixth centuries through the seventh and eighdi, to the ninth and tenth, is paralleled also in a shift in roman alphabet Insular inscriptions from the use of capitals, arranged across the slab with little or not attention to word-boundaries, to book-
INTRODUCTION
1H
hand laid out like text in a contemporary manuscript. As with the roman alphabet inscriptions, die trend is from a distinctively epigraphic towards an ever-more manuscript-derived attitude to letter form, letter spacing, scale of lettering and layout of text. This is as true in post-seventh century Ireland as it is in Scotland and a similar phenomenon is traceable in the development of the runic script in contemporary Anglo-Saxon England.
In conclusion, there is evidence that the ogham script was in practical use for seven or eight hundred years, from as early as the fourth century into the eleventh (from the twelfth century there are texts about ogham, but not texts in ogham). The vagaries of archaeological preservation mean, however, that the surviving examples are very unevenly distributed over time. The vast majority date from before the seventh century, and of these, most span a period of just 100-150 years. This has everything to do with shifting power relations in immediately pre-Christian Munster, much less to do with the development of the ogham script. Previous attempts to distinguish between Irish and Scottish ogham have focussed on features which are purely a function of date and have nothing to do with geography or ethnicity (length of vowel-stroke, presence or absence of stem, other 'developed features' including forfeda, word-division, etc.). Ogham inscriptions in Scotland are little different to Irish oghams, of a similar date. Only onefeaturehas been identified as possibly distinctively 'Scottish', namely the systematization of slope for b- and h-aicme consonants, but further work is required to evaluate the role of gradient in Irish oghams before even this can be stated categorically.
THE FORFEDA
'Spelling reformers who have attempted to add new letters to old alphabets have received a mixed reception' [Sims-Williams 1992:29] Until recently, scholars have been dismissive of the miscellaneous assortment of characters invented to
INTRODUCTION
liii
supplement the original ogham inventory of twenty characters. In the light of later medieval traditions it was thought that these additional letters (for-feda) 'were designed with the Latin and Greek alphabets in mind, in particular to accommodate letters of the Latin and Greek alphabets not already matched by Ogam characters' [McManus 1991:2]. Their creators were thus 'out of touch with the objectives of the original framers of Ogam who were concerned to create an alphabet for the Irish language1 [ibid]. However, a comprehensive reassessment of theforfeda by Sims-Williams has shown convincingly that, on the contrary, the 'driving force behind their development was the desire to enhance the ogam alphabet as a means of representing the sounds of Irish itself [1992:75]. The fantasy alphabets of the Book of Ballymote are indeed scholar's whimsy, but the serious intent of the forfeda found in practical use is clear. Taking each in turn, Sims-Williams has shown how it was created and developed to fill a perceived gap in the inventory of symbols available to represent the Irish language, and how the additional letters aided Irish grammarians 'in die taxing task of conceptualizing the sounds of the evolving Irish language in terms of written graphs' [1992:75].
It has been customary to compare unfavourably the 'more complex visual shapes' of theforfeda 'smelling strongly of the ink-horn rather than the stone-cutter's blade' [Hamp 1954:312 cited by McManus 1991:2], to the simple lines of the original letters. As Sims-Williams has emphasized, however, some additional characters, for instance the double or cross-hatched letters, 'use the resources of the original alphabet', and even those which are modelled on the shapes of foreign letters 'were not entirely unsuited to stone or wood' [loc.cit.]. That the letters are of miscellaneous shape is not surprising given the ad hoc nature of their creation, for theforfeda were not created as a single system and introduced by fiat, but rather evolved gradually over a long period of time. The later manuscript treatises on ogham tend to give only five supplementary letters, a codification of an Aicmena Forfed clearly modelled on the quinary groups of the original twenty characters. Examination of (be forfeda in practical use, however, reveals that there were many more than five.
INTRODUCTION
liv
The first forfid was introduced when it was acknowledged that the gap between the ever-evolving language and the graphemes available to represent it had become intolerably great Once it had been accepted that the interests of orthographical precision might over-ride the elegance of the original system, and thus that grammarians were free to modify the alphabet in line with their changing perception of Irish phonology and orthography, it was perhaps inevitable that there would be an accelerating process of 'plugging the gaps'. The impact of the new (Christian) manuscript-based orthography must have been profound and may have prompted the first major re-think of the ogham letters and their values. The introduction of new symbols was only pan of a wider process of orthographical reform which saw the assignment of new values to the redundant old letters H, Gw, and Sl. The reforms, however, were piece-meal, or at least occurred in several stages over a long period. Many of the innovations appear to have achieved canonical status, but the occurrence of unique symbols on certain inscriptions suggests that not all were universally adopted (for instance, Inchyra A (i) 77, Bressay 19120, Lunnasting 4). Discrepancies between the manuscript tracts suggest competing traditions [Sims-Williams 1992:32-5]. What is clear, is that post-classical oghamists in Ireland and Scotland drew on a common pool of innovation and experimentation.
Many of the
developed traits present in Scottish oghams had already appeared in Ireland, or were later to feature there. The first forfid (& -*-, •>-<-) is well-attested in Ireland inscriptions (and in 'practical' manuscript use), and even the much rarer 'second forfid\ (&•&} is epigraphically attested (Maumanorig CIIC 193). Both have canonical status in the later manuscript tracts. The principal of cross-hatching, later extended to various characters, is established in Ireland by the end of the classical phase (Ballyhank CIIC 102) and appears in practical use in Irish manuscripts of the ninth century (St. Gall). Macalister was wrong to say that the forfeda were evolved by Picts to represent sounds not present in Irish [1940:186], since it is clear that many of them first appear in Ireland. This is not, however, to rule out the possibility that new features were invented in Scotland, or the values of existing ones modified. Angled vowels (which may draw on the form of the non-phonetic 'feather-mark') were widely used in Scotland, but appear in practical use only in a much later Irish manuscript (see entry on Lunnasting for further discussion). The innovation of bound
INTRODUCTION
lv
strokes first seen in Scotland occurs also in an inscription from the Isle of Man and in later Irish manuscript tracts (Book of Ballymote). Even the unique 'integration sign' character (retrograde
fifif)
seen at Bressay has its parallel in In Lebor Ogaim. The lack of extant Irish evidence for the key, early post-classical, period, makes it difficult to discern whether the borrowing was all one way (from Ireland to Scotland), or whether what we have, in fact, is only one side of a conversation whereby innovations generated on both sides of the Irish sea were passed back and forth. Since some at least of the Scottish oghamists were grappling with the problem of using the ogham alphabet to represent a non-Goidelic language, it would be thought that they, more than most, would have given close attention to weaknesses in the graphemic arsenal. Their handling of H, for instance, suggests that they did.
The use of forfeda has, in the past, been taken as one of the primary contrasts between ogham in Ireland and Scotland (or Pictland) since, in Padel's words, 'Pictish' oghams are 'far more heavily endowed with them' than Irish [1972:1]. As long as forfeda were considered baroque and degenerate the popularity of forfeda in Scotland was considered an embarrassment. The new dignity accorded them by Sims-Williams entails that their proliferation in Scotland need no longer be dismissed as merely a fondness for 'showing off'. Instead it implies oghamists possessed of the same 'keen instinct for phonetics and a refreshing independence of mind' [McManus 1991:25] attributed to their Irish counterparts. It would be wrong, however, to characterize Scottish ogham (or even oghams solely from the area of Pictland) as 'stuffed' with forfeda.
Of course, if all classical ogham in Scotland is at the outset excluded as Irish-looking' then,
naturally, the 'Pictish' ogham corpus will have more forfeda than the Irish since the use of forfeda was more prevalent in the post-classical period. Of the thirty-five or so Scottish oghams sufficiently legible to discern, fifteen or sixteen have no forfeda at all; seven or eight have either a single x-forfid or at most one further developed character (e.g. hammer-head A, angled vowel); only six have more than two. Even in this last group the use of supplementary characters is scarcely otiose. Golspie has about twenty-seven characters, only three of which are forfeda and these of the most widely attested son: an x-forfid, an
INTRODUCTION
Ivi
angled vowel, and a hammer-head A. The same three occur at Burrian, accompanied by a single crosshatched R, a character with well-established Irish parallels. Nonetheless, even if one is strictly comparing only like with like, it is true ihaiforfeda are rather more common in post-classical ogham in Scotland, than in Irish oghams of the same period. Why might this be ?
For a start, the Irish sample of post-Classical ogham in practical use (i.e excluding manuscript tracts about ogham), is very, very small and the texts uniformly shon. Thus it may be entirely fonuitous that specific characters are not attested in practical use when later manuscript tracts testify to Irish knowledge of them. Secondly, the demandforforfeda is, of course, greater die further one is from the Primitive Irish of the period of the original inventory, whether simply by date (i.e. development within the Irish language) or by language family (as explained above, die incompatibilities are greater if trying to represent non-Goidelic names or words). Brittonic-speaking oghamists may well have felt the inadequacies of the original inventory more keenly. Thirdly, it may have been that contemporary manuscript spelling had an inhibiting effect on the use of forfeda in Ireland. It has been noted that at least some Irish post-classical oghams are simply a caique on the contemporary manuscript spelling of the words, for instance, die use of h3 and hi in St. Gall's LATHEIRT, rather than the old ogham convention of using simplex T to show lenition. Similarly with the use of h4 and hi in St. Gall's M1NCHASC, Ballyspellan's CNAEMSxCH, and Killaloe's BxNDACHT and CxLLACH, though in the case of the last two, it is interesting to note the preference for the x-forfid over EA. The prominent early use of this firstforfid may have established its use beyond die learned environment of die grammarians. The contrast with Scotland is marked for here there appears to have been an effort to persevere with old ogham conventions, for instance the contrast between simplex and geminate consonant symbols. The letter H was not used as h was in die manuscripts to indicate lenition of a preceding T or C. It is open to discussion whedier this lack of manuscript influence on Scottish ogham onhography reflects the lack of a local tradition of manuscripts in die vernacular, or radier that Scottish ogham was developing outside die ecclesiastical milieux where such
INTRODUCTION
lvii
manuscripts might be expected.
Further progress in interpreting the texts of Scottish oghams is required before we can establish the extent to which the orthographies used in Scotland conformed to or diverged from accepted Irish practice. Unfortunately, the value ascribed to a character in a given instance depends on our interpretation of its context, and, of course, our interpretation of the text depends on the values ascribed to individual characters. As discussed above, it is possible that in Scotland new values, as yet invisible to us, were assigned to characters such as Q or V. More obvious is the rehabilitation of characters redundant since before the oldest classical pillars. The earliest evidence for the practical use of H comes from Scotland where, as stated above, its value is not that later assigned to it in Ireland. The only epigraphic example of m4 in practical use comes from Scotland at a date at least as early as the first manuscript appearance of the character (late eighth or ninth century Berne Codex), interestingly with probably the same value, /s7 (see entry on St. Ninian's Isle).
The abundance of geminate consonant symbols has long been a recognized feature of Scottish ogham orthography. Padel recognized that the geminate/simplex distinction was not random and sensed intuitively that geminate symbols were avoided in word-initial position [1972:29], Subsequent independent work by Harvey has shown statistically the avoidance of word-initial geminate symbols in classical ogham in Ireland, a feature he sees as modelled on a similar convention in written Latin. In classical ogham Harvey suggests geminate symbols were used to indicate unlenited consonants; simplex, lenited [1987], an explanation which would go some way towards explaining Scottish examples of the distinction. In Ireland, however, this usage does not appear to have survived into the post-classical period. Instead Sims-Williams and others have detected a convention by which gemination could be used to indicate a preceding diphthong [Sims-Williams 1992:70, McManus 1991:91, Harvey 1987:69-70].
INTRODUCTION
lviii
The correlation between geminate characters and developed features such asforfeda and word-division, suggests that in Scotland the distinction is a 'learned* symptom. The implication of this is that, conversely, knowledge of ogham may have existed in 'non-learned' circles, a possibility fiixther supponed by the domestic, or non-ecclesiastical, contexts of Scottish oghams lacking geminate symbols, e.g. Ackergill, Bac Mhic Connain, Gurness, Inchyra C, and Pool. Given the very long period over which ogham was in use in Scotland, the wide geographical area and diverse range of contexts represented, it is not surprising that we should find evidence of more than one orthographic standard. But there were standards: innovations such as angled vowels or the hammer-head A are observed across a range of inscriptions (angled vowels: Birsay 1, Bressay, Brodie, Burrian, Cunningsburgh 2 and 3, Formaston, Golspie, Lunnasting; hammer-head A: Birsay, Buckquoy, Burrian, Formaston, Golspie, Latheron, Lunnasting, possibly Scoonie. See Formaston entry for further discussion of both features). The present work appears to support SimsWilliams's contention that 'it may be only the relative scarcity of extant Irish and Pictish ogam inscriptions from the later ogam period that has given the impression that the forfeda were peculiarly scholastic in character' [1992:75]. Initial forays suggest that further study of the orthography of Scottish ogham would be rewarded, not only with a greater understanding of language and literacy in early medieval Scotland, but also of the development of ogham as a whole in the crucial early post-classical period.
A NOTE ON WORD-DIVISION The use of a pair of dots arranged vertically flanking the stem (—4—) to indicate word-division is a wellknown feature of both Scandinavian and Anglo-Saxon runic writing [illustrations in Page 1987]. It appears in Shetland on the Scandinavian runic inscription of Papil [Moar 1952], in Orkney on the Birsay 2 runestone, and elsewhere in Scotland at Inchmamock, Bute and Laws, Monifieth [ECMS:412-3; 280]. At Kilbar, Barra single and double dots are used [ECMS: 114-5], and at Thurso there are triple dots [RCAHMS 191 Ib: 123], a form well-attested in Scandinavia. While word-division is not indicated in every runic inscription in Scotland, it is very prevalent [Shetelig 1954b]. Another device is the medial saltire
INTRODUCTION
lix
found on Iona 69 [RCAHMS 1982:190]. [For all these, see also Barnes 1993].
Word-division in Scottish ogham inscriptions is shown only on inscriptions from the Northern Isles. The Birsay 2 example has only a single dot and is doubtful, and the others, Bressay, Cunningsburgh 3, and Lunnasting, are all from Shetland, and all use double dots. The northern distribution and the similarity to the runic device led previous commentators to identify the word-division marker in ogham as a borrowing from the runic tradition [a suggestion first made by Graves 1855]. This is quite possible, Lunnasting and Cunningsburgh are undated, but Bressay and Whiteness demonstrate that oghams continued to be carved into the Norse period, and the later oghams of the Isle of Man, and of Hiberno-Norse Dublin (unpublished inscribed comb) and Killaloe (CIIC 54) suggest that the Norse (?or 'Celto-Norse') in the British Isles took up ogham with enthusiasm. If it could be proved that the device was a borrowing from Norse runic tradition then, obviously, it would have important dating implications. Any inscription exhibiting the feature would have to date after the Norse settlement in the ninth century. Unfortunately, however, the device need not have been a Scandinavian borrowing, since it is present in Anglo-Saxon runic inscriptions of earlier date, for instance on the Thornhill cross, dated c.800 [Elliot 1959:fig.35]. Less consistently, single, double, and triple dots occur on the Franks casket, dated c.700 [Page 1972]. The device need not have been a borrowing at all, but may have arisen within the ogham tradition itself. The ninth century ogham marginalia in the St. Gall Priscian includes a single dot, above the stem, half-way out across the ogham-band, indicating word division in the phrase a chocart • inso [Nigra 1872:196a (reproduced Sims-Williams 1992:64 fig.8)]. The ninth century Ballyspellan brooch (CIIC 27) uses ->to indicate the beginning of each word, whether at the start of the text, or in the middle. If Irish oghamists were experimenting with word-division in the ninth-century then it is possible that Scottish oghamists may have come up with the double-dot device themselves, quite independently. The use of a single dot to separate letters is found in the fourteenth century Book of Ballymote text of In Lebor Ogaim, if this practice had become established already by the time of the Lunnasting text, then the double dot as an
INTRODUCTION
lx
intensified version, separating not just letters but words, may have been arrived at without external input. Norse runic influence cannot ^ ruled out, but neither is it the only possible explanation. Thus, while word-division is certainly a late feature, it need not be a reliable indicator of a date after the first Norse cultural contact. There are no Irish examples of word-division earlier than the mid-ninth century, but then there are very few oghams from the later seventh and eighth centuries, so the practice could have originated earlier but evidence simply not survived. The concentration of examples in Shetland may simply reflect a later date for these inscriptions still within the Pictish period.
OGHAM IN SCOTLAND As stated in the Preface, the present intent is to provide the materials from which a history of ogham in Scotland may be written, rather than to write that history. Nonetheless, a few general observations are in order.
THE STUDY OF SCOTTISH OGHAM 'The so-called Pictish inscriptions constitute a problem so thorny that many writers who have discussed the Pictish question, however, confident they may have been, have carefully avoided attempting to deal with them. Perhaps this was discreet, but it is obviously the duty of anyone investigating the Pictish question not to play safe by ignoring entirely a very material piece of evidence1 [Jackson 1955:38]
The history of modern scholarship on ogham has yet to be written, and here is not the place to attempt it. The ogham script first came to 'outside' scholarly attention at the end of the eighteenth century with the notorious Mount Callan forgery [de h6ir I983] t but not until the work of the Rev. Charles Graves in the mid-nineteenth century, was it placed on a solid foundation. In the 1850s increasing awareness of ogham in Ireland led to the recognition that horizontal marks on certain standing early medieval monuments in
INTRODUCTION
lxi
Scotland were, in fact, ogham inscriptions. The first to be recognized appears to have been Bressay, followed by Golspie, Logie Elphinstone and Newton. In 1870 the first ogham stone to be found in an archaeological investigation in Scotland was uncovered at Burrian. In the following decade chance finds in Shetland prompted local antiquaries to search further, and by the end of the decade half-a-dozen had been recovered there, though some were subsequently lost. When the stones from Scoonie and Fonnaston were discovered they were instantly recognized as bearing ogham.
Till this point, Scottish ogham had been studied solely by Irishmen (Graves, Brash, Ferguson), whose starting point was the corpus of orthodox pillars of Munster, and inevitably the extant Scottish examples seemed strange. The first publication to treat the Scottish oghams separately was The Ogham Inscriptions of Scotland', by James Carnegie, the ninth Earl of Southesk [1884]. Southesk, antiquary and poet, famous for his collection of ancient gems [DM?], tackled the eleven oghams then known in Scotland with down-toearth pragmatism, but he was labouring with a number of defective readings (e.g. Scoonie, Newton), and lacked sufficient philological knowledge to make much of the texts. His publications on the subject in the Proceedings of the Society of Antiquaries of Scotland did much to raise scholarly awareness of ogham in Scotland, at a time when early medieval sculpture of Scotland was attracting increasing attention. Shortly after, the oghams on the Brodie stone were recognized, and, at about the same time, those at Auquhollie.
Until this point, talk had always been of the oghams of Scotland, of 'Scotch* oghams, and, despite the fact that all the inscriptions known at that point came from Pictland, the 'Picts' had scarcely been mentioned. Interpretation of the texts had always been on the assumption that they were in Irish, or at least some form of Celtic (excepting possible Norse influence at Bressay). All that changed, however, in 1892 with the publication of Sir John Rhjs's 'The inscriptions and language of the Northern Picts* (PSAS), a long and detailed discussion of the inscriptions (ogham and roman) leading to the conclusion that they were written in a language allied to Basque. From then on, these were 'Pictish oghams'. The publication of Allen and
INTRODUCTION
lxii
Anderson's monumental Early Christian Monuments of Scotland in 1903 placed the known ogham inscriptions in the wider sculptural context, butforhis readings and interpretation Allen relied on Rhys.
In the period up until the Second World War, oghams continued to come to light, increasingly in the course of archaeological excavation (Bac Mhic Connain, Gurness, Birsay 1-2, Poltalloch). Francis Diack used oghams in an unsuccessful attempt to support his theory that the prehistoric language of Scotland had been Goidelic, thus that Pictish was ancestral to Modern Scottish Gaelic. A decisive moment came with R. A. S. Macalister's publication on The inscriptions and languages of the Picts' [1940] in further support of Rhjs's views on 'Bronze Age' non-Indo-European Picts. Macalister divided the corpus into 'Pictish' and 'Irish-looking', covering the former in his article, and the latter (Gigha and Poltalloch) in his Corpus Inscriptionem lnsularum Celricantm. The Pictish oghams were thus consigned to a ghetto where diey have languished ever since.
The first to subject the oghams of Scotland to modern linguistic investigation of any rigour was Prof. Kennedi Jackson, unfortunately at a time when the prevailing paradigm saw the Picts and their supposedly non-Indo-European characteristics as very much a 'Problem'. In the decades after the War, new oghams were excavated (Whiteness, Inchyra, Buckquoy) and further inscriptions were recognized on standing monuments (Altyre, Dunadd, Blackwaterfoot). In each instance Jackson's opinion was sought. He published short notes on the last four mentioned, but beyond this his writing on the topic was restricted to a scant two pages in an article on the Pictish language [1955], and a short encyclopedia entry [1983]. Jackson was firmly of the opinion that the Pictish oghams were written in a non-Indo-European language, and it is to be lamented that such an influential opinion was not argued through at length in print.
The first oghams found in Scotland were predominantly formal public monuments using highly developed forms of the script. Increasingly, however, it was informal inscriptions, even graffiti, written in simpler
lxiii
INTRODUCTION
scripts which were coming to light, especially as a result of archaeological excavation. The palaeographical gap between the Irish and the Scottish examples was closing. Over thirty years after Macalister's 1940 article, a fresh survey of the material was made, an Edinburgh MLitt. under the supervision of Prof. Jackson. Its author, Oliver Padel subjected the ogham inscriptions ofPictland to thorough and sensitive examination, producing the most accurate and sympathetic set of readings to date. He did not, however, question the underlying assumptions of his study which remained those of Jackson a quarter century earlier.
In the late 70s and early 80s early medieval archaeology in Scotland was making great strides, and old orthodoxies, especially those concerning the Picts, were being challenged.
In general, however,
archaeologists steered clear of the oghams, 'owing to the appalling linguistic problems they present' [Ritchie 1987:62]. The Irish historian Alfred Smyth in attempting to over-turn Jackson's argument for non-Indo-European Picts, attacked his presentation of the ogham evidence [1984:58]. A number of his criticisms are telling, though his were brief passing comments in a more general discussion.
At about this period, the Edinburgh social anthropologist Anthony Jackson, was putting forward a novel explanation of the Pictish symbol stones which, at least at first, was given some credence [1984]. Bound up with it was his eccentric explanation of Pictish ogham whose import, he maintained, was not linguistic, but calendrical. Without explaining why an alphabetic system should be adapted in this way, he maintained that the component strokes of letters could be grouped together in sevens and multiples of sevens and thus that the ogham inscriptions were monumental manifestations of the Pictish ritual calendar. As Ross Samson has pointed out to me, a script based on multiples of between one and five strokes will inevitably produce occasional groupings of seven, or multiples of seven. Jackson explained the infrequency with which these patterns 'worked-out' in his corpus of readings, which appear to be based on Padel and are not always accurate, was due to their fragmentary preservation. He justified his heavy reconstruction of 'defective' readings on the grounds of 'the relative ease with which it can be done' [1984:184]. His theory is
INTRODUCTION
lxiv
unworthy of serious consideration, its author, in any case, having abandoned it in favour of another numerological explanation [pers. comm].
A real methodological breakthrough was made by Anna Ritchie when she argued that, though the interpretation of Pictish oghams was 'highly controversial', 'fortunately for the archaeologist they can also be treated as artefacts on the same level as any other item of material culture' [1987:64, original emphasis]. Ritchie looked at ogham inscriptions as just one of a 'class of artefacts' to be interpreted as 'intrusive Irish influence in Pictland' [62]. This influence, she concluded was not conveyed via the Scotti but through the Irish church (see below). While pointing out the fundamental similarities between Irish and Pictish oghams, she suggested that differences might reflect a Pictish lack of understanding of the script they sought to adopt [65]. On the contrary, it is clear that the Picts, and other ogham-using inhabitants of Scotland, were adepts, but in all other respects Ritchie's anicle, with its emphasis on the oghams of Scotland as archaeological artefacts first and foremost, has been the stepping off point for the present work.
THE CORPUS 'the object of this comparison ... has been to show that Pictish ogams are not as odd or inexplicable as they may seem' [Ritchie 1987:66] There are thirty-seven separate ogham inscriptions extant in Scotland (a further three are documented but lost), on thirty-four objects, from thirty sites. In the west ogham is found as far south as the island of Arran in the Firth of Clyde, or even further if Lochnaw is genuine, and as far north as North Uist. In the east examples come from the northern shore of the Forth (the southern boundary of Pictland), to the Northern Isles of Orkney and Shetland. Ogham is thus found both throughout Pictland and in the territory of the Gaels of D£l Rfada. Unless I mean to refer specifically to Pictish-speakers or dieir territory, I have tried wherever possible to avoid the terms 'Pictland' and 'Pictish' preferring instead die, for our purposes,
Cochgpilhead Poltalloch Dunadd
Gigha 1-2'
Blackwaterfoot 1-2
THE OGHAM INSCRIPTIONS OF DÁL RIADA
INTRODUCTION
lxv
ethnically-neutral term 'Scottish'. As argued above, it is impossible to identify an ogham inscription as 'Pictish' on the basis of its script. The geographical extent of Picdand varied over time, thus to apply the label 'Pictish' requires a greater degree of chronological certainty than is possible in many instances. In any case, how meaningful is the label 'Pictish' for an ogham monument from Norse Shetland, orforan ogham from Pictish territory written in the Irish language, as I maintain Buckquoy may be ? Surely 'Pictish' is meaningful only if this is the language of the text, and the language of the text can be established only rarely and with difficulty.
Unlike pre-seventh century Ireland (and southern Britain), where ogham stones constitute a single homogeneous class of monument, the ogham-inscribed objects of Scotland have little in common with each other beyond their lettering. Almost a third of the extant total are not even 'monuments' at all. Three of the inscriptions are carved on small non-prestige domestic objects (Padel's 'chattel oghams'), namely the Bac Mhic Connain knife-handle, the Buckquoy whorl, and the Gurness knife. A further seven or eight are carved on stone, but so informally as to be classed almost as 'graffiti', certainly, they are not in any sense 'monumental': Birsay 1-3, Blackwaterfoot 1-2, Inchyra A, and Pool. It may be appropriate to class Burrian with this group also. The remainder are formally carved on public sculpture, but from a wide variety of monumental genres. There are unshaped pillars (Auquhollie, Gigha, Newton, Poltalloch), Class I Pictish symbol stones (Ackergill, Brandsbutt, Inchyra, Logie Elphinstone), Class II Pictish symbolinscribed cross-slabs (Brodie, Formaston, Golspie, Latheron, Scoonie), Class III cross-slabs and crosses (Abernethy, Altyre, Bressay, Burrian, Cunningsburgh 2, Dupplin, Lunnasting, Whiteness), and a heterogeneous collection of miscellaneous inscribed slabs or fragments (Cunningsburgh 1 and 3, Lochgoilhead, St. Ninian's Isle 1). Dunadd is a formal monumental inscription carved on the living rock at a site of great symbolic importance.
The earliest datable inscription is Pool which may date to the sixth century, Birsay 3 is from die seventh century, others, including Auquhollie, Gigha, and Poltalloch, are probably at least as early as these two,
THE OGHAM INSCRIPTIONS OF ORKNEY
INTRODUCTION
lxvi
probably earlier. At the other end of the spectrum, the Whiteness cross-slab fragment is dated on an historical grounds to the tenth century, Bressay is ninth or early tenth, and it is possible that other Shetland oghams, from Lunnasting and some or all from Cunningsburgh, also post-date Norse cultural contact. The others range in date between these two extremes, with many dating probably to the eighth and ninth centuries.
Ogham is found at major ecclesiastical sites, including Dupplin, and St. Ninian's Isle, and also at churchsites of more humble status, e.g. Bressay, Brodie, Formaston, and Golspie. What is striking however, is the appearance of ogham at secular sites. Secular households of high, possibly royal, status, might be expected to have included ecclesiastics, which may go some way to explaining the use of ogham at Dunadd and Birsay, but this seems less likely of humbler domestic sites such as Gurness and Bac Mhic Connain. Ogham graffiti in a cave at Blackwaterfoot further supports this view of demotic literacy in ogham. At least three or four of the extant ogham monuments of Scotland were found in association with burials, isolated or in cemetries: Ackergill, Inchyra, Poltalloch, and possibly Newton. An interesting group in the north-east comprises re-used prehistoric standing stones: Auquhollie, Brandsbutt, Logie Elphinstone. This great diversity of contexts, over a wide area and long period of time, suggests, as Padel put it, that 'knowledge of Oghams ... was fairly widespread at some point' [1972:3]. What is certain is that it was not restricted to a learned ecclesiastical milieu.
As argued above, the primary difference between the ogham monuments of Ireland and Scotland is one of date. If monumental oghams had been carved in Ireland in the eighth, ninth and tenth centuries, they would undoubtedly have looked something like the extant examples from Scotland (for confirmation of this, see the Killaloe and Colman bocht slabs). That they were not is a reflection of a fundamentally different attitude to the script in the two regions. In Ireland up until the seventh century ogham was the sole monumental script, thereafter new monument forms were introduced which employed exclusively the roman alphabet. The subsequent development of the ogham script indicates that knowledge of ogham did not die
THE OGHAM INSCRIPTIONS OF THE EASTERN MAINLAND (SOUTH)
lxvii
INTRODUCTION
out, but never again did it enjoy centre stage. The Celtic inhabitants of Scotland had three writing systems at their disposal: die toman alphabet, die ogham alphabet, and die Pictish symbol system [Forsyth 1996b]. The Pictish preference for die latter two reflects a distinctive attitude to literacy which I have discussed elsewhere [Forsydi 1996d].
Widi regard to its ogham inscriptions, Scotland may be divided into half-a-dozen regions: Argyll, Atlantic coast (including Orkney), die mainland south of die Mounth, the nordi-east mainland, die northern mainland, and Shetland. Very briefly, taking each in turn:
Only two inscriptions in Argyll have a strong claim to be early (Gigha and Poltalloch), one is certainly late (Lochgoilhead), and die others, though simple and thus difficult to date, are unlikely to pre-date die seventh century and could be later.
Gigha, Poltalloch, and Dunadd are probably all Goidelic,
Blackwaterfoot and Lochgoilhead could be, but have not be satisfactorily interpreted. All were recognized on standing monuments or exposed rock, except Poltalloch which came to light after an excavation but was not recovered under controlled conditions. This is in marked contrast to die eight inscriptions from Orkney and Atlantic Scotland, all of which were recovered during archaeological excavations. All of this group, which includes the three chattel oghams, are informally carved, even die rather casual Burrian ogham which is inscribed on a cross-slab. Those that can be dated, Pool, Birsay 3, Buckquoy, are early postclassical (probably sixth, seventh, and eighth centuries, respectively), and some, at least, of the others (Gurness, Bac Mhic Connain) are unlikely to be late. One, Buckquoy, appears to be written in Irish, one, Burrian, in Pictish (i.e. Brittonic). This is in keeping with die accepted historical model of a Pictish population touched by Irish Christianity. The remainder have not yet been interpreted, some are too fragmentary to yield much, others, such as Bac Mhic Connain contain recognizably Celtic segments. That this group should be unlike oghams elsewhere in so many ways may reflect die feet diat while they were uncovered through controlled excavation, this is true of none of the other Scottish oghams. The density of oghams in Orkney may reflect nothing more dian die disproportionate amount of archaeological attention
THE OGHAM INSCRIPTIONS OF THE NORTH-EAST
lxviii
INTRODUCTION
which has been focussed on the archipelago this century. Three examples came from a single site, Birsay, the fourth from no more than a few hundreds away.
Similarly intense investigation of parts of the
mainland might well have uncovered comparable collections.
At the other end of Pictland, and at the other end of the monumental spectrum, there is a tight little group of oghams from around the mouth of the Tay. One is the puzzling Inchyra stone, anomalous in so many ways, re-used for three separate ogham inscriptions, none of which have been fully interpreted. On the opposite bank of the river the major Pictish ecclesiastical site of Abemethy has produced an oghaminscribed fragment of de luxe sculpture, unfortunately too small to yield linguistic information. The appearance of ogham on the magnificent early ninth century Dupplin cross above the royal and ecclesiastical centre of Forteviot is the strongest evidence for the prestige of the ogham script in Pictland. unfortunately the ogham is illegible. The cross-slab from Scoonie is somewhat of an outlier, but its text is readily interpretable as a Celtic personal name.
It is striking that there are no oghams to the north of Inchyra on the north bank of the Tay until one reaches the Mounth. That this area is densely populated with fine early Christian monuments, four of which bear roman alphabet inscriptions, proves that the lacuna is not archaeological, and indicates that interest in ogham was not uniform or universal [see Forsyth 1996d for further discussion]. Enveloped by die southern slopes of the Mounth, the Auquhollie stone is the only pillar of purely orthodox Irish type in the east. Its text is highly damaged but may be Celtic. Deeside has produced a single ogham-stone, die cross-slab fragment from Formaston which appears to be written in Irish. In the Garioch is another tight little group, comprising two, possibly three, re-used prehistoric standing stones. One, Newton, has an ogham text which is Celtic (probably Brittonic, but possibly Goidelic), though its non-ogham text has resisted interpretation, another, Logie Elphinstone has a short circular text which cannot be regarded as straightforwardly 'linguistic', the third, Brandsbutt, has long presented a puzzle, though according to a new interpretation may, in fact, be in Irish. In the hinterland of the great Pictish coastal fortress of Burghead
CAITHNESS AND SUTHERLAN] Location of sites mentioned in text
ORKNEY PENTLAND FIRTH
PAPI Of O
ULISTEH
SUTHERLAND
CAITHNESS
Ogham inscriptions
[after BATEY 1991:fig.l]
INTRODUCTION
lxix
in Moray are two formal Christian monuments with long ogham texts. Unfortunately one, Brodie, is largely illegible, apart from one Celtic personal name; although the other, Altyre, is also damaged, it too appears to contain a personal name and patronymic.
The region from the Moray to the Dornoch Firths is without ogham, though, as in Strathmore, there is fine sculpture, including one roman alphabet inscription. Beyond this are two cross slabs, at Golspie and Latheron, both of which contain personal names with patronymics, one certainly, the other probably, Celtic. Furthest north on the mainland is the Class I Pictish symbol stone from the linear cemetery at Ackergill, inscribed, it appears, with a single Celtic personal name. It appears that the latest of all Scottish oghams come from Shetland, where some, if not all, may even date to after the Norse settlement in the early ninth century. Four, Whiteness and Cunningsburgh 1-3, are too fragmentary to yield much linguistic information, the latter group are notable for the eccentricity of their different scripts. The slab from St. Ninian's Isle contains a fragmentary name and a patronymic and may be Celtic. Bressay's intriguing linguistic mix of Gaelic and Norse, or even Pictish, Gaelic and Norse, is in keeping with the mixed artistic influences of its sculpture. The remaining slab, Lunnasting, contains on recognizable Celtic personal name, and much to puzzle.
THE TEXTS
'the case for a pre-Celtic language lurking behind these inscriptions is not proven' [Smyth 1984:58]
As early as 1582, the great Scottish Humanist scholar (and native Gaelic-speaker) George Buchanan declared his belief that the Picts were 'sprung from the Gauls' [quoted by Ferguson 1991:18]. Notwithstanding the sometimes feverish popular adherence to the Germanic and Goidelic theories of Pinkerton, Skene and others, Celtic scholars were more or less unwavering in their support of Buchanan's
Cunningsburgh " 1, 2, 3, [4] St Ninian's Isle
1, PL PI
THE OGHAM INSCRIPTIONS OF SHETLAND
INTRODUCTION
lxx
view. Until, that is, 1892, when Sir John Rhys put the cat among the pigeons by proposing, on the basis of his study of the inscriptions of Pictland, that Pictish was a non-Indo-European language. He subsequently revised his conclusions, abandoning attempts to link Pictish with Basque. Although he persisted in his belief in non-Indo-European Pictish, he conceded Brittonic as the dominant Celtic influence [1898]. His theory was taken up by Zimmer who was more concerned with supposed non-linguistic evidence of non-Indo-European-ness [1898]. F. C. Diack attempted, unsuccessfully, to revive the Q-Celtic hypothesis [1944], but in general the P-Celticists held the day, especially after the publication in 1926 of W. J. Watson's The History of the Celtic Place-Names of Scotland which demonstrated the importance of the P-Celtic component in the Scottish onomasticon. Already in 1892 Alexander MacBain had declared the Pictish problem solved - Pictish was clearly a P-Celtic language [1892:288]. Yet the non-IndoEuropean thesis refused to die [e.g. Price 1984]. Eoin MacNeill's study of the names in the Pictish KingList led him to identify a strong non-Indo-European component in Pictish, though he accepted the presence of Brittonic as far north as the Moray Firth [1939]. Macalister was firmly of the opinion that Pictish was a non-Indo-European language and attempted to reconstruct elements of its grammar on the basis of his interpretation of the oghams. Kenneth Jackson attempted to resolve the contradiction between evidence for Pictish Brittonic and apparently non-Indo-European Pictish words and names. In his seminal article 'The Pictish Language', one of the most enduring contributions to the epoch-making collection The Problem of the Picts [1955], he came to the startling conclusion that there had in fact been two, quite separate languages in Pictland: one the language of the pre-Indo-European inhabitants, the other, the Gallo-Brittonic tongue of Iron Age invaders. I have argued in detail elsewhere why I believe Jackson's theory of the 'two Pictishes' to be misconceived [1996c] and will not rehearse my critique of its theoretical underpinnings. I maintain that Jackson was driven by non-linguistic considerations, that his use of both the classical and modern place-name evidence was tendentious, and that the evidence of the names of Picts preserved in documentary sources points not to non-Indo-European, but solely to Brittonic.
The supposedly
corroborative social and cultural evidence has been shown to be illusory, the argument for non-Indo-
lxxi
INTRODUCTION European Pictish, first made on the basis of the inscriptions, stands and falls by them alone.
It is simply not true to say, as Jackson did, that all Pictish oghams are 'unintelligible gibberish* [1977:222].
Difficulties of reading and interpretation are legion, but a number are perfectly
comprehensible (most obviously Scoonie). Earlier scholars have had a misplaced expectation of how easy it should be to understand them. Both Jackson and Padel talk of looking for the 'key' to Pictish ogham, as if there is some mystery to be unlocked which would cast the inscriptions in a new light, revealing them at a stroke to be as comprehensible as the orthodox inscriptions of Ireland. In this Jackson and Padel may be compared to the first generation of ogham scholars looking for the answer to the 'riddle' of the Irish inscriptions. The answer came only when hard-earned knowledge of Irish historical phonology revealed the orthodox inscriptions to be, not cryptic messages in druidic code, but pellucid texts in an archaic form of the language. Sadly, the exiguous nature of the evidence means we are unlikely ever to gain such detailed understanding of the historical phonology of Pictish.
There are three stages in interpreting an ogham inscription, and at each level there are difficulties likely to have a negative impact of the legibility of the text. The first stage is transcription (which characters are intended), the second, transliteration (what is the orthographical, i.e. sound, value of these characters ?), the third, interpretation (what are these words and what do they mean ?). Problems of transcription and transliteration have been discussed above and it should be clear why Jackson's peremptory dismissal of St. Ninian's Isle's bes and nanammowez is unwarranted. Inspection of the stone reveals bes to be not a complete word, and a more sympathetic transliteration renders nanammowez into the somewhat less rebarbative /nanamowest/.
Our present state of partial ignorance concerning Scottish/Pictish ogham
orthographical conventions is a handicap, an equally serious problem is our lack of knowledge over what kind of text to expect.
Ixxii
INTRODUCTION
The homogeneity of the classical Irish and southern British ogham stones has led scholars to look for a false consistency in the Scottish oghams when it is clear from the great diversity of contexts in wliich it is found in Scotland that the script was used to write all manner of different texts. When Irish oghams of any period depart from the narrow range of standard formulae their interpretation becomes suddenly more difficult, and there is no small number of 'baffling' Irish oghams. Page's comments on the difficulty of interpreting Anglo-Saxon runes are a salutary reminder that epigraphic opacity is not an exclusively Pictish preserve [ 1973]. An additional problem in Scotland is that the script was used to write texts in more than one language and it is not always obvious which should be expected in a given case (see for instance Buckquoy). A disappointingly large proportion of the thirty-five Scottish oghams are toofragmentaryor too weathered to yield any linguistic information whatsoever. Short disjointed sequences of letters on worn and broken slabs may look outlandish, but these should not colour our perception of the rest of the data. Only about eighteen inscriptions are sufficiently undamaged to be philologically significant. A linguistic argument based on the curious circular ogham from Logie Elphinstone, which seems more likely to have a decorative or symbolic/ritual function rather than semantic meaning, would be of doubtful validity. Even when a reading is utterly unambiguous, inscriptions consisting almost entirely of personal names are prone to linguistic problems which are specifically onomastic. A case in point is Inchyra C's SETU. Tliis informal, four-letter inscription, seems likely to be a personal-name. No such name is actually attested, but that alone is insufficient reason to reject it especially since it is likely to be a hypocoristic form. Jackson's epigraphic super-scepticism is reflected in his unwillingness to accept even Poltalloch's -CRONan- as 'certainly Gaelic* [1983:220].
His explicit rejection of EDDARRNONN is surely
unwarranted, similarly his implicit denial of the Celticity of DUNNODNATT and TALLUORRH. The Scottish oghams present many difficulties of interpretation, but these must be put in perspective. The blame lies in pan with their state of preservation, in pan with our ignorance of their orthographical conventions and linguistic contexts. Even if some were to prove to be written in an otherwise unknown language, this should not be the explanation offirstresort. Jackson assumed Buckquoy would be in Pictish and thus was satisfied that its apparent unintelligibility showed it to be non-Indo-European. It is now clear
INTRODUCTION
lxxiii
he read the text back-to-front, and, if the new interpretation of the inscription is accepted, it is not Pictish at all, but Old Irish. In the endeavour to interpret the rest of the oghams of Scotland it is not 'superhuman ingenuity' that is required [Jackson 1955:141], but a change of attitude and philological hard work.
THE INTRODUCTION OF OGHAM TO SCOTLAND 'it has become conventional received doctrine that ogam was introduced to the Picts by the Scots' [Ritchie 1987:64] There is no doubt that knowledge of the ogham script was introduced to Scotland from Ireland, the question is, when and how did it get there ? As far as Jackson was concerned, there was no issue, without further discussion he stated: ogham was invented in Ireland in the fourth century 'and subsequently imported to Scotland by the Dalriadic settlers of Argyll in the fifth' [1955:139, cf. 1983:220]. Padel, too, thought it 'fairly clear' that the Picts had borrowed ogham from Scottish D&l Rfada, and did not probe the matter further [1972:1]. There are, however, a number of problems with this hypothesis: firstly the comparative lack of inscriptions from Argyll, secondly the handful of early inscriptions well outwith areas of Dalriadic influence, and thirdly the fact that the bulk of the Scottish oghams are far later than the fifth century and, visually, have little in common with their supposed models in D&l Rfada. It is clear that Jackson's view was based on the premiss of a mass folk-migration from Ulster c.500, an event which was a 'given' in the early Scottish history of the 1950s, but whose historicity is now acknowledged to be unfounded. In its place, archaeologists and historians have put forward a model whereby an intensification of long-established links across the North Channel, led to a shift of emphasis in the power base of the ruling elite, culminating in their permanent transference to Scotland in the late fifth century. Even if the introduction of Dalriadic culture into Scotland was a process rather than an event, it seems unlikely that the ogham script was one of its components. As early as 1970, Leslie Alcock had pointed out that there was no evidence that the Scorn had a tradition of using the script in their homelands, ogham
INTRODUCTION
lxxiv
inscriptions being completely absent from Irish Dál Ríada [1970:59]. As all previous commentators have pointed out, ogham in Ireland is densely concentrated in the south of the island (five-sixths of all extant Irish examples are from Munster). The mechanisms by which it spread elsewhere in Ireland are poorly understood, but only a handful of inscriptions have been found in die north (eight extant examples from the whole'province of Ulster). There is no need to suppose that the borrowing came to Scotland across die shonest stretch of water, and it is at least as likely that ogham reached Argyll independently from southern Ireland by the same kind of processes which brought it to Connacht and Ulster. Nor need we suppose that ogham's route to the Western or Northern Isles was necessarily via Argyll (though it might have been). Our ignorance of the role of ogham in contemporary Irish society is a major stumbling block. Until we better understand who in Ireland knew or wanted ogham we will make little progress in understanding its introduction into Scotland.
The extent of Gaelic ecclesiastical influence in Scotland (east and west) is well-known, and the Church has been put forward as a likely conduit by which ogham entered and was spread. If it was the Church, then, as far as Alfred Smyth was concerned, the culprit was Iona 'the great disseminator of ideas and motifs throughout Pictland' [Smyth 1984:79, 58], The influence of Iona in Scotland was indeed strong, but ogham does not feature at any known Columban foundation in Ireland or Scotland. Furthermore, if it was through the Columban church that ogham reached Orkney, then, as Anna Ritchie said, 'the long arm of Iona reached far indeed' [1987:64], and, one might add, 'early', since the Pool ogham is at least as early as the sixth century. The role in Scotland of non-Columban ecclesiastics such as Maelrubha of Bangor has long been acknowledged, but there is a growing recognition of the extent of ecclesiastical links with southern and western Ireland (the voyages of Munster ascetics to the Northern Isles is discussed in the entry on Buckquoy, for links between Leinster and Abenay, see Abemethy and Inchyra). The appearance of ogham in the marginalia of Irish manuscripts and on a slab from Clonmacnoise testifies to ecclesiastical knowledge of ogham in Ireland in the ninth century.
Some of the Scottish oghams (for instance,
Formaston) have almost certainly sprung from a similar milieu, but a number (Auquhollie, Pool, Bac Mhic
INTRODUCTION
lxxv
Connain) are surely too early to have come from Christian contexts.
Ogham took root in Scotland, as reflected in its chronological, geographical, contextual, and stylistic diversity. To a certain extent, it followed its own avenues of development there, palaeographical and orthographically as outlined about, and also in the new range of high-status contexts in which it was employed. It is clear, however, that Scottish oghamists remained up-to-date with developments in Ireland, as reflected in their use of the new value of m4 at least as early as the first extant Irish examples (see entry on St. Ninian's Isle). It would be wrong to imagine that all interchange took place on Scottish soil. The Life of Fintan reveals Pictish ecclesiastics in the early ninth century undergoing training in Ireland and learning Irish (see entry on Buckquoy), they may well have picked up knowledge of ogham there, whether or not it was pan of the 'official' curriculum. That no single explanation fits all circumstances merely underlines what Padel has pointed out, that ogham in Pictland (I would say Scotland) was the result not of a once-off borrowing, but rather of on-going cultural contact. It may, in fact, reveal as much about Irish ogham since the extant Scottish inscriptions presuppose a continuing ogham tradition in Ireland after the seventh century, even though there is little direct archaeological evidence for it. The archaeological survival of post-classical ogham in Scotland reflects the script's high-status which led to its being preserved on monumental sculpture. The marginal status of the script in Ireland means that examples of post-classical oghams in Ireland are restricted to chance finds of informal inscriptions (Kilgulbin, Ballyspellan), graffiti (Knowth) or the marginalia which are the manuscript equivalent of graffiti (St. Gall).
A detailed
investigation of post-classical Irish ogham is required before the oghams of Scotland can be seen in their true light, but they in turn may illuminate ogham in Ireland.
Ixxvi
INTRODUCTION SCHEDULE OF READINGS
The following is intended as no more than a rough guide. A number of the transcriptions are tentative and readers are urged to consult the detailed description in the entries concerned. For a list of conventions used see Guide to Entries. For further discussion see Introduction: Transliterating Ogham, Transcribing Ogham.
Abernethy
-QMI- or -IMN
Ackergill
NEHTETRI-
Altyre
-]AMMAQQ(q/tx /c)A(ll/n)MvsMAHHR( /a)R(a/aa/o)sSuDDS
Auquhollie
VU(u/o)nON [?auoa/a(c/t)o?] TEDOV(u/o)(n/s/r)(i/e/r)
Bac Mhic Connain
M(a/o)QUNTEN( /a)CoT
Birsay 1
(m/h/a)ONNORRA(n/vv/lv)RR
Birsay 2
-(b+/a+/m+)(a/ )QI(:/ )(t/ )A(b+/a+/m+).
Birsay 3
-m(a/ )(b/ )QQHO(b+)-
Blackwaterfoot 1
H(i/e)OMEQ(i/e)H
Blackwaterfoot 2
VUEDLA
Brandsbutt
IRATADDOARENS(-
or
NEHTETRE(b+/m+)
INTRODUCTION Bressay
lxxvii
CRRO( /s)SCC : NAHHTVVDDAddS : DATTRR : aNN(-) BENiSESMEQQDDRoANN(-)
Brodie
(a) -](v/v+)ON(a+/a+m+/a+b+/a+h+)(e/i)( /v)CC(i/obh/oo)[(b) see entry -]EDDARRNONN[- -(o/a+/g+)(Il/s+/SÍ+)(t/u)]Tl[(m+/rr)[-
(c)
Buckquoy
ENDDACTANIM(v/lb)
Burrian
I(t/d/o/u/?)( /e/b)IR( /a)RANN(u/)(./ )RRACT(k/e/m/?)EVVC>
Cunningsburgh 1
-]I( /d/h)R(u/e/i)[-
Cunningsburgh 2
Arris: -]d+(o/u/e)V( /o)DDR(s/n)( /oa/u)[Face:
Cunningsburgh 3
(circular, possibly bENDDACTANIMl)
-EHTECONMORs
(a)
xTTEC(o + /g + )[-
(c)
-](a+/m+)VVR[-
(b) -]a + (v/bl): DATT(v/u)(v/b+/m+)[-
Cunningsburgh 4
lost
Dunadd
Upper
(ae/i/q)S(d/o)(u/at/i/q/e/c)[—]V( /a/o)(n/lv)A( /d/o)[crack](t/ch/v)
Lower
FI(nn/rr)MaNA(ch/q)
Dupplin
illegible
INTRODUCTION
lxxviii
Formaston
MAQQoiTALLUORRH | NxHHTVROBBACCxNNEVV
Gigha
vi(q/c)ulaMAQ(u/ /i)(c/t)0( /m)GI(n/l/bb)i
Golspie
ALLHALLORREDDM(a/e)QQN(i/ia/uu)V(h/a/)(rr/ni)(c/e)(?)(rr/qq)
Gurness
IN(e/c)IT( /a)TEM(o/om/ob/u/e/?)N( /.) |
Inchyra A (i)
(e/uu/ou)TTL(i/uo)ETR(e/s)N(o?)IddOR(s/v)
Inchyra A (ii)
-](u+)HTu(o/i)( /a)GED[-
Inchyra B
INEHHETESCIE(t/c/d)(t/.)INNE
Inchyra C
(s/n)ETU(+)
Latheron
DUnNODNNATMAQQNET(u+)-
Lochgoilhead
(/h)MUD(?)ALI( /d)
Logie Elphinstone
QFTQU (circular)
Lunnasting
XTTuCUHXTTS : AHXHHTTJNNN : HCCVVEVV : NEHHTONs+
Newton
IDDARRNNNVORENNIxO(t/c)(c/e) | (i/r)OSR(r/n)
Poltalloch
-CRONaV+-
Pool
-sROTATR- or -RVAVORc-
(-CRONaN-)
MATS or CFAM
INTRODUCTION
Ixxix
St. Ninian's Isle 1
-b+]ESMEQQNANNAMMOVVE&
St. Ninian's Isle 2
lost
St. Ninian's Isle 3
lost*
Scoonie
-)EDDARRNONN
Whiteness
-](v/s/n)N(:/d)AR[?- or -?]RA(:/l)Q(t/c/q)[-
lxxx
GUIDE TO THE ENTRIES
All known ogham inscriptions from Scotland are listed in alphabetical order of site. If there is more than one inscription from the same site these are numbered 1, 2, etc. (e.g. Cunningsburgh 3). If an object has two or more separate inscriptions these are labelled A, B, etc. (e.g. Inchyra B). Doubtful inscriptions are excluded from the main body of the catalogue and dealt with in an Appendix (Dubia). Typically, each entry is presented as follows (some headings optional);
NAME Name of adjacent farm, village, parish, or island where found. Despite the inconsistencies of naming, no attempt has been made to change names established in scholarly literature.
DISCOVERY Description of current location, including National Grid Reference (NGR) to six figures, or eight where possible, civil parish and pre-1974 county (in keeping with NMR convention for identifying sites). Where appropriate, ownership or scheduled status is noted. If in museum collection, date and circumstances of donation and catalogue number. If on display in NMS at time of writing this is noted (N.B. the collection is shortly to be moved to the new Museum of Scotland where it will be rearranged, some pieces may be taken from the store for display, and vice-versa). Description of original location (where known) if different from current location, also details of any intermediate moves (including date and circumstances of any subsequent damage or modification). Date and circumstances of discovery or first mention. Publication details, especially first publication. Alternative name for item if appears thus in literature.
SITE / LOCALITY Description of site where object found and details of any archaeological excavation conducted there. Details of any early medieval inscriptions in non-ogham scripts (runic inscriptions are refered to by their
GUIDE TO ENTRIES
lxxxi
number in Barnes's catalogue of Scandinavian runic inscriptions of the British Isles [Barnes 1992]). Notice of local material likely to have bearing on the use of the site in the early medieval period, including artefacts, settlement remains, historical references, place-names, and saints dedications where appropriate. Discussion of the wider context of the item, as above, also including, if appropriate, a physical description of surrounding area (relief, land quality, access), strategic importance, and relationship to major contemporary sites and monuments, and natural features. Historical references to region in the period, especially Irish connections.
DESCRIPTION OF OBJECT Brief general description of object and location on it of inscription. Stone/material: Geological identification taken from secondary literature. Dimensions:
Maximum height x width x thickness, given in metres (m), to nearest 0.1m. Most measurements are by the author. In the case of those which are not, if metric measurement is an approximation from published imperial measurement these are given also.
Condition:
Assessment of completeness of object and inscription, including degree of weathering and other damage.
Fuller description of weathering, etc. Estimation of size and nature of original object if fragmentary. Discussion of function of object if known. Scottish sculpture is classified acording to Romilly Allen's scheme: Class I = unworked stone incised with Pictísh symbols; Class n = cross-slabs bearing Pictish symbols; Class III = related sculpture with no Pictish symbols. The additional Class IV (simple crossmarked stones) suggested by Henderson [1987] is also adopted.
CARVING TECHNIQUE Where possible, a detailed description of physical appearance of incised lines (both inscription and other carving), including dimensions of individual strokes, description of cross-section of strokes. Measurements
GUIDE TO ENTRIES
lxxxii
of carving are given in millimeters (mm), at least to nearest 5mm. Where it can be assessed, description of technique used (e.g. pocking or scoring). If appropriate, comparison with carving technique of other examples. It was not always possible to take undertake detailed analysis and measurements of inscriptions on display in museums. Relation of ogham to other carving including assessment of whether or not inscription is primary, i.e. if the monument was intended from its inception to bear the text.
DESCRIPTION OF CARVING - DECORATION Description of form of all carving on object other than inscription (figurative, geometric, etc.), including comparison with similar motifsfoundelsewhere. Unless otherwise specified, references to early medieval Scottish sculpture are to Part HI of ECMS (see ECMS index). Location of Scottish sculpture given using standard three-letter county abbreviations (see list of abbreviations, above). References to different types of cross-shape, interlace and other geometric forms of ornament are to the numbered classifications laid out by Romilly Allen in Part II of ECMS. Where appropriate a note is made of iconographic themes and wider artistic affinities (e.g. Irish or Anglo-Saxon influence).
DESCRIPTION OF OGHAM INSCRIPTION The length of the inscription and dimensions of typical strokes are noted where possible. If the inscription is not complete the likely length of missing portions is assessed. The direction of reading is established. Detailed description of ogham inscription stroke by stroke. Each letter or group of strokes is assigned a number (italicized) which is keyed to an accompanying diagram [see fig. for key to numbering]. A note is made of any unusual letter forms but values of forfeda are discussed under 'Form of script'. All readings are based on independent personal examination by the author unless otherwise stated. Comparison is made with descriptions by previous authorities. See introduction for detailed discussion of conventions followed when describing ogham inscriptions. N.B. Unless otherwise specified, all directions (right/left, forward/backward, above/below) are relative to the direction ofreading. Since most inscriptions are placed
GUIDE TO ENTRIES
lxxxiii
venically and read from bottom-up, not all directions are meaningful relative to objea as it currently stands.
The following conventions are used in giving readings of inscriptions (see Introduction for fuller explanation): Uppercase letters reasonably certain. Lowercase letters doubtful.
If a reading is uncertain, alternatives are given in brackets in decreasing order of preference, separated by an oblique stroke, e.g. (t/u), most likely - T, alternative reading - U; (/a/h), most likely - no letter, alternative reading - A, least likely - H.
a \ t + , s + , etc. = 'A or any subsequent letter of the draicme', T or any subsequent letter of the h-aicme\ etc.
Underline - two letters in one character compendium, e.g. £A, RR.
-]
portion of unknown length lost at beginning
[-
portion of unknown length lost at end
-)
beginning of text intact, or possibly portion missing
(-
end of text intact, or possibly portion missing
[..]
missing portion, one point for each missing letter
[—]
missing portion, number of letters unknown
|
end of line
:
word-division given in text
#
word-division supplied
GUIDE TO ENTRIES
lxxxiv
FORM OF SCRIPT Description of the palaeography of script used, including the possible orthographic value of any special characters. Discussion of distinctive characteristics and comparison with scripts of other inscriptions, especially if these might have dating implications.
INTERPRETATION OF TEXT Any difficulties of transliteration. Linguistic comments including possible word-division, if this is not specified in text. Identification of possible personal names, name-elements, or other words, and discussion of other attestations (especially epigraphic). Translation of text where possible/appropriate. Identification of persons named where possible.
DISCUSSION Discussion of general significance of inscription and monument. Any other points of interest.
BIBLIOGRAPHY OS Record card - (number of file in NMR). Select list of references by author in chronological order, format = SURNAME date:page. The following are excluded: secondary works which merely list the inscription or which contain entirely derivative information (unless the original work is obscure); works containing passing references are listed if these are of significance for locating the stone at a given time. While complete consistency has been impossible, the general principle was to include all scholarly references likely to be of use to those with interests other than the purely antiquarian. Since the theories of Anthony Jackson are not endorsed by die author and since his readings are derivative of Padel, his Symbol Stones of Scotland (1984) has not been included in the bibliography of individual entries. He treats the Pictish ogham inscriptions on pages 174-200. In general, all known published illustrations are included, unless these are derivative or very poor. If a work
GUIDE TO ENTRIES
lxxxv
features only an illustration of the item, this is noted. Any works discussing site or object but not inscription are listed separately.
PREVIOUS READINGS All previous readings based onfirst-handobservation are included in chronological order, except those of purely antiquarian interest. Since the readings in A. Jackson (1991) appear to be entirely derivative of Padel's these are not included.
ILLUSTRATIONS The illustrations for each entry follow on unnumbered pages (see List of Illustrations). Typically, illustrations include photograph and/or drawing of object and inscription and a schematic diagram of lettering giving key to numbering. Where appropriate site plans, and illustrations of other material from the locality are also given.
Ixxxvi
"In lapidary inscriptions a man is not upon oath" Samuel Johnson to James MacPherson, 1775
1
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
(IN ALPHABETICAL ORDER)
1
Abeniethy
21
Latheron
2
Ackergill
22
Lochgoilhead
3
Altyre
23
Logie Elphinstone
4
Auquhollie
24
Lunoasting
5
Bac Mhic Connain
25
Newton
6
Birsay 1-3
26
Poltalloch
7
Blackwaterfoot 1-2
27
Pool
8
Braodsbutt
28
St. Niuian's Isle 1-3
9
Bressay
29
Scoonie
10
Brodie
30
Whiteness
11
Buckquoy
12
Burrian
13
Cunningsburgh 1-4
14
Dunadd
15
Dupplin
16
Formaston
17
Gigha
18
Golspie
19
Gurness
20
Inchyra
THE OGHAM INSCRIPTIONS OF SCOTLAND
2
ABERNETHY
DISCOVERY This fragment was discovered in 1890 by a Mr Marr of Abernethy, Perthshire [Southesk 1895:245]. According to Romilly Allen it wasfoundin the ancient churchyard (NGR NO 190164) [ECMS 308; 1891], but die local minister Dugald Butler maintained, alternatively, that the stone had come from 'a house in the Station Road* [1897:230]. He may have been confusing it with one of three other fragments found built into masonry in the village [ECMS 308]* Dr Alexander Laing of Newburgh-on-Tay presented the stone to the National Antiquities Museum on 9 May 1892 (NMS Cat. No. IB 98), [For a second allegedly ogham-inscribed fragment from Abernethy, see Appendix - Dubia].
SITE Documentary sources indicate that Abernethy, situated near the confluence of the Tay and the Earn, was one of the principal ecclesiastical sites of early medieval Scotland. The name, significantly, is P-Celtic* attested in the Pictish Chronicle as Abur-nethige, The first element is the Brittonic aber 'confluence' [Watson 1926:458], the second the genitive of Nethech or Nettheach (fem.), a Gaelicization either of Neithan or a British river name of the same root (i.e. < *Neciona fpure one*) [Watson 211], Nicolaisen has commented on die possibly cultic origins of sites with aber names [pers. comm J . Barrow [1973] and Driscoll [1991] have used documentary and place name material as well as archaeology and air reconnaissance to reconstruct the extent of the thanage of Abernethy and the nature of the relationship between its constituent parts [see map]. Driscoll draws attention to the density of P-Celtic place-names and to the proximity of the hill fort of Clatchard Craig [100-102].
Alan MacQuarrie has recently discussed in detail die irreconcilably muddled' foundation legends of Abernethy [1992:115-8]. Textual analysis of die relevant sources was provided by Marjorie Anderson [1973:92-6]. These are (i) two notes embedded in Version P of the Pictish King-list preserved in the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ABERNETHY
3
fourteenth century Poppleton MS (Paris, Bibliothéque Nationale, MS Latin 4126), thefirstof these is dated by Anderson as 'later than about 724, and... considerably earlier than 1093* [1973:95], it appears to have been written by a Brittonic speaker, possibly at Abernethy [A. Anderson 1922 (I990):cxx-cxxi], and (ii) a fifteenth century interpolation made by Walter Bower (abbot of Inchcolm) in his copy of Fordun's version of the Pictish King List [Scotichronicon I V.12; McQueen 1989:302, 457-9], Bower claims to have used quondam cronica ecclesie de AbirnetM 'a certain chronicle of the church of Abernethy'.
The two traditions diverge in details, but agree on the following points: Abernethy was founded by a Pictish king (possibly Nechton or Gartnait) at or near a Pictish royal centre; it was in some way associated with St. Brigit of Kildare; and it considerably pre-dated Dunkeld. MacQuarrie is inclined to accept the link with Brigit, which he sees in the context of contemporary activity in Scotland by other Irish ecclesiastics, such as Columba, Maelrubha of Bangor and Brendan of Clonfert, and takes it as evidence of a foundation at Abernethy by the seventh century [117]. Marjorie Anderson, however, is more sceptical. She points out that there were several saints called Brigid and that the identification of the Brigid of Abernethy with the famous founder of Kildare may be a later rationalization [1973:95]. Simon Taylor has pointed out to me the possible dedication to a Leinster saint (Maedoc) of the parish of the parish of St. Madoes, on the opposite bank of the Tay [see entry on Inchyra]. This, he suggests, may point to a genuine early ecclesiastical connection between the Firth of Tay and south-east Ireland. The alleged link with Kildare is the only evidence for a monastic church at Abernethy in the Pictish period. MacQuarrie thinks that the royal connection may be anachronistic also, but highlights the proximity to Abernethy of an Iron Age hill-fort possibly re-used in the Pictish period [115, 117]. The stress in the legends on the antiquity of the foundation and the involvement of a great Pictish king may spring from an attempt to affirm the authority of Abernethy in the face of increasing assertions by the royal foundations of Dunkeld or St. Andrews.
The emphasis on Kildare is interpreted by the McQueens as 'an attempt by 8c Pictish
ecclesiastical authorities to play down the importance of Columba and Iona for the Pictish church' [458].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ABERNETHY
Bower claimed Abernethy as the principal episcopal centre of the Picts, but already by the mid-ninth century it seems to have been eclipsed by Dunkeld. The Annals of Ulster for 864 note the death of one Tuathal son of Angus, primepscop Fortrerm and abbot of Dunkeld. Though Abernethy appears to have become marginalized after the elevation of Dunkeld, it remained an important and wealthy ecclesiastical centre. By the early twelfth century the house there consisted of a community of céli dé with a number of ordinary clerici [Lawrie 1905:11-12].
A Class I symbol stone provides further evidence of activity at Abernethy during the Pictish period [ECMS 282]; the foundations^ continued importance in the post-Pictish period is attested by four fragments of high-quality sculpture including two cross-shafts, one clearly Irish-influenced [ECMS 308-10, 312]. All have been severely cut-down, and all except the ogham-inscribed stone were found built into modern structures. A small, probably later, cross-incised slab was discovered in the churchyard in 1894 [ECMS 341-2; Butler 1897:237 pl.VII]. The most striking archaeological evidence for Irish influence at Abernethy is provided by a well-preserved round tower (one of only two in Scotland), dated by Ralegh Radford to the second half of the eleventh century [1942:3-4], and by Fernie to c.1100 [1986:393].
DESCRIPTION OF OBJECT A sandstone fragment carved in relief on one face with the lower parts of two equine legs and a raised band carrying the remains of three bound ogham letters [see fig.]. Stone:
Old Red Sandstone
Dimensions:
360 x 200-230 x 70mm
Condition:
Fragmentary but otherwise in good condition, letters clear.
This stone, like several of the others recovered from Abernethy, has been severely cut down, probably for use in much later building work. The scale of both letters and horses implies that the extant fragment represents only a small proportion of the original, but there are virtually no other clues to the size and nature of the monument. None of the edges of the fragment are original and, since the back has been
4
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ABERNETHY
5
shaved off, it is impossible to tell if the stone was sculptured on both sides. Free-standing cross-slabs are the most common monuments of the period in the area, but the Abeniethy fragment could have been a recumbent (cf. Meigle 9, 11, 12,26 PER; Strathmanine 2 FOR; St. Vigean's 14 FOR) or an architectural element such as a frieze (cf. Meigle 10, 22; St. Vigean's 8 or Murthly PER).
DESCRIPTION OF CARVING The band prepared for the ogham is 90mm broad, approximately one quarter of the total height of the extant ponion. Below it is a blank surface at the same level as the background to the carving above the band. In its bottom left-hand comer is a semi-circular ponion which has been deliberately defaced. The placing of ogham on a broad raised band separate from the rest of the carving is closely paralleled on the fragment from Whiteness, and the slab from Brodie, and, to a lesser extent, by the stones from Formaston and Golspie.
The upper half of the fragment is badly damaged. All that survives in this area is the lower pan of two legs of a horse, or possibly a deer, facing left. Macalister hints at the hoof of a third leg [1940 pi. Vic], but there is no trace of it on the stone. Horses, with or without riders, are a common motif on Pictish sculpture, especially in the south of the Mounth [Carrington 1995]. They are depicted in a number of stereotyped ways, one of the most prevalent forms seems particularly relevant to the high-stepping horse(s) of the Abernethy fragment. Examples of this type are seen, for instance, on the St. Andrews sarcophagus, on the Meigle 11 recumbent [see fig.], and on such cross-slabs as those from Rossie Priory, Logierait, and Meigle (5 and 2). In it, the horse is depicted in profile facing left, with legs arrayed in the following sequence - first the front right (far) leg raised and slightly flexed, then the front left (near) leg straight down on the ground, thirdly the back left (near) leg raised and well flexed, and finally the back right (far) leg bent slightly back and touching the ground.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ABERNETHY
6
The relative slope and position of the two legs on the Abernethy fragment suggests that they represent either the front and back near legs of a single animal, or the back far leg of one horse and the front far leg of another, immediately behind it. In favour of the latter interpretation is that the front of the two Abernethy legs has a distinct slope like the fourth leg of the archetype. Although the slender legs and sharply rendered hooves of the Abernethy fragment are strongly reminiscent of, for instance Meigle 11, there are several other different stock horse-types and it would be unwise to labour the point with so little surviving.
There is no reason to doubt that the ogham is contemporary with the rest of the carving, and the whole is most naturally interpreted as part of one over-all design. The difficulty comes in the attempt to reconcile the apparently conflicting orientation of text and horse. The most natural explanation of the carving is that one or more horses are walking in procession along the ogham band. This is the interpretation favoured by Rh$s, Allen, and Macalister, as implied by their preferred reading of the ogham as IMN. Their third letter, however, slopes the wrong way to be an N. In Type II inscriptions, when m- and b-aicmi letters are not perpendicular to the stem, their slope, if marked, is always forward relative to the direction of reading. To maintain this convention, the ogham must be taken the other way round and - I M N - becomes - Q M I - . If the ogham is not to be read upside down, thefragmentwould have to be rotated 180° which would put the horse upside down with its hooves in the air. Of course, ogham inscriptions are typically placed vertically, and one possibility would be to have the Abernethy text reading vertically up the left side. This, however, would entail a horse standing on its hind legs with its fore-legs in the air. The seahorses which appear on the Aberlemno Battle stone, Meigle 1, Meigle 11, Murthly and Largo do hold their fore-legs in the air, but they face towards each other in pairs, and never have their hooves against an outside edge.
One final scenario, the only one which permits both horse and ogham to go forward in opposite directions simultaneously, is suggested by the form of the recumbent Meigle 11 [see fig.]. If the Abernethy fragment
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ABERNETHY
7
was pan of the upper face of a recumbent with thick raised border, and horsemen riding to the left in a sunken panel, just as they are on Meigle 11, then an ogham inscription which began in the bottom left-hand corner, rose vertically (as is normal) continued across the top, down the right-hand side and along the bottom [see fig.] would have the portion under the horses hooves reading forward right-to-left. This arrangement of text, with letters facing in, is familiar from seventeenth century Scottish gravestones. A medieval example is the eleventh century Latin grave-slab at Stratfield Mortimer, Berkshire [Okasha 1971:114-5].
In evaluating the relative merits of the above interpretations the following precedents should be borne in mind: ogham inscriptions are always read from left to right; there are no clear exceptions in the Scottish ogham corpus to the rule that sloping consonants slope forwards relative to the direction of reading; there are no horizontal oghams in Scotland except for (a) the latter pan of Golspie, which is a continuation of the vertical text and therefore conforms to standard Irish practice, and (b) Dupplin, which is also a continuation of a text which starts vertically and follows the arris. Unless one is prepared to argue that the ogham and the horse are not contemporary (and there seems no warrant for that), then only in the final scenario can the evidence be interpreted in keeping with all the above conventions.
DESCRIPTION OF CARVING - OGHAM INSCRIPTION The stem-line occupies the full width of the extant fragment (225mm). Since it protrudes beyond the outer strokes of the first and final letters, presumably it extended further in both directions. The individual strokes consist of deep U-section grooves that have been first pecked then smoothed. The carving, comprising three complete letters, is very clear and well-preserved. The slope of letter / (b-aicme) [see fig. for key to numbering] indicates the direction of reading (see discussion above). Padel drew attention to the uneven spacing of the letters [1972:42], but with such a short section extant it is difficult to judge the significance of this. Letter 2 follows hard on letter 7, but there is an unexpectedly large gap between it and letter 3 (50mm at the stem, almost as broad as letter 3 itself). The left tip of 2 is, however, roughly
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ABERNETHY
8
equidistant from the tips of the two strokes to either side and it is possible that the oghamist arranged the letters strictly according to the space between the tips of the strokes rather than making due allowance for differences in the slope and width of letters. An analogous contrast would be between the regularly spaced print of a typewriter, as opposed to the proportional spacing of text produced on a word-processor.
The inscription is as follows [see fig. for key to numbering]: 1
Five short oblique strokes to the left of the line, slopingforwards,bound. Individual strokes 3540mm in length. Even spacing between strokes - total width of letter 65mm. - Q.
2
One long (50mm) oblique stroke across the stem, slopingforwards- M. Gap of 50mm.
3
Five long strokes perpendicularly across the stem, bound. Individual strokes 52mm in length. Even spacing between strokes - total width of letter 55mm. Since 2 indicates that m-aicme letters are sloped, 3 must be a vowel - I. Allen has the left bind-stroke protruding beyond the final stroke, but this is mistaken [ECMS 309 fig.323].
This gives a reading:
--QMI--
or, in the unlikely event that the text should be taken in the opposite direction:
-IMN -
FORM OF SCRIPT With only three letters surviving there is little to be said about the script of the Abemethy fragment other than that it was written in an accomplished, monumental form of Type lie. The stem-line is visible, the straight vowel-strokes occupy the full ogham band, the h-aicme letters slope in the direction of reading. The component strokes are equally and generously spaced. The spacing of the letters is surprisingly uneven, but given that only a short sequence is extant it would be unwise to make too much of this. Abemethy is one of only two examples of the use of bind-strokes outside the Northern Isles, it is significant that the other is at nearby Inchyra.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ABERNETHY
9
INTERPRETATION OF TEXT The extant text, though clear, is too shon for any meaningful linguistic analysis. The appearance of die letter Q in an ogham always suggests some version of the formula word MAQQI. With the dubious possible exception of some very worn sections at Brodie, Q has not yet appeared in Scotland except in this word, which consistently spelled with Q(Q) rather than C [see Golspie, Latheron, Bressay, et al.], despite the later date of most Scottish texts. Such an interpretation would entail a reading 'x maq]Q MI[-'. Though any gap in die carving occurs between die M and I this has little bearing on possible worddivision, which is not usually indicated by spacing of letters alone, and, as explained above, is more likely created by a desire to keep die letter tips evenly spaced. The Aberaethy fragment would long post-date apocope, so we should not be looking for a genitive ending -I.
A handful of Irish male and female personal names begin A//- [see, for example, die indices of CGH\ Ó Riáin 1985; Ó Corráin & Maguire 1981]. Two loosely connected with Scotland are Columba's sister who, traditionally, was called Minchloth [Anderson 1922 (1990):23 (Continuation of Adomnán)] and the father of two of his twelve original companions, Midgna son of Meti [Anderson 1922 (1990):40 (Irish Life of Columba in Lebar Brecc)]. The biblical borrowing Miche(a)l (< Michael) was rare in the early period [Ó Corráin & Maguire 1981:136], though it might be considered appropriate to an ecclesiastical context like Aberaethy. In the form Mael Micheil and Mael MHchiU 'devotee of St. Michael', die name appears on three inscribed stones from Clonmacnoise [CIIC 606, 738, 801]. In addition to several early dedications to Michael in the north and west of Scotland, there is a Kirkmichael in Perthshire and one in Buchanan, Stirlingshire [Redfoid 1988:249].
DISCUSSION Although very little of this text survives it provides imponant evidence for die use of ogham at a major ecclesiastical site. Ogham is striking in its absence from the principal monastic and episcopal sites of early medieval Scotland (incl. Iona, St. Andrews, Dunkeld, and Brechin). This may have some significance for
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ABERNETHY
10
an evaluation of the nature of the royal involvement at Abemethy, which should perhaps be compared with Foneviot as a centre of both secular and ecclesiastical power. Both sites enjoyed prominence in the Pictish period proper but were eclipsed after the mid-ninth century rise of Dunkeld. The presence of equestrian motifs in Pictish an has been attributed to the taste of the secular nobility. Perhaps the Abemethy fragment, with its collocation of ogham and horse motif, may, like Meigle 11, be pan of a monument to a Pictish magnate who had arranged to be buried within the precincts of the great church.
BIBLIOGRAPHY
OS Record card - NO 11 NE 20
Browne 1891; Allen 1892:254; Rh?s 1892:261,268; Southesk 1895; Nicholson 1896 (B - 2, 4, 26, App. 85); Butler 1897:230-1, pl.II; Rh?s 1898:345-6; ECMS 308-310fig.323;Macalister 1940:215, 216 pi.vie; Diack 1944:74 XIV; Padel 1972:42-3.
Site Only:
Radford 1942:3-4; McQueen 1989:1 V.l2,pp.457-9;Feraie 1986:393; Driscoll 1991:100-
102.
PREVIOUS READINGS
Rh?s
IMN or QMI
Nicholson
QMI
ECMS
IMN or QMI
Macalister
IMN
Padel
...QMI... 'or less probably ...IMN...'
[(meq)q Mi-]
THE SHIRE OF ABERNETHY Chape&and cforctas d^enleni onfibarvury.*»12thCerttry A
FGfts
•
A Psrfes
-f-i- Qucf^Cross
ABERNETHY - Conjectural map of the early medieval shire of Abernethy [Driscoll 1991 fig. 53]
Coiitia^)
[BUTLER pl.II]
ABEINETHY - Ogham-inseribed fragment
(key to numbering)
«tf.-'-V**
JS*
*c*
Side view of recumbent grave monument, Meigle 11 [ECMS fig. 345a]
w 4^f ABERNETHY - Possible reconstruction of monument
i
11
ACKERGILL
DISCOVERY According to Mrs Duff-Dunbar of Ackergill [OS record card], this stone, a Class I Pictish symbol stone, formerly stood upright at the north-west end of a long mound near Ackergill Tower in the parish of Wick, Caithness (NGR ND 3483 5499). It was found on the links towards the south side of Keiss Bay in August 1896 by John Nicolson of Nybster, broken into several portions [ECMS:28]. In 1897 the stone was presented to the National Antiquities Museum by Sir Francis Tress Barry (NMS Cat. No. IB 168).
SITE The long mound at Ackergill is a natural elongated sand dune (c. 120m long, max. 27m across and 2m high), much mutilated by Second World War defence works. It was excavated by Edwards in the 1920s and found to contain several long cist graves in rectangular cairns [see fig.]. Funher graves were later uncovered during the building of a road, and as a result of coastal erosion in the early 1980s [OS record card]. In conjunction with the graves found at Keiss Bay [Ashmore 1980:349], the linear inhumation cemetery at Ackergill comprises the southern pan of a concentration of burials almost two miles in length [Batey 1991:53]. Edwards uncovered sixteen burials in ten separate graves; two isolated long cists, one isolated round cairn and the rest long cists covered by square or rectangular kerbed cairns of a typically Pictish form which Close-Brooks calls 'platform cairns' [1984:99]. Of the latter group, all contained extended inhumations without grave-goods. The skeletons were of individuals of both sexes and all ages from babies to the elderly. The people were lightly built, with an average height for males of 5'5" and forfemalesof 5'. Edwards considered the remains represented a 'native population of mixed origin* [1926]. A bronze chain found round the neck of one of thefemaleskeletons was thought, at first, to be comparable with Norse jewelry and dateable therefore to some time in the tenth century. This comparison has been undermined subsequently. Close-Brooks compared the chain to one from the Roman fon at Newstead dating to thefirstor second century AD [1984:97]; Batey preferred the comparison with a chain
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
12
found in the Nome's Law hoard of Picrish silver [1991:53]. The recent recognition of die low, rectangular, kerbed cairns as being a Picrish type has made it Virtually certain* that Ackergill is a Pictish cemetery [Close-Brooks 1991:97].
In addition to the ogham-inscribed slab, another Class I symbol stone was found lying on the surface of the centre of the mound near the head of Grave 1 (c.ND 3487 5497). This stone, now in the NMS (IB 206), is a much-weathered fragment of clay slate, 0.30 x 0.42 x 0.03m [OS record card ND 35 SW 12]. It is incised on one face with a 'notched rectangle' and pan of another symbol [Edwards 1926:179, CloseBrooks 1984:97]. Though the two symbol stones are closely associated with graves, a direa correlation cannot be proven beyond doubt due to the vagueness of accounts of the circumstances of their discovery. The same is true for other northern symbol stones associated with burial cairns, such as Watenan and Dunrobin. There is, however, mounting evidence - for instance the Inchyra and Picardy ABD stones - to show a link between Class I symbol stones and graves. A third Class I symbol stone was found at Keiss, upside down as pan of the paving of a rectangular structure at nearby Birkle Hills [ND 3392 5847]. This fragment of sandstone is incised with two symbols, the 'mirror* and the 'triple ovals', and is now in the NMS (IB 188) [ECMS 27-8, RCAHMS 191 Ib No.577, Close-Brooks 1984:99, OS record card ND 35 NW 5].
As Batey pointed out, the undoubted importance of the Ackergill cemetery has been 'overshadowed by the apparent lack of associated settlement remains' [1991:54]. This is pan of a general lack of Pictish settlement archaeology throughout Sutherland and Caithness. Batey mentioned three brochs within a one mile radius of Keiss, the cellular structures surrounding which may date possibly to the Pictish, or 'protoPicrish' period [49], The painted pebble found in the area is further evidence of Pictish activity [49], Further away, a few miles to the north, is the important Norse settlement of Freswick Links which appears to have superceded a Pictish settlement [Batey 1991:51,58]. The place-name Papi Ceo on the coast just south of Ackergill raises the interesting possibility of an early (Pictish) ecclesiastical site in the area because
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
13
MacDonald identified it as derived from ON papi 'priest' [ 1977]. Waugh, however, suggests an alternative etymology, from ON pap 'breast' [1985:406].
DESCRIPTION OF OBJECT A slab of Caithness slate, incised on one side with the remains of a pair of Pictish symbols and an ogham inscription [see fig.]. Stone:
Grey slate
Dimensions:
1.20m tall x 0.60m wide x 0.08m thick
Condition:
The variousfragmentsof this slab have been cemented back together, but the surface of the upper part is missing, taking part of a symbol and perhaps part of the ogham with it. The remaining carving is clear and well-preserved.
This is a typical Class I Pictish symbol stone, an undressed stone of irregular shape (because of local geology a slab rather than the more familiar boulder of further south). Though the upper third of the surface has been lost, die original silhouette of the stone has been preserved. The suggestion by Stephens [IV 1901:35] that the slab bore runes was dismissed by Olsen [1954:197] after a personal inspection in 1911 in the company of runologist Haakon Shetelig.
DESCRIPTION OF CARVING Leaving aside the ogham, the Ackergill slab is a typical example of the Class I symbol stone. It displays the standard pair of symbols arranged so as to be almost touching. There can be no doubt, however, that the symbols and ogham are contemporary. Not only are the symbols placed slightly to the right to allow for the ogham, but the same carving technique, the pocking and smoothing so commonly employed for Class I stones, has been used for both. As the stone currently stands on display there is an uneven horizontal 'shelf across the carved face just above 'ground level', below which die stone is an inch or two thicker. This feature may have been original to the quarried slab, or may be a deliberate device designed to frame the carving or to give die upright slab greater stability in die ground.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
14
Much has been made of the unusual position of the ogham on this stone. Though its backwards-leaning diagonal alignment is unique, it preserves three important features of the normal arrangement: it is close to the left edge of the face, reads from the bottom up, and the reader's head is turned down to the left. All that would be needed for the inscription to be in line with the norm is for the lower end to be swung round to the left slightly. Macalister was quite wrong to maintain that the oghamist 'evidently held his model upside down when he copied it on to the stone1 [1940:214]. One possibility is that the artist required the ogham to come close to touching the symbols, just as it seems to have been necessary that they almost touch one another. Alternatively, since the stone curves sharply just after the end of the extant portion of ogham, a straight stem of this length placed parallel with the left edge would need to have been begun quite a bit lower than present, perhaps too low in relation to the symbols to have been either visible or aesthetically pleasing. This implies that the straightness of the stem and its proximity to the left edge overrode the requirement of true verticality. It is impossible to tell whether the unwillingness to curve the stem or place it along the much straighter right edge is to be taken as evidence of lack of confidence, and hence lack of familiarity with the use of the script, or the converse, immersion in a tradition and unwillingness to abandon its conventions for practical considerations. In the clash between Class I and ogham conventions the outcome is a compromise.
DESCRIPTION OF CARVING (SYMBOLS) In addition to the short ogham text, the Ackergill slab is carved with two of the standardized Pictish symbols - the remains of the fish and below it the so-called 'rectangle1 - two reasonably common symbols. The rectangle, which has been identified, with no real justification, as representing a book or book-satchel, is filled with spirals in Ultimate La Téne style. In addition to the Ackergill example, which for some reason he omits, Romilly Allen lists the appearance of the rectangle on ten other Class I stones: Sandness SHE, Firth ORK, South Ronaldsay ORK, Clynekirkton 1-2 SUT, Little Ferry Links 1 SUT, Benbecula, Grantown MOR, Old Deer ABD, Newbiggin Leslie ABD. To these we could add the two occurrences at Inchyra. The symbol occurs in four caves: Covesea MOR, East Wemyss (Jonathan's, West Doocot, and
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
15
Sloping Caves) FIF; and on a single Class II cross slab: Golspie SUT [ECMS:66]. The Ackergill example is typical of its class and belongs to the sub-type which differentiates between embellished upper and plainer lower halves.
Anderson identified the incised lines above the rectangle as the 'belly, ventral, pectoral and anal fins' of a fish. The 'tail* which appears to the right on the ECMS photograph is in fact a fracture in the damaged surface of the stone. The shape of the fins make it clear that the fish, in common with all other Class I fish (except Edderton which is anomalous in its general layout) faces right with the tail to the left. Allen, again omitting Ackergill, lists the appearance of fish on thefollowingClass I stones: Dunrobin SUT, Edderton ROS, Drumbuie 2 INV, Keith Hall ABD, Kintore 1 ABD, Percyclieu ABD, Rhynie 1 ABD, Stonehaven 2 KCD; on two caves: Covesea MOR, East Wemyss (Jonathan's Cave) FIF; and on five Class II cross-slabs: Ulbster CAT, Golspie SUT, Glamis 2 ANG, St Vigean's 1 ANG, Meigle 1 PER. The two symbols have a wide distribution weighted, especially in the case of the rectangle, to the north. They do not appear in combination on any other Class I symbol stone apart from Ackergill, but they do occur together on the Class II cross-slab from Golspie. In the latter case the rectangle is the most prominent symbol, the largest and placed at the top, and is paired with the 'beast'; the fish is smaller and lower and appears to be paired with either the dog/lion or the 'flower*.
DESCRIPTION OF OGHAM INSCRIPTION The stemstans cleanly at the first stroke and continues, in a slightly wavering line, to die fractured upper surface. The letter strokes overlay the stem, which sometimes disappears under them [see fig. for key to numbering]. As Padel pointed out, the last two strokes of 7 are unbroken by it. The vowel strokes (2, 5,8) are slighter and more closely spaced than the consonants, perhaps to enhance the distinction between these two groups already indicated by the difference in length. The strokes of the last two consonants (6, 7) appear thinner and more cramped than those of the first three (7, 5, 4). The strokes of the first consonant (7) in particular, appear broad and generously spaced. The increasing compression of strokes
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
16
and the curtailment of the first stroke of 8 because of the overhang of the last stroke of 7 suggests that the carver was running out of space. This implies the letters were carved straight onto the stone without a preliminary marking out. Padel has suggested that the slightly-less-than-expected slope on the R (7) was also a space-saving measure. This may be the case, but it should be noted that in some earlier inscriptions m-aicme letters slope less markedly than in later cases.
The stem stops c.60mm shon of the outer edge of the stone, but nothing of the original surface remains in the area beyond 8.
The rough fractured surface of this portion is markedly less fresh than the
brokensurface of the upper third of the slab, and it is possible that the vertical edges are original. If this is the case then the ogham has been carved right to the outer limit of the smooth surface and no further letters have been lost. The stem of the ackergill ogham has not been re-cut on top of the letter strokes.
If precedent were not sufficient to justify reading from the bottom up, the slope of the consonants proves it. All slope as expected except for 2, which is only slightly less than vertical. The letters are as follows [see fig. for key to numbering]: 1
Five strokes to the right of and perpendicular to the stem - N.
2
Four vowel strokes across the stem, sloping very slightly backwards - E.
3
A single stroke to the left of the stem, sloping backwards - H.
4
Three strokes to the left of the stem, sloping backwards - T. Groups 3 and 4 are carefully spaced.
5
Four vowel strokes perpendicularly across the stem - E.
6
Three strokes to the left of the stem, sloping backwards - T.
7
Five long strokes across the stem, sloping forwards - R.
8
All previous authorities have interpreted this group as five vowel strokes across the stem -1. The final stroke is thicker than the previous four and separated from them by a distance greater than that which separates the four from one another. Though the difference is slight, given the general compression towards the end of the inscription, it may well be telling. A third consideration is
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
17
that the upper (left-hand) portion of the final stroke appears to cross the stem only barely, certainly less than the previous four strokes. If the carver intended to differentiate two letters, the first must be E, and the second probably B. There seems insufficient room for the final stroke to have continued significantly across the stem (M), or to have been followed by a second stroke (L or G), but since the edge is damaged it is impossible to be categorical. Previous authorities may have been predisposed to read -I because of the frequency with which final (genitive) -I is encountered on the classic Irish pillars.
While -I is possible, other possibilities should be
entertained: the dating of Class I symbol stones is debated but is almost certaintly post-apocope. Which gives a reading:
NEHTETRI»
or
NEHTETRE(b + /m + )
FORM OF SCRIPT The form of the script of the Ackergill ogham is a simple Type Ila, notable for having vowel strokes which, though longer than notches, are still shorter than the strokes of the m-aicme The vowels occupy not more than the middle two quarters of the ogham band, the m-strokes occupy its full extent. The haicme strokes are well-sloped, the b-strokes a little less so, perhaps because they come at the very beginning where there is amply room. Successive letters of the samea/cme, e.g. 3 and 4> are well-spaced, but letters of different acimi are tightly packed, even with less space than between the component strokes of the letters in question. A notable feature of this inscription is the marked cramping of letters towards the end. The first letter is very generously space, almost over-generously, leaving only minimal space for the last few. Brandsbutt offers a fairly close comparison of script, though it has longer vowel-strokes, and b- and h-consonants which slope minimally if at all. Scoonie and St. Ninian's Isle are close too, though the former has less slope on its b- and h-aicmi and the latter has more explicitly-spaced letters.
INTERPRETATION OF TEXT The extant letters of the ogham text are clear and there is no doubt over the reading except for the final letter(s). Precedent would lead us to expect the eight or more letters to form a male personal name in the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
18
genitive. The single personal name formula is a common enough sub-type among Irish ogham inscriptions [McManus 1991:52]. Neither Nehtetri nor Nehtetri- suggests an otherwise attested personal name, but depending on how the word is segmented, possible known personal name elements can be identified.
Neht- may be compared with Formaston's Nehht-y Latheron's Netfu*)- and possibly with the (I)Neit(ateon the Gurness knife. Previous authorities have usually connected these with the well-attested Pictish name Nechtany or in less Gaelicized spelling Neithon. These are from *Netíonos [O'Rahilly 1946:368]. Since the -anf-on in these names is not a diminutive suffix, but an intrinsic part of the name, it is not strictly speaking detachable. It appears, however, that a false analogy may have been made with diminutives and the -an came to be considered detachable, cf. Talorc/Talorcen, Drust/Drostan.
In Irish tradition Necht appears as the name of the divine figure Nuadu Nechu though wrongly associated with Nechtarty it is in fact an epithet, the Olr. verbal adjective meaning 'propre, pur, bland 'clean, pure, white', from the root *nikto- 'lavé', 'washed* [Vendryes 1983:s.v.]. As a noun, Olr. necht means 'niece' but it is not attested as a personal name element. There is, of course, no certainty that the -H- in the Ackergill inscription represents l%l [see Introduction for discussion of transliteration of H]. If the spirant is very weak, Ackergill's EH may be equivalent to the diphthong -e/- in the Neitano of the Peebles inscription [Steer 1969]. Neht might be comparable with the Irish Neit listed in O'Mulconry's Glossary as nomen wri, and the name of the young poet NeidelNéde in the Accallam in dá Thuraid [Stokes 1905:14, 8]. Vendryes links this name with the common Olr. noun néity 'combau lutte\ [1983:s.v.] and the name of the Irish war god Néid mentioned in Connac's Glossary, compare with the Gaulish divine name Neto [Evans 1967:370], the war god of the Accitani [Vendryes 1960: s.v. nia]. The related element Netaoccurs in composition in such names as ogham NETTA-SLOGI (CIIC 109 > NadSlúaig) [McManus 1991:109-10 §6.15], and uncompounded in Olr. Niath, Nioth, gen. Neth [Stokes & Strachan 1903:274.38, 267.39, 269.34, 273.7]. These are associated with Olr. nia 'champion', 'hero' or its homonym 'sister's
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
19
son, nephew'.
After the first syllable, the segmentation becomes problematic. One possibility is Neht-Etri> with the second element comparable to MW Ethri < *Ettorigos [Williams 1937:cxvi], the latter attested as Ettorigi, with probably mistaken genitive formation, on the early sixth century inscribed pillar from Llanbabo, Anglesey [CIIC 318, Nash-Williams 1950:53-4, Jackson 1955:566]. This would imply a three-element proto-form *Neta-Etto-rigi. An alternative segmentation, Nehtet-ri, brings us back to Rhps's comparison of Nditet with the Uoret of the Drosten Stone at St Vigean's, and with the Namet and Morbet of the Kinglist [Anderson 1973:246-7]. The latter two, however, appear as second names and are perhaps best interpreted as epithets rather than names proper. The syllable -ettl-edd occurs in several of the Scottish oghams, apparently on the end of personal names, but its significance has not yet been satisfactorily explained. If Nehtet is to be taken as a unit, what remains is r/-, which recalls the very common Celtic personal name element 'king, ruler'; -rix -rig etc. in Continental Celtic [Evans 1967:243-9], r(h)i- and -n in Welsh and Breton [Lloyd 1888:51]» -ri in Irish [Uhlich 1993:294]. There are numerous epigraphic examples of names containing this element [McManus 1991:104 §6.7] and the forms are discussed in detail by Jackson [1955:624-8 §180].
Is the language of the inscription Gaelic or Brittonic, and is the case nominative or genitive ? The answer depends in pan on the reading of the last two letters. If the inscriptions ends with -ri this could be Irish nominative, or Brittonic nominative or genitive. There is certainly not room for the pre-apocope < *-rigas but the two strokes necessary for a final -g, giving the correct Irish genitive rig, would just about fit. Pursuing this possibility, the Ackergill text may represent a compounded Nehtet-ri(g) or an uncompounded Nehtet Ri(g) with the second element operating perhaps as a title, 'Nehtet the king*. There are no ogham examples of uncompounded forms of rig, but Macalister has proposed that the three line text of the fifth century pillar from Llanber, Merionethshire, CIIC 413 (CAELEXTI/MONEDO/RIGI), be read as three
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
20
separate words meaning 'Caelextus King of the Mountains'. Jackson, however, would take MONEDORIGI as a compound personal name [ 1955:355]. Several late Insular script inscriptions from Ireland contain the word rig, usually qualified by the name of his kingdom. Four refer to 'the king' in rig and give his name, but they are all late: the famous slab at Fahan Mura (CIIC 951), a tenth century cross at Clonmacnoise (CIIC 849) and two twelfth century crosses at Tuam (CIIC 522-3). The distinguished position of the Ackergill stone in relation to the rest of the cemetery might reflect the honoured burial of the local petty king, someone with status perhaps analogous to the ri of the Irish niath, but if -ri is to be taken simply as the final element of a compound name it is to be interpreted less literally. The etymological meaning of the name was unlikely to be foremost in the mind of those who used it, Evans produces a few Gaulish examples of names compounded with -/it belonging to women [1967:244],
If the alternative reading NEHTETREB is accepted, then the final TREB immediately demands comparison with Celtic *trebo- 'inhabitation, 'settlement', cognate with Latin tribus, and English thorp. The Welsh reflex tref' homestead, hamlet' is very common as a place-name element. Its Irish cognate is, however, extremely rare as a place-name [Watson 1926:357], though common enough as a noun, treb, meaning, perhaps under influence of the Latin, 'house, farm, holding, household, tribe, stock' [DIL]. The word must also have existed in Pictish since it occurs as a place-name element in eastern Scotland [Watson 1936:357-65; Nicolaisen 1972], notably as the second element in the territorial name Moray < *mori-treb 'sea(board) settlement' [Watson 1926:115-6]. It seems unlikely that a place-name would be recorded in an inscription of this period, but the element also occurs as an ethnic name, the Atrebates of southern England, and as the name of a deity, Contrebis recorded in inscriptions from Lancaster and Overborough [Rivet & Smith 1979:259].
DISCUSSION The Ackergill ogham stone must surely be seen in the context of a sphere of cultural and perhaps political
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
21
influence centred on Orkney, including Caithness and much of the northern pan of the Scottish Atlantic province. The rectangle symbol has a predominantly northern distribution and the form of the script employed at Ackergill is closest to that of the chattel oghams of the Atlantic province and the older Orkney ogham stones. Were it not for the ogham inscription, Ackergill 1 would be a typical Class I symbol stone. The symbols are accomplished renditions of two 'mainstream', though not particularly common, symbols. The uneven spacing of the letters and unusual angle of the stem gives the ogham a slightly scrappy appearance in comparison with the elegant spirals of the rectangle.
Ackergill has been compared
unfavourably to Brandsbutt. Padel thought that In general confidence of execution' the former 'does not at all match up to the other stone' [107]. Ackergill is certainly more modest in scale and more lightly incised, its symbols, however, are no less accomplished than those of Brandsbutt. The contrast is primarily between the two oghams themselves. The design element in the Brandsbutt inscription is very strong, the layout and spacing is careful and the over-all impression is calligraphic. Ackergill on the otherhand, seems closer to ordinary writing and the distinction between the two is one of greater informality as against greater monumentality.
Though it cannot be proved definitively, the association between Ackergill 1 and one of the graves in the dune seems likely. It is merely an assumption that the ogham referred to the individual buried underneath. Note that, in common with the Brandsbutt stone, Ackergill does not have the typical 'X MAQQ Y' formula. If the symbol pair is taken as referring similarly to the person buried, then the implication would be that the ogham was a 'transliteration' of the message of the symbols and that both represented the name of the deceased. Ross Samson's recent suggestion that the Class I symbol pair might represent the two elements in a di-thematic personal name is of interest here [1992]. Since neither the fish nor the 'rectangle' appear on any other ogham inscribed stones (except Golspie where it is one of eight symbols) there is no control by which to check a possible correlation between fish and Neht- and a 'rectangle', whatever it is supposed to represent, and -Etri. In fact, if, like the ogham, the symbols are to be read from the bottom up, then perhaps the correlation would be rectangle/Afecfcr, fish/Em*.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ACKERGILL
22
Since the vast majority of Class I symbol stones do not bear ogham inscriptions, one might wonder why one was required at Ackergill. The bi-lingual ogham pillars of Wales spring to mind as an example of the same text conveyed to two audiences, one speaking British, the other Irish, through two scripts side by side. Might the ogham text at Ackergill have been intended for an audience who would not have understood the symbols ? If so who might these people have been ? Could this be evidence for small numbers of people of Irish origin settled amongst the indigenous population of the far north (knowledge of ogham betrays influence which is, at least ultimately, from Ireland) ? Alternatively, the ogham might be considered in some way supplementary to the message of the symbols, expressing, perhaps, something which could not be conveyed by the symbol system alone. Either way, the prominent position of the stone atop a mound by the shore next to what was probably a coastal routeway suggests the message was supposed to be read by passers-by.
BIBLIOGRAPHY
OS Record card - ND 35 SW 12
Anderson 1897:296 fig.2; Rh?s 1898:370; ECMS 28-9 ill.; RCAHMS 191 Ib: 190-1 No.587; Edwards 1926:179; Macalister 1940:214, fig.7; Diack 1944; K H Jackson 1955:141; padel 1972:107-110; CloseBrooks & Stevenson 1982:27 ill.
Site Only:
Edwards 1926, 1927; Ashmore 1980:348-9, 351-2; Close-Brooks 1980; Close-Brooks 1984:97-9 ill.; Batey 1991:49, 53-7.
PREVIOUS READINGS Rh?s, ECMS, Macalister, Padel
NEHTETRI
N
0
i 0
i........ i i
i
i
30 m«tr«i i
i i
i
'i • r
i
i
| 100
f««t
ACKERGILL - Plan of cemetery [CLOSE-BROOKS 1984 fig.5.6]
0>
i en
1 CO
"8 *c .S "T
E cd O i
C/3
S
o <=* w u
I 1
1
I
ACKERGILL - Schematic representation of the ogham (key to numbering)
23
ALTYRE
DISCOVERY This tall Class in cross-slab currently stands surrounded by a wooden fence in a field by the drive from Altyre House to Forres, in the parish of Rafford, Moray [NGR NJ 0391 5537]. This is not, however, its original location and there is some confusion over where it had been prior to this. At the beginning of the nineteenth century, probably c.1820, the stone was brought to Altyre from somewhere in the Laigh of Moray by the Laird, Major Cumming Bruce of Dunphail, or his brother [Calder & Jackson 1957:246]. In the main text of ECMS, Romilly Allen is no more specific about the earlier location than 'Dufftis parish', but in afootnotequotes the Burghead antiquarian Hugh W. Young, as saying it came from the college field at the village of Roseisle, there is no further authority to support this tradition [136 n.l]. Norman Atkinson, Curator of Angus Museums, has suggested1 that the stone may, in fact, have come from the Ogstoun. The site now occupied by the 'Michael Kirk' of Ogstoun was prominent in the later Middle Ages, as indicated by the surviving sculpture there, but fell out of use in the early modern period. According to Mr Atkinson the old kirk became a lair for the Gordon family in 1705 when the parish was united with Kinneddar and the congregation moved to Drainie. The link with Altyre was made in 1795 when the Cummings of Altyre married into the family of Gordon of Gordonstoun. Calder surmised that the reasonforthe slab's transference was its suitability as a rubbing postforcattle [Calder & Jackson 246]. While it was performing this function at the time Romilly Allen wrote, it seem far more likely that the original impulse to move it was a romantic or antiquarian one. Atkinson's hypothesis is certainly plausible, but, since Calder claims a co-ordinated search for additional documentary information proved fruitless, it is unlikely that it could ever be proved. But how strongly we must dissent from Calder's opinion that in its current 'charmingly bucolic context it would be churlish to complain that no record was kept of the provenance of the pillar, or of its siting or setting' [Calder & Jackson 247]!
1
In his lecture The Picts in Moray*, to the conference of the Pictish Arts Society, Aberdeen, 28 November 1992.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
24
Though the Altyre cross-slab has been knownfora considerable time, only comparatively recently was the ogham inscription noticed. Without recognizing them, William Rhind may have been referring to the ogham letters when he commented that 'there appear to be faint marks of Runic knots on this stone, or other carvings1 [1839J (here frunicf does not mean written in the Germanic rune-alphabet, but rather, following contemporary usage, 'in the style of Celtic or Northumbrian interlace' [Chambers Dictionary; on antiquarian use of word 'rune' see Page 1973:2]. In quoting Rhind's comments, Stuart added 'the marks of ornament seem now to have disappeared' [1856]. Romilly Allen thought it 'doubtful' that there had been any ornament; even if there had, 'none ... now remains' [loc. cit.]. The first intimation of the ogham inscription comes in the form of a hand-written note in the NMS Library's copy of ECMS by J. Graham Callander, Keeper from 1919-38: 'this stone has an ogham inscription, much defaced. J . G . C . It is not known how this fact had come to Callander's attention. Having seen this note, Callander's postwar successor R.B.K. Stevenson visited Altyre to confirm the observation. At his behest. Sir William Calder, in his capacity as school governor, called up 'a task force of masters and boys' from nearby Gordonstoun School to 'have the thick coating of lichen which concealed and preserved the inscription removed, and beat out an impression of it' [Calder & Jackson 247]. On the basis of this rubbing Kenneth Jackson attempted a reading and interpretation of the ogham, published, with a sketch of the monument in the Proceedings of the Society of Antiquaries of Scotland [Calder & Jackson 1957]. Sir William published an account of the Altyre ogham in die Inverness Courier for 23 August 1957 [Wainwright 1961.7n3]. To my knowledge, no photograph of this monument has been published.
In the early 1990s the artist Marianna Lines visited the stone and made an organic 'impression', i.e. plantbased rubbing on cotton (the image subsequently marketed in silk as a neck-tie!). When I visited the stone in August 1992 I was fortunate in having with me a copy of this rubbing which greatly facilitated my reading of the stone, which is now thickly covered in lichen. On the better preserved of the two faces die cross-shaft is decorated with a thin vertical line 15-20mm in from each long edge. Despite the lichen this
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
25
is clearly visible on both edges. The left-hand line is crossed by short perpendicular and oblique strokes, carved to the same proportions as the vertical line. Because of weathering and lichen these are visible only in short, disjointed stretches, but the few that are clear are unmistakably artificial and deliberate. They have not been noted before. The neatness of the extant lines, and their disparate gradients, suggest they are not the remains of abstract geometric ornament. I believe that they are, in fact, the remains of another ogham inscription. Confirmation of this will come, however, only when the lichen has been removed (since the cross-slab is known not to be in its original location there can be no excuse for not moving it indoors, given the severe weathering the stone has already suffered this must surely be an urgent priority).
The name Altyre, stressed on the first syllable, has been consistently applied to this stone, and since nothing definite is known about its earlier position the label should be kept, even though it is known not to be the stone's original location. There has been a tendency in the older literature to confuse this stone with the much better known 'Sueno's Stone' from Forres about three miles away (NJ 046 595) [as pointed out on OS Record Card, also Calder & Jackson 247].
LOCALITY The fertile soil and favourable micro-climate made the Laigh of Moray as attractive to the prehistoric and proto-historic farmer as it remains today [Jones etal. 1993:51; see map]. The combination of rich natural resources and seclusion from the power-bases of the south fostered an independence which was not finally reigned in by central authority till the end of the twelfth century [Shepherd 1993:75], In the post-Pictish period, Norse sources refer to the mormaers of Moray as 'kings of Scots' [Crawford 1987:64], and the Irish Annals label an early eleventh century mormaer of Moray as ri Alban [Cowan 1993:119]. Moray has always been open to influence from the west through die Great Glen. Gaels appear to have expanded along this route during die Pictish period for in the tenth and eleventh century the rulers of Moray claimed
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
26
descent from the Cenél Loairn [Cowan 1993:119]. The use of ogham at Altyre and at Brodie, a few miles to the west, is perhaps to be placed in this context of increasing Gaelic influence in late Pictish Moray. For a general discussion of Pictish Moray see Shepherd 1993 and refs. [see also map].
The precise original location of the Altyre slab is unknown, but it appears to have come from the western half of the portion of land which juts out into the North Sea between Burghead and Lossiemouth. Jones et al. [1993] have shown how this area was cut off from the rest of Moray by a sweep of dune and moss 'unavailableforearly settlement' and in the east, the large Loch of Spynie, now drained [see map]. This sandstone ridge of a few square miles of habitable land has been described as 'the core of Pictish Moray' [Shepherd 1993:85], 'an island of Royal power with a range of inter-related sites', viz. Burghead, Covesea, and Kinneddar [Shepherd 83]. The enormous early Pictish promontory fort of Burghead, was almost certainly the major royal centre of the region. The series of bull plaques from Burghead are justly famous; sometimes overlooked is the corner-post shrine of late seventh or early eighth century date [Thomas 1971:152], otherfragmentsof ecclesiastical sculpture [ECMS 137-42], and the ninth century Anglo-Saxon blast horn [Graham-Campbell 1973]. The cliff cave at Covesea has Pictish symbols carved on its walls and simple crosses which may date back to the Pictish period [ECMS 129-31]. The most prominent ecclesiastical site in the region is, however, Kinneddar (Drainie) a few miles further east. Twenty-five fragments of sculpture have been recovered from later structures at this site, including one extremely fine depiction of David rending the jaws of the lion which establishes an intimate link with the St. Andrews sarcophagus and the Nigg slab [ECMS 142-9]. The evidence points to Kinneddar being a major monastic site from the eighth century [Shepherd 1993].
If Young was correct in his claim that the Altyre slab had come from Roseisle then it cannot be disassociated from what appears to have been a Pictish long cist cemetery at Easterton of Roseisle. Romilly Allen gives an account of the circumstances leading to the burials and the Class I Pictish symbol
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
27
stone found re-used in one of the graves [ECMS 124-6]. Allen also stresses the very different toponomy of the area before the draining of the Loch of Spynie, suggesting that some of the little hillocks around Roseisle may formerly have been islands. If Roseisle was the original home of the Altyre slab it may have stood on the route towards Burghead.
DESCRIPTION OF OBJECT A tall, thin, Class III cross-slab, sculpted in relief on two faces with crosses, having one ogham inscription incised vertically up one narrow edge and possibly another up the side of one cross-shaft [see fig.]. Stone:
Grey sandstone [ECMS]
Dimensions:
Approx. 3.3m tall from current ground level, width 2'lCr (0.86m), thickness T (0.18m) [Calder 1957]
Condition:
Severely weathered and covered in lichen. While some portions of the ogham are comparatively clear, others are very worn and doubtful. There may be a section missing from the beginning.
The nearby Sueno's stone is carved from a block of grey sandstone over 20' (6.10m) high. Southwick gives its probable source as 'die cliffs of Permo-Triassic rocks between Cummingston and Covesea' [1981:3]. Roseisle is only a mile or two inland from Cummingston and these cliffs seem a very likely source for the Altyre pillar too. Stuart gives the total height of the monument as 15' (4.57m), but today only slightly over IT (3.35m) now appears above ground level. His plate shows die cross complete but it is hard to credit a loss of over a metre in only fifty years. Though the top is broken, the damage probably pre-dates the stone's move and the disparity may as easily reflect inaccurate measurement by Stuart rather than a loss since die 1850s. Taking the better preserved of the two faces as the 'front', the ogham occupies most of the lower half of the narrow left edge. Much of the upper half is severely weadiered but the termination of the ogham is clearly visible so we may be assured that nothing has been lost at the end. The beginning, however, is imperfect. At the base the stone is severely 'waisted' due to
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
28
splash-back, and its girth has been reduced by several centimetres. Any carving in this area has been lost. There is no trace of any ogham letters on the right edge, but much of its surface has weathered away. There is no basis forjudging whether the ogham was part of the original design of the sculpture, or added later. Insufficient remains to compare the carving technique with the rest of the stone, but the ogham is formally carved and is no graffito. Its position on the left edge is the traditional one. There seems no reason to doubt that it is primary.
DESCRIPTION OF CARVING - DECORATION The cross on the reverse is cut in very shallow relief and only its venical shaft survives. (Calder refers to a 'horse-shoe shaped marking, low down on the back of the stone* which he concludes 'may not be ancient1 [1957:247]. I could see no trace of it). The cross on the front, however, stands 15mm proud of the background and much of its outline is still preserved. It is a square-armed cross with double square hollows for arm-pits [Romilly Allen's shape 98a, Cramp's A2 (1984)]. The shaft is of the same thickness as the arms and the stepped hollows in the angles of the arms create the impression of a square panel at the cross-head. The form was a popular one on Pictish cross-slabs, particularly among later slabs and is seen elsewhere north of the Mounth at Skinnet CAT, Edderton and Nigg ROS. The Altyre cross-head is tiny in comparison with its exceedingly long shaft (extending almost the whole height of the slab), a feature even more pronounced than on the nearby Sueno's Stone.
Romilly Allen doubted that the cross was ever decorated. Since the line which follows the entire perimeter of the cross about l5-20mm in from the edge is clearly visible weathering cannot be held responsible for substantial losses. The background is too pitted to rule out the possibility that it might have been decorated with geometric or figurative work, but there is no trace of any such carving. If the slab was indeed plain it would be unique in such a grand monument.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
29
DESCRIPTION OF CARVING - OGHAM INSCRIPTION Face If the incised line on thefrontface is indeed the stem of an ogham inscription then it performs this function for only part of its length. Its primary function is to delineate the entire outline of the cross, suggesting the ogham, if that is indeed what it is, may have been a secondary addition. The extant carving is very faint and weathered, for much of its length it is too densely encrusted with lichen to be made out. A single perpendicular cross-stroke is clearly visible, and a little higher a group of at least four m-aicme strokes. The surviving strokes are too disjointed, however, to allow reconstruction of the text. The 'ogham* on the front seems cut in a slightly finer script than that used on the side, but the differential wear patterns mitigate against direct comparison. If the face ogham is accepted as genuine, and if it is judged secondary, this may imply the edge ogham is also a later addition. The edge ogham is already one of the longer extant texts. If an ogham of similar length survived on the face also, then the Altyre texts would approach nearby Brodie in total length.
Edge At the base, the whole of the edge has been lost to a depth of several centimetres. The inscription is visible at the lowest point on the intact surface, about 600mmfromthe current ground level, and it is clear that lettering has been lost at the beginning. Contrary to Jackson's implication and Padel's assertion, the text is not complete as it stands. The stem is visible for approximately 1250mm and stops cleanly with the last letter. Above this the surface is intact for about 1.5m (5r). How much has been lost below this depends on whether or not the bottom of the slab is intact, how the current ground level compares with the original siting. It would perhaps be surprising if the stone was re-erected at precisely the depth to which it was originally sunk. At present it sits in the ground directly, but like other tall later Pictish slabs, such as Sueno's Stone, or Aberlemno 3 ANG, it may well have had a separate stone base. An inscription is unlikely to have begun at ground level, but another 300mm of text, at least five or six letters, could
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
30
quite happily be accommodated.
The stone has weathered along its bedding planes to produce a number of vertical fractures. On the right edge this has resulted in what looks like the remains of a stem-line, though it cannot be since the original surface has been lost entirely. On the left edge there are fissures which Jackson, presumably on the basis of Calder's nibbing and drawing rather than personal inspection of the stone, interpreted as guide-lines for the outer edges of the ogham band. As Padel points out, not only are these irregular, but the letters do not keep to them; they are entirely natural. Calder mentions 'a considerable number of punched cavities or dots, which appear to bear no relation to the characters and are probably due to the exuberance of an amateur new to the chisel' [247]. A more likely explanation is that these are due simply to the weathering of the sedimentary rock.
The carved surface is too weathered and lichen-encrusted for a detailed examination of carving technique, but it appears that the usual pock-and-smooth method has been employed. The stem is more substantial than the strokes, both wider and deeper. The less weathered of the letter-strokes are quite thin, and even when fresh, are unlikely to have been deep. The inscription is, however, very competently and evenly cut in a formal measured style.
The following description is based primarily on personal observation but full consideration was taken of Calder's drawing (from the nibbing), made after lichen had been removed and, as he claims, with an objectivity 'guaranteed by the absence of any temptation to theorise about Picrish words or Ogamic forms' [247], There can be no doubt that the inscription reads from the bottom up, as one would expect. The b- and h-aicme strokes slope consistently for this direction of reading. The stem emerges from the weathered surface and after a short distance the letters begin, as follows [see fig. for key to numbering]: 1
A single long stroke, perpendicularly across the stem - A. There is sufficient clear space before
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
31
this letter to be reasonably certain that it is a singleton rather than the final stroke of a longer group. 2
A single long oblique stroke across the stem sloping forwards - M.
3
As above - M. These two letters are very clearly spaced and there is no doubt that they are to be taken separately.
4
A single long stroke, perpendicularly across the stem - A.
5
Five short strokes to the left of the stem, sloping backwards - Q. These are of equal length and neady and closely spaced.
6
As above, - Q. There is a clear gap between these two letters, easily enough for two more strokes, so there can be no doubt that they are to be taken as two groups of five.
7
There is a fairly clear stroke to the left of the stem, sloping slightly backwards, which is followed by a badly weathered patch where the carving is difficult to make out. In Calder's drawing, the stroke is perpendicular. Jackson thought it 'apparently an A', but conceded that 'it is not very clear, and is scarcely traceable below the base-line*. He is right that its perpendicularity would be sufficient to identify it as an A (since the inscription is otherwise so consistent regarding slope), but it is not clear to me that it is perpendicular. The shadow on the opposite side of the stem, is too high for it to be the right distal tip of a perpendicular vowel-stroke, and the slope would be in the wrong direction for an m-stroke. On balance, therefore, the first stroke of 7 is most likely an h-stroke. Calder has an obscure patch immediately following this, then two oblique strokes sloping backwards the first crossing the stem, the second not. Jackson was forced to posit an error and read either AH or D. There is a shadow to the right of the stem, but it is faint, and may not be a proper stroke. The carving to the left is more definite. Padel thought the obscure patch may have contained two h-strokes, a hypothesis tentatively supported by Marianna Lines's nibbing. The next twohe thought 'almost pan of the E-forfid [i.e. x ] but not quite', or perhaps two more h-strokes with the first crossing the stem only accidentally. Taken in conjunction with
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
32
the previous two doubtful strokes and the first clear one, this group could be read, Padel suggested, as a third Q. This is plausible, and though the first two strokes are rather cramped the letter as a whole occupies only 5mm less of the stem than the previous Q. The weathering is such that it is difficult to judge how wide the strokes were originally. From Lines's rubbing the third and fourth strokes appear more widely-spaced than the others (TD), but since the gap between the three previous letters is at least twice as big in each case it seems more like that 7 is a single group. The sloping is against the shadow to the right being a continuation of the fourth stroke. The fifth stroke is so close it must have been part of the same letter, yet if it were produced across the stem it would collide with the single stroke of 8. On balance, the most likely explanation, supported by the slope of the clearer strokes, is a single group of five h-aicme strokes - Q. Given the growth of lichen and the poor condition of the visible carving the other suggested readings, especially Padel's T + x , cannot be ruled out. C is perhaps less likely, but not impossible. 8
The next letter, a single long stroke, perpendicularly across the stem, is clearer - A. Its unambiguous verticality relative to the stem, reinforces the argument against the preceding group incorporating vowel-strokes.
9
Two short strokes to therightof the stem, sloping slightly forward - L. All three authorities have read 9 and 10 as two pairs, there may, however be a fifth stroke between them making a single group - N.
10
As above, but with a steeper gradient - L. Between this letter and the next there is the shadow of a possible stroke, perpendicularly across the stem (A), it is, however, very short, and far less substantial than £. The letters in this section are very well spaced so its admittance would not cause undue congestion, but the whole is so worn that the observation is tentative.
11
A single long oblique stroke, sloping forward across the stem - M.
12
The next 220mm is even more badly weathered than 7, and really all that can be detected are disjointed portions of strokes either side of the stem, which itself is not clear. Immediately
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
33
following 11 there is a stroke to the right of, and perpendicular to, the stem. It appears to have a small serif at its outer end, which may be nothing more than weathering. If it were projected across the stem it would immediately collide with 11. Padel read four strokes to the left of the stem, the first more or less perpendicular, the rest sloping forward in parallel with each other. He noted a left-hand serif on the second and an open right-hand one on the fourth, traces of this are just visible on Lines's rubbing, but again it may be wear. Calder had three strokes at this point, the middle straight and sloping forward the outer pair with a dog-leg curve to left and right respectively. Jackson interpreted this as V, Padel suggested BL. 13
The next section is long enough for one or two letters but the whole thing is so worn that little can be discerned at all. There could be about four b-strokes (S), but this is little more than a guess. Jackson read BV, which is possible, but, as Padel said Very obscure'.
14
Faint traces of a single oblique stroke, sloping forward across the stem - M.
15
A degree of clarity returns with a single long stroke, perpendicularly across the stem - A. I see no trace of Jackson's doubtful second A.
16
A single stroke to the left of the stem, sloping backwards - H.
17
As above - H. This stroke could be taken together with the preceding one to read D, but at 20mm they are rather widely spaced for this.
18
Five long oblique strokes across the stem, sloping forwards, closely spaced - R. The entire group is 50mm wide.
19
The next two groups are confused because of the weathering of the surface. The spacing and gradient is such that it is difficult to know how many groups there are and where they are to be divided. From Lines's rubbing, there appears to be a stroke perpendicularly across the stem. If it is a separate stroke (A), it would be very close to the following group. The next six strokes are more or less parallel with one another and slope forward across the stem. The first of these oblique strokes appears to be very long to the right of the stem, but this may be an illusion of
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
34
wear. The third and fourth oblique strokes are not visible to the right of the stem, but the area is worn and they are so closely spaced to the others that they cannot be a separate h-aicme letter. The fifth stroke now appears more substantial than the preceding ones, though it is equally-spaced from them. The sixth stroke, almost identical to die fifth, follows at a slightly greater distance. Immediately beyond it, to the right of the stem there is a shadow. Calder represents this sixth stroke as perpendicular, presumably taking the shadow as the right half of the stroke. Calder has only six strokes in total for this group, he appears to omit the small perpendicular stroke at the beginning. For 18119 Jackson read RR 'pretty clear on the whole', then A. Padel commented that 19 which he read as R seemed 'rather careless' and the strokes not strictly parallel. It is really only the first which is out of alignment, however. If it is not a separate letter the lack of slope may be a deliberate way of differentiating two successive letters of the same aicme, a device for which there is widespread precedent. If this perpendicular stroke is an illusion then 18119 is unproblematically RRA. If it is a separate A, then the group is a cramped RARA, or an implausibly spaced RRAA or RRO. There are no unambiguous examples of double vowels on Scottish ogham inscriptions (with the possible exception of Auquhollie), though in Archaic Irish manuscript orthography vowels were occasionally doubled to show length [Thunieysen 1946:20-1 §27). 20
From here to the end the stem is clearly visible. Unfortunately the first group of strokes is uncertain. There appear to be four strokes to the right of the stem sloping forward (S), die first stroke, however, is damaged, and the second is extremely long, projecting across the stem and, apparently, terminating in a small curl at each distal tip. Jackson interpreted the portion to the left as being above the third stroke, though not continuing in a straight line. He doubted that it was intentional.
Padel considered it 'certainly due to weathering', and this is the most
straightforward explanation. 21
Four strokes to the right of the stem, sloping forward - S. This group is clear and unambiguous.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
22
ALTYRE
35
Three long oblique strokes, sloping forward across the stem. Only the third is to any extent certain, but the spacing is such that there must have been at least two and probably three strokes. The third appears to have little flaps at its distal ends and calls to mind the 'integration sign* character seen at Latheron, Lochgoilhead, Birsay. Calder figures all three strokes as /-shaped curves, which is reminiscent of the five retrograde /s of Bressay. Jackson and Padel take this group as a vowel, which, despite the slope, is more likely than the third member of the m-aicme /gw/ (which is virtually unprecedented in epigraphic ogham).
23
Two strokes to the left of the stem, sloping backwards - D. From here to the end the carving is very clear.
24
As above - D. The letters 24 and 25 are clearly spaced.
25
Four strokes to the left of the stem, sloping very slightly backwards - S.
Which gives the following, very tentative, reading: -]AMMAQQ(q/t x /c)A(ll/n)MvsM AHHR( /a)R(a/aa/o)sSuDDS
FORM OF SCRIPT The Altyre ogham is cut in a rather simple Ila or lib form of the script, the letters very generously spaced along a stem-line, with long straight vowel-strokes occupying up to the whole width of the ogham band, and all consonants well-sloped in the expected direction. The stem stops cleanly with the last letter. Starting cleanly with the first is a feature seen, for instance,at Brandsbutt, but so often the end of the stem is not intact and there are no other examples in Scotland of the stem ending like this. The general deterioration of the carving and the particularly bad weathering in two key stretches makes it difficult to determine whether apparent irregularities are original, or merely illusions due to wear. To take one example, the well-preserved letters are of standard length, but there are a few strokes which appear considerably longer than their neighbours, for instance 18y 19, and 20. These, however, are restricted to the most severely worn sections. There are three possibly non-standard characters but each is in a patch
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
36
of wear and very doubtful:
Character 12:
The stone is so worn at this point that it would be wrong to make too much of the
supposed serifs on two of the b-strokes of this letter. They may well be not original, and their form is completely unprecedented. None of the other letters have any trace of a bind stroke.
Character 20:
Once again, the stone is so worn at this point that it is not at all clear that the apparent
flourishes at the distal tips of the second stroke of this letter are original. The other strokes are not visible at all on the left of the stem, and on the right are markedly shorter than the second. There is precedent for a /-stroke in the middle of a group of five straight strokes, but in both cases these are m-aicme strokes.
Character 22:
Jackson said that the oghamist had carved these three strokes with curved lines rather than
straight. The extant carvings are fluid, unlike other instances of /-letters which are straight with flaps or curls. Elsewhere, the /-stroke appears as a singleton, or as the central member of a group of otherwise straight strokes. Only at Bressay is there are group consisting entirely of /-strokes. The phonetic value of such a letter is unclear. Here it may be some kind of modified vowel (U), [for further discussion of characters similar to 20 and 22, see Latheron].
The generous spacing between letters is a marked feature of this inscription; some letters have sufficient roomforup to two additional strokes between them. The substantial gaps between the intact letters must be a guide when interpreting the confused sections. It would be wrong to read too much into subtle difference in spacing between strokes and thus the most obvious interpretation is in each case the likeliest.
The component strokes of the b- and h-aicme consonants are quite short and widely-spaced. These letters thus have a squarish aspect typical of more developed forms of the script (some lib and most lie). The
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
37
closest parallel is the script of the St. Ninian's Isle slab which is well-spaced, and squarish, but lacks supplementary letters and developed features. Other Type Ila oghams are more closely-spaced and don't slope, while spaced and sloping oghams tend to have supplementary characters. Buckquoy is similar to Altyre, though it has a single forfid. If the text of St. Ninian's Isle is correctly interpreted it may provide linguistic evidence for a date in the seventh century, Buckquoy is probably eighth century. Both of these seem rather early, however, for Altyre which, on an historic grounds is unlikely to be earlier than the late eighth century, and is more likely ninth.
INTERPRETATION OF TEXT The main obstacle to a satisfactory interpretation of the Altyre ogham is the uncertainty over the weathered sections of text. The intact portions are very clear, but where there is doubt it is considerable. If some of the doubtful characters are supplementary letters, then we have the "dditional problem of how to transliterate ///, for instance. The appropriate transliteration of H is discussed in the Introduction where it is suggested that it may have the value /x/. The transliteration of V is also discussed in the Introduction: in Gaelic texts it is to be transliterated /w/ or HI depending on date, in Picdsh /w/ regardless.
Since no word division is indicated it is left to us to segment the text. Letters 3-6 can be isolated straight away as MAQQ, a form of Irish MAQ(Q)I 'son', appearing elsewhere in Scottish ogham at Fonnaston and Latheron. The use of QQ rather than CC or C is unhistorical, but is standard in Scottish ogbams of the mainland and Northern Isles. The balance of the text from 7 on is surely too long to be a single name and requires funher segmentation. The sequence MAHH /maxx/ echoes the earlier MAQQ /mak(k)/, perhaps the text is of the form 'X son of Y son of Z\ but why l%l rather than /k/ ? Thefinalconsonant of mac(c) is never lenited. This would leave RRASSUDDS as a separate word, which is awkward since doubleconsonants were generally avoided in word-initial position [Harvey 1987]. MAHHRRASSUDDS might be an Irish name of the Mac-N. type (cf. Mac-Dara, Mac-Táil, Mac-Cáirthind)y but this would still not
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
38
explain the lenited /x/. RRASSUDDS calls to mind the Resad of the St. Ninian's Isle chape, but is otherwise not recognisable as any attested personal name (though possibly related to Brittonic Ris > Rhys ?). The final -S is as problematic here as at Brandsbutt, Cunningsburgh 1, Bressay, Guroess, and Lunnasting. In all but the last case (where it is word final), the -S is at the very end of the text. In the last three cases it comes at the end of a sequence - vowel, dental stop, S (though this may be merely coincidental).
Turning to the beginning of the text, Jackson interpreted Am as an unattested Pictish personal name, rejecting Olr. am 'I am', 'not indeed on grammatical grounds but because such a formula would be entirely without parallel in Ogam inscriptions anywhere* [250]. The reading, however, is not Am, but -am. That survives is the end of what was a personal name of as many as half-a-dozen further letters. Personal names ending in -am are few in both Gaelic and Brittonic sources, but the Pictish King-List has a king Galam immediately before Bride mac Meilochon, AU has him die in 579 [Anderson 1973:130]. One of the donors in a charter dated c. 738-40 in the Book of Llandaff was a certain Riatam (Rhiadaf) [Davies 1979:183].
The section immediately following the MAQQ is almost impenetrably obscure: (q/tx/c)A(ll/n)Mvs.
If 7
is a Q at the beginning of the next word then it is the only clear example in Scotland of a Q outside the formula word MAQQ. This is sufficient to throw some doubt on the interpretation. It might be a third Q on the end of MAQQQ, leaving ALL- the opening of the Golspie ogham, but a third Q would be entirely superfluous. If the first three strokes of 7 are T then we have the Celtic Tall [see Formaston for discussion of Pictish personal names in Tal-]. This, however, would leave the indistinct lines between 8 and 9 unaccounted for. Callm- might be related to Olr. calma 'strong, brave, valiant* [DH]% but the latter is not attested as a personal name element. Jackson suggested the Irish adjective doll 'blind' in an otherwise unattested usage as a personal name, but such a reading leaves too many strokes unaccounted for.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
39
< Jackson concluded by saying that, apart from 'Am son of Dall' 'the inscription is impenetrably obscure, and will remain so until a future Ventris or Chadwick provides the key to the "Pictish" Ogams' [250]. The problem, however, is not so much the lack of a 'key* but the highly weathered state of the carving.
DISCUSSION Even at its current height of 3.3m, Altyre is the second tallest sculptured slab from early medieval Scotland, shorter only than the giant Sueno's Stone (6.10m). It is over a metre taller than the next tallest Class III slab, the Farr Stone CAT (2.29m), though there are a handful of Class II slab that are within 0.15-0.30m shorter. The northern distribution of these tall thin slabs is marked, though there are a few from south of the Mounth, including the Fowlis Wester slab PER (3.15m) which has the same A2 form of cross. That such an enormous slab, weighing several tons, could have been moved a distance of more than 10 miles is a salutary warning to those who would make too much of the current position of sculptured stone monuments.
Since we don't know the precise original location of the Altyre slab we are severely hampered in attempting to understand its significance. Its general context is a late Pictish Moray increasingly drawn into Gaeldom, more specifically the environs of the major secular power centre of Burghead and die important ecclesiastical site at Kinneddar. Too little of the text survives to make clear the identification of the language of the inscription, whether Pictish or Irish. Though the form of the monument suggests a date in the later eighth or ninth century the ogham is written in a simple form of the script (not withstanding doubts over possible supplementary letters), and perpetuates the old convention of placing the ogham up the left narrow edge. It is striking that the Altyre and Brodie ogham-slabs should be so close to one another and yet so far from other ogham-inscribed stones. To the south, across the Grampian mountain, the nearest ogham is the Newton Stone. To the north, across the Moray and Dornoch Firths, the nearest is at Golspie.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ALTYRE
40
BIBLIOGRAPHY OS Record card - NJ05NW34
Rhind 1839 (stone only, no mention of ogham); Stuart 1856 pi. 114 (no mention of ogham); ECMS 136 (no mention of ogham); Calder & Jackson 1957; Padel 1972:49-52.
PREVIOUS READINGS
Edge: Calder & Jackson1 Padel
(/d)AMMAQQAahALLMVBVMaAHHRRASSUDDS (/t/q)AMMA(q/bl)Qa(a/l/v)L(lv/v)( /b/a)blmaHHRRASSUDDS
Face: None
1 The readings published p.249 omit the final A, but, as Padel points out, this is a misprint, the letter is referred to in the text of the preceding paragraph.
Location map, Moray and Easter Ross showing extent of areas of prehistoric and proto-historic settlement [Jones et ah 1993:48 ffg.3.1]
ROSEISLE (original location of Al tyre slab?) BURGHEAD
m
Class I Class II Class lit
Stores
AP Barrows Forts (undaUd) symbols Covssea finds (S3
15km
Land abova 150m Placa Namas
15ml
\
Prat
PICTISH MORAY [after Shepherd 1993:77 fig, 4.2]
I Pit
— Carddan
«**$ m\ /í/ An* -y^*< * • * " it'- \ 4i
ALTYRE - Ogham-inscribed cross and detail of inscription [Calder & Jackson 1957 figs. 14-15] %
(key to numbering)
19 18
Vfi*.:
E^
ALTYRE - Ogham inscription, based on impression by Marianna Lines (plant matter on cotton) [reduced]
12
41
AUQUHOLLIE
DISCOVERY This pillar, a List B Scheduled Monument, stands 380m north-west of Nether Auquhollie farmstead, in the parish of Fetteresso, 5 miles north-west of Stonehaven, Kincardineshire (NGR NO 8232 9079). The stone is of such massive proportions that it is unlikely to have been casually moved far, if at all, from its original position. It is visually unprepossessing and is unlikely to have attracted the kind of Romantic attention which led to the moving of the Altyre cross-slab. The presence of the carving was first pointed out in 1886 by Rev. John G. Michie, Minister of Dinnet, Aberdeenshire [Southesk 1886:37]. The Earl of Southesk visited it on May Day of that year and gave an account to the Society of Antiquaries of Scotland. If, as claimed, the Auquhollie stone really was part of a circle, perhaps it was during die alleged removal of the other stones, shortly before 1886, that the inscription was first noticed. Rhys complained of the thick lichen which covered the stone when he visited it [1892:271], which perhaps goes some way to explaining why the inscription had gone undetected.
It was the Earl of Southesk who reported, without naming a source, that the stone was 'said to have formed part of circle recently removed* [1886:37]. Though no details are given, a local account of a near contemporary event must be given a large measure of credence, though one might wonder why a single pillar was left standing while all the rest were taken away. It is doubtful it would be saved for the sake of its inscription; carving far more noticeable and accomplished has not prevented the nineteenth and twentieth century destruction of stones much finer than the Auquhollie pillar. Diack was sceptical about the alleged presence of a stone circle because of his assertion that 'all experience is against it, since alphabetic writing has never been found on such circles' [1925:258]. This, of course, is not the case: a number of prehistoric standing stones have been reused in the early medieval period for ogham and other carvings. The practice seems to have particularly common in the North-East; several symbol statements and the Brandsbutt and Logie Elphinstone oghams were all carved on stones that were components of stone
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
42
circles.
Before being demolished, the adjacent croft was called 'Langstanes', which might be taken in suppon of Southesk's claim of a stone circle. Diack, however, maintained that the name derived from an unrelated circle of stones 'not far off [1925:258]. Presumably he meant one of the two stone-circles surrounding ring-cairns at nearby Raedykes [Ritchie 1923]. This explanation seems unlikely since these stone circles are a mile away, surely too distant to give rise to the naming of so insubstantial a building as the croft. The name cannot be taken as conclusive proof of the prior existence of a plurality of stones until it can be ruled out that it is not a possessive (Lang Stane's) derived from the name of the single stone, known locally as the 'Lang Steen' [Southesk 1886:37].
SITE The Auquhollie stone stands 467 feet above sea level [142m OD], on the north-west side of a farm-track which forms a field boundary. Though the stone is several miles inland at the head of a valley, because of the configuration of the surrounding hills, it is in fact in sight of the sea, which may be significant in considering the overwhelmingly coastal distribution of the Scottish oghams. There are no associated settlement or early ecclesiastical remains, but the vicinity was favoured for prehistoric ritual activity. At neighbouring Raedykes there is a group of four ring-cairns, two surrounded by stone-circles. Close by there is also the well-preserved remains of a Roman camp. The significance of the siting of the camp and, probably also the ogham, is its location on the route north across the last pass into Deeside (A957, the modern Slug Road). The great barrier of the Mounth comes very close to the sea at this point and beyond Stonehaven the coastal route becomes virtually impassable. This narrow corridor must have been of the greatest importance, linking, as it did the two major regions of Pictland [Alcock & Alcock 1992:269]. The siting of the Pictish citadel Dun Foithir immediately south of Stonehaven, either at Dunnottar castle or the neighbouring rock promontory of Bowduns [Alcock & Alcock 1992:276-7], is an indication of the strategic importance of the area in the period. Two sieges of this major site were important enough to be mentioned
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
43
in the Iona Annals (AU 680, 693).
DESCRIPTION OF OBJECT A natural, rectangular monolith, undressed and unshaped. The only carving is the ogham inscription up the south-east arris [see figs.]. Stone:
Quartzose gneiss [ECMS 203], highly stratified with frequent beds of quartz.
Dimensions:
0.43 x 0.66 x 0.7m at the base, 2.35m tall.
Condition:
Good, some wear at arris (carving sharpened in modern times).
The stone is hard and consequently the inscription is in reasonable condition. Only at die arris is weathering a problem. In particular, notches have become obscure in the middle portion and at the end. When Crawford visited Auquhollie for the OS in July 1939 he noted that the stone had been recently cleaned [OS record card]. Doubtless this was for the taking of R. A. S. Macalister's photograph in which the strokes appear remarkably fresh [1940 Fig 3]. It is clear that most, though not all, of the strokes and notches have been re-cut with a sharp tool in modern times, and it looks as if this happened before Macalister's, possibly even before Diack's, picture was takea The re-cutting has obscured only the base of each stroke, the original sides are largely intact. Since the more doubtful strokes were not touched up, they now look even more questionable in contrast to die ones which were sharpened.
DESCRIPTION OF CARVING Both Diack and Macalister claim to have found Pictish symbols on the stone, though in different places. According to Diack there were 'two small circles connected by two bars' on the nonh-east face 'at the height of the last w, about 2lA feet from the top of the stone', each circle was of I xh" diameter and the whole 'symbol' was 5te" across. Below this symbol he claimed to see circular traces of at least one farther symbol and likened the whole to the rudimentary symbols found in the Fife Caves [1925:262]. Macalister's symbols were 'squeezed away in the corner' of the 'B-surrace' (i.e. the west face), he described them as 'small, tentative, ill-made'. Padel could find no trace of these alleged symbols, nor
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
44
could Ian Smith when he examined die stone for the RCAHMS in 1983 [OS record card]. There is a small keyhole-like mark in the position described by Diack. If this is what he was meaning, it can hardly be a symbol and there is nothing to indicate it is not natural. Given the durability of the stone it is unlikely that any symbols seen as recently as the 1940s would have weathered away. The presence of Pictish symbols can be dismissed.
DESCRIPTION OF OGHAM INSCRIPTION The ogham inscription, on the other hand, is very clear. It is carved up the south-east edge of the stone, beginning about 0.3m from the bottom and continuing for a further 1.25m. It is difficult to be absolutely sure where the inscription ends but there is probably about 0.8m clear at the top. The stem is not a carved line, but the natural edge of the stone. The lettering follows the most sharply defined arris, but even this wavers considerably, especially at the outset, and confuses the interpretation of some of the earlier letters. The consonant strokes average about 50mm in length and, when due allowance has been made for the meandering arris, are generally perpendicular to the stem. Vowels are indicated with notches on the arris. The strokes are generally broad and chubby, and of proportions which comfortably accommodate fingers. The detail of their incision profile has been obscured by modern re-cutting, but they were probably of the standard 'pock and smooth' technique [Gordon 1956].
Padel is quite right to assert that the arris 'is so rough that one can often see what vowel-nicks one wishes' [53]. In cases of doubt, the strongest guide is the available space. The apparent gap between letters 2 and 3 is quite exceptional in a Type 1 inscription and usually one is quite safe to assume in such cases that all strokes were initially more or less evenly spaced, with no gaps between letters.
There is no evidence of the initial notch that Diack and Rh£s interpreted as A. Macalister noticed a 'small circle' before the initial V and identified it as 'surface disintegration'. Allen and Padel were probably right to disregard it. The inscription continues as follows [see fig. for key to numbering]:
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND 1
AUQUHOLLIE
45
The inscription proper begins with three strokes to the right of the stem - V. Padel and Macalister were of the opinion that they sloped slightly forwards but this is an illusion caused by the gradient of the arris at this point.
The next eleven strokes, comprising letters 2 to 4 are confused because of the sharp left-wards twist of the stem and the unusual spacing of the extant strokes. 2
Three shon strokes across the arris. Each is very slightly longer than the previous one, perhaps because the stem is beginning to veer to the left. This group is clearly differentiated from 1 and is to be interpreted as a vowel - U.
3
After a shon gap there are two strokes similar in length and relative position to those of 2. They are followed by a very shon stroke, really just a nick. If the nick is the remains of a stroke that was originally more akin to the previous two, it would be taken with them and the group read as three vowel strokes - U. The gap between 2 and 3 would then be seen to be intended to differentiate between two successive vowels. This interpretation, which would require the portion on the vulnerable arris to have survived while the outer ponion on the face was lost, does not tally entirely with the pattern of wear on the surface of the stone, but this is because of the false impression given by the re-cutting of pans of the strokes.
The interpretation is further
complicated by the fact that these vowel-strokes, though shoner than the flanking consonants (7 and 4) are longer than the vowel-notches later in the inscription, which are almost circular. On balance, however, the most likely interpretation is - U. 4
Five strokes to the right of the arris. The initial impression is of a pair of shoner strokes to the right followed by three longer ones continuing further to the left. The impression, however, is a false one because the pan of the first two strokes nearest the arris is fainter and has not been recut. Furthermore the stem twists sharply to the left over a slight overhang at this point. All recent authorities agree in taking these five together as (N) and this seems the most likely explanation.
5
One clear vowel-stroke - A. The gap between this notch and the next stroke is larger than might
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
46
be expected. Contrary to Padel's assertion, there is ample room for a second notch. There is a slight indentation on the arris where a hypothetical second notch would fall. Unlike the first it has not been re-cut and is therefore easily missed. I am inclined to read - 0 . 6
Five clear long strokes to the right of the stem - N.
7
There follows a 200mm interval over which the arris has become so worn that virtually nothing survives. A word boundary indicated by a blank space, such as Macalister suggested, would be unprecedented and it is safe to assume that this portion contained letters now obscure. From a distance vague notches on the arris can be discerned, but whether these are complete vowelstrokes, or the remains of longer consonant strokes is unclear. One would expect traces of any consonantal strokes to have survived on the flat surface of the stone. Since there are none, it may be more likely that vowel-notches have been lost from the vulnerable arris. There is room for more than five notches, i.e. certainly more than one letter, but exactly how many there were, or how they were spaced is very difficult to ascenain. Padel thought five was possible though he conceded there was room for more. Macalister read the second half of the gap as I (surmising that the carver had cut I in the first half then erased it and re-cut it further along to avoid confusion, an outlandish theory dictated by his preferred reading). Rhpsused Allen's measurements to show that the interval in question was the same as the interval between the V and the N (i.e. between 1 and 4), and argued therefore that the missing vowel had the same number of strokes. He read the first as I (or perhaps UO) and so read the second very tentatively as I also. About four notches in the middle appear more definite than those on either side. It is possible that they are the remains of an h-aicme letter, perhaps C or T, with a short vowel on either side (A or 0 ) . An alternation of vowels and consonants would be preferable to a string of vowels, though in the light of the apparent double vowel at the beginning of the text, the latter cannot be ruled out. If the letters are all vowels, the little which remains perhaps suggests the following division -1,3,2,1 (i.e. A U 0 A), but this is not much more than a guess.
8
Clarity returns with three strokes to the left of, and perpendicular to, the stem - T.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
47
9
The four vowel-notches which follow are feint but not in doubt - E.
10
Two strokes to the left of, and perpendicular to, the stem - D.
11
Two vowel notches - O.
12
The three strokes to the right of the stem here have some claim to be regarded as sloping forwards, but they are still clearly - V. A number of authorities stop here, but there is no doubt that the inscription continues.
13
Though two vowel notches are apparent in the next section (O), there is room for a third - U.
14
From here to the end, the configuration of the strokes is moot. Padel saw two separate strokes to the right with space for another in-between, but surely there are four, if not five - N (or S). Macalister read these as R but they do not seem any longer or more sloped than the preceding V.
15
Macalister thought he saw the tops of a second set of R strokes protruding out of the fractured arris. These are more reasonably interpreted as vowel notches, though they are extremely difficult to determine, I read four or possibly five - 1 (or E).
The above gives the following tentative reading: VU(u/o)nON [?auoa/a(c/t)o?] TEDOV(u/o)(n/s/r)(i/e/r) One disquieting aspect of the above reading is, apart from Macalister's doubtful R, there are no letters from the M-aicme. This is not necessarily a problem, but it should alert us to the possibility that we have, for instance, misinterpreted some of the longer vowel strokes.
FORM OF SCRIPT Auquhollie is written in a classic Type la form of the script, with very short strokes/notches for vowels and no drawn-in stem line. The stem wavers to such an extent that it is not always easy to evaluate the slope of letters, but most seem roughly perpendicular to the stem. Only certain b-aicme letters appear to incline. Weathering, especially of the middle section makes it difficult to gauge the spacing of strokes. Within letters they seem evenly spaced, between letters there tends to be a slightly greater gap. Consonants consist of strokes of no longer than medium length which, combined with the generous spacing of strokes
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
48
gives the script a rather 'boxy' appearance. In this it is quite unlike Newton, whose letters can be quite long and thin. The script of Gumess is comparable, though it lacks the slope on b- and h-aicme letters and, like Bac Mhic Connain and even Dunadd, distinguishes much more sharply the length of vowelstrokes and consonants.
INTERPRETATION OF TEXT - Historical Context Before attempting to make sense of the very tentative reading suggested above, it would be helpful to make explicit our initial assumptions about the language of this text based on our understanding of its historical context. Since Type I ogham pillars appears to be exclusively associated with Irish-speakers, whether in Ireland or in Irish colonies abroad, it would be natural to assume that the Auquhollie ogham was an Irishlanguage text, using the standard Irish orthography of the appropriate date. One would also expect the names carved to be Irish names.
All this is thrown into some doubt, however, by the unexpected
geographical location of the Auquhollie pillar within sight of the eastern coast of Pictland.
Place-name and other evidence shows that this area was Gaelic-speaking in the Middle Ages, but the first influx of Goidels is usually dated after the Dalriadic take-over of the Pictish kingship in the mid-ninth century, certainly far too late for the Auquhollie pillar. If found in Ireland, a monument of this form would be dated up to the sixth or, at the latest, the seventh century. The current model of Scottish history cannot explain such strong Irish influence in the North-East at this date, since we understand the only extensive Irish settlement in this period to be in Argyll. The Auquhollie ogham may be the only physical remains from an early Irish settlement in eastern Pictland. It takes on added significance in the light of literary evidence for a link between the area, the Mearns, and south-west Munster [Watson 1926:218-219].
In the Munster genealogies preserved in Rawlinson B.502 [CGH 148 a 21, 30 37] and also in the Books of Leinster, Lecan, and Ballymote, a link is made between the Éoganacht Locha Léin and the Éoganacht Maigi Gerrgind i nAlbae. Loch Léin is the Lakes of Killarney and Mag Gerrginn in Scotland is the plain
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
49
of the Pictish province of Circenn, in other words, the Mearns (Kincardineshire) [Hogan 1910 s.v.v.]. This connection is expressed in narrative form in the saga episode the Exile of Core son of Lughaidh, king of Munster, preserved in the twelfth century manuscript, the Book of Leinster [287 b 1 -acephalous] [Hull 1941, 1947], The story goes that Core is sent to his death in Alba but the trap is foiled by the King's chief poet Gruibne. That the plot device is an ogham message carved on a shield is, for our purposes, merely a serendipitous coincidence. Core marries the king's daughter and stays in Alba until he has fathered three sons, then he returns to Ireland with wife, sons, and treasure. As Watson explains, the story is an old one, as shown by a reference under the head-word gorn in Cormac's Glossary (c.900) to the retoiric of Gruibne's/d/te, 'welcome', to Core [1926:18].
David Sproule has shown how Core's heroic biography is constructed from politically motivated fblkloric commonplaces. The reason for including a Pictish element in the story 'presents a problem like that of the hen and the Qgg' [Sproule 1985:16], though I cannot agree with him that the connection is 'explicable ... in purely narrative terms' [17]. As Dr Clancy has suggested to me, the impetus is more likely to have come from Picts desirous of a non-Dalriadic Irish connection, than from Munstermen keen to link themselves with the Picts. But whatever the origin of the tradition it is indisputably early, accepted, as Sproule and Hull make clear, by Irish genealogists by the eighth century, and used against the Éoganacht Locha Léin who were already in decline by the second half of eighth century and were politically marginalized by c.800 [Byrne 1973:216-20]. Dr Clancy writes: "It strikes me as likely that the creation of a fictional Eóganacht ancestor must belong to a time when a) the Picts were strong, b) the Eóganacht were one of the strongest in Ireland, and possibly, c) Dál Rfada was either at an ebb, or undesirable. The most likely period, it seems to me, is the 730s, during the rise of Óengus m. Ferguso [f 761], and the rise in Munster of Cathal m. Finguine (f742), by far the most powerful Irish king at the time." [pers. comm. 31.3.93] In fact, this same genealogy states that Oengus ri Alban was a scion of the Éoganacht Maige Circenn. These and other nebulous links between Ireland and Pictland need to be untangled. In the story, the king of Pictland is called Feradach Finn Fechtnach, which M. O. Anderson has suggested may be echoed in the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
50
name of a prehistoric Pictish king commemorated in the king-lists, Uuradech uecla/uetla [1973:90]. Thomas Clancy has pointed out to me that Gniibne 'éces di Mpairi appears again as the audior of a quatrain used to illustrate a particular metrical form in a Middle Irish tract on metrics [Thumeysen 1891:33].
Watson supplies some local place-name evidence he claims corroborates the tale [220]. Drumforber, not far away in the parish of Laurencekirk, is 'Cairbre's ridge' (Drumquharbir, 1539). Watson tries to link it with Core's Pictish son Cairbre Cruithnedián, 'little Pict, or Pictish fellow', though, given Cairbre was the sixth most popular male name in early Ireland [Ó Corráin & Maguire 1990, s.v.], this fails to convince. Similarly, the link between Balmackewan, parish of Marykirk, (< Baile mac Eoghain 'stead of the sons of Eoghan') and the Eóganacht seems forced given that Eoghan was one of the twenty most popular names [op.cit]. The old name for the parish, Conveth, is certainly of interest (< Olr coinnmed 'free quartering, billeting' > legal term 'coigny' [DIL]\ but diere is no reason to believe the name predates the Gaelic ascendancy of the later ninth century.
The literary evidence points to the existence in Munster in the eighth century of at least a tradition of a connection with eastern Pictland south of the Mounth. Even if one is not prepared to accept an early (sixth? seventh? century) Irish settlement in Kincardineshire, the Auquhollie ogham pillar is unequivocal evidence for some form of Irish influence. One would imagine that such influences would need to be rather strong to produce, not only the ability, but die desire on the part of Picts to erect an Irish type of monument using an Irish type of script. If one is unprepared to accept this alleged Irish settlement then must one come up with another historical mechanism by which someone living in the foothills of the eastern Mounth would have wanted to and had the means to erect an ogham pillar.
INTERPRETATION OF TEXT - Linguistic Discussion Our first expectation would be that the Auquhollie text consisted of Irish names in die usual Irish
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
51
orthography, but even if the names were Pictish, the orthography would have to have been derived from Irish ogham, and so an Irish aspect would be inevitable. If the latter were the case the orthography might be rather experimental. Thomas has discussed the interplay of Goidelic and Brittonic on the ogham and other monuments of Dyfed and Dumnonia as the Irish settlers there Vent native', at least to a certain extent [1994].
Turning now to the ogham text, though much of this inscription is uncertain, it is clear that it lacks the central MAQI so typical of the pillar oghams of Ireland. The worn patch in the middle of the text is almost long enough for MAQI, but there is nothing in the little that remains to support such a reading. The AuquhoUie text is of fairly typical length and an interpretation 'X [*] Y', the [*] being the word for some interpersonal relationship, is highly likely. Though MAQI 'son' is by far the most common relationship expressed in the formula 'X [*] Y', a significant number have AVI 'grandson* and fewer have CELI 'client1, 'dependent', 'follower', one has NIOTTA 'nephew' [see QIC index verbomm, and McManus 1991:118-20 §6.27]. There is one example, from Eglwys Cymmin, Carmarthen (CIIC 362), of INIGENA 'daughter', the sole example of the ogham commemoration of a female. There would almost be room in the weathered central section of the AuquhoUie text for AVI, but again insufficient remains to support such a reading.
Of course, there are numerous Irish words expressing family and other relationships which might be deemed appropriate in this context, and there is, also the outside possibility that the language of the link word is Pictish. If the basic structure 'X [link] Yf is accepted, and the text is surely too long for a single name, this implies that the names on either side are something like: Vuunon [-] Tedovoni or Vuunon (l-]te) Dovoni The final -/ looks like a masculine genitive ending, if VUUNON is not complete as it stands, a genitive ending may have been lost in the worn patch:
i.e. Vuunon[- (-]te) Dovoni etc.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
52
An alternative interpretation is to take the letters at 7 as the possible remains of some form of NETTA or NIOTA 'sister's son* [McManus 1991:109-10 §6.15]. In Irish ogham inscriptions the word is usually the first element in a compound personal name of McManus's Type C. The most common Type C formation is that with MAQQI-, but as McManus explains, this 'does not denote a filial relationship to the second element - often a dependent genitive of a divine name or the name of a tree or a word associated with a trade - but probably originally had the meaning 'devotee* or the like* [1991:108-9 §6.14]. Names formed from NETTA, e.g. Nad~fraich, Nad Sétna, Nad-sluaig [CGH\7 had similar connotations. If, as discussed below, DOVONI is a form of the divine name DOVVINIAS, the case for a NETTA- name is strengthened.
The form in the classic ogham pillars is usually NETTA, though there are two examples with 10 vocalism, CIIC 202 and 252, in the latter NIOTTA is not pan of a name, but is the link word indicating family relationship. The relevant portion of the Auquhollie inscription is, unfortunately, obscured. There is cenainly room for 10 (NIOTE), or if I am right to suggest that there may be the remains of h-aicme strokes amongst the markings at 7, then NETTE. Final E rather than the expected A is a problem. The word may be comparable with Formaston's NEHHT-, Bressay's NAAHT-, and Gurness's -NEITA. To accept 6-9 as some form of NETTA- or NIOTTA, would necessitate separating off thefirstname VUUNO and the last name DOVONI.
WUNOIVUUNON1 Depending on how one wishes to segment the inscription, the ending could be -ONI, or simply - 0 . The former might be some form of the diminutive suffix -IGNI, -OGNI (VENDOGNI 'fair' [CIIC 422]) and especially -AGNI, which is so very common in the formation of Irish personal names [for 0 for A see McManus 1991:118 §6.26]. McManus notes a few examples of ogham -ONI for -OGNI, but these are all doubtful [1991:107-8 §6.12]. If the -I belongs to the next word (I[.]E, cf. St Vigeans/pe ?), the form may be VUUNON. McManus gives examples of early post-apocope inscriptions with -I dropped in one name but not the other [1991:94 §5.26], but the readings at Auquhollie are so uncenain that it would be unwise
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
53
to attempt a linguistic dating on these grounds alone. Final -O might reflect the genitive of a Old Irish /- or «-stems (-0 < Primitive Irish -ós) [McManus 116 §6.25]. No satisfactory explanation of the first element, with its double vowel, is immediately apparent, though we might compare VUUNO- with ogham VOENACUNAS (recte VOINA-CUNAS [McManus 121 §6.28]), CIIC 164. In Archaic Old Irish texts vowel length may be shown by doubling. Thurneysen suggests this may be intended to show something more than mere length 'perhaps a pronunciation bordering on disyllabic in certain positions of the word in its clause or in slow speech' [Thurneysen 1946:20 §27]. Whether the third character of the b-aicme is to be transliterated /w/ or HI is a function of date and linguistic context (see Introduction). Since neither of these have been established, I will use the conventional V
DOVONI Dovoni presents fewer problems. Names in DOV-, 'black', appear in a number of inscriptions from both sides of the Irish sea (Uhlich [1989] explains why the form is consistently DOV- rather than an expected DUBU-), for instance, DOVAIDONA (> Dubáed) CIIC 503, DOVALESCI (> Duiblesc) CIIC 63, DOVATUCI (> Dubthoch) CIIC 37 [Uhlich 1993: s.v.]. The Auquhollie name may be a form of DOVAGNI (> Dubári), cf. CIIC 432, with -ONI for -OGNI/-AGNI, the diminutive suffix, as discussed above [for a discussion of ogham DOV- see McManus 1991:122 §6.29]. Of particular interest in an area with supposed Eóganacht connections is the name of the ancestor deity of the Eóganacht sept, the Corcu Duibne, which appears in ogham MUCOI DOVINIA [CIIC 175, 178, cf. 156].
DISCUSSION In every sense, Auquhollie is a classic example of the orthodox Irish ogham pillar, except, that is, for its location. It is a rough pillar, the inscription runs up the (left) corner of the stone, there is no stem-line, vowels are represented with short notches, and the consonants have similar proportions to the standard Irish type and are not elongated like many Scottish examples. The strokes are equally spaced along the stem and there are no gaps between letters. The overall length of the inscription is typical, and even if the text
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
AUQUHOLLIE
54
is not of the common 'X MAQI Y' variety, that is far from the only formula used, McManus list fourteen [ 1991:52 §4.6]. In archaeological terms Auquhollie can be dated most probably fifth- to seventh-centuries. A fuller interpretation of the text would be required before linguistics could offer more precision. The main significance of the inscription is to demonstrate that ogham was known in mainland Pictland at this early date. Further corroboration for early knowledge of ogham in Pictish territory is provided by the radiocarbon dating of the Pool and Birsay 3 stones. The literary evidence for a connection between Munster and the Mearns has already been discussed; see Logie Elphinstone for discussion of the re-use of prehistoric monuments in the early medieval period.
BIBLIOGRAPHY
OS Record card - NO 89 SW 10
Southesk 1884:37; Rh?s 1892:270, 1898:348; Nicholson 1896:4 (D); ECMS 203-4; Ritchie 1923:27-8 fig.5; Diack 1925, 1944; Macalister 1940:190-1 fig.3 pl.IIa; Padel 1972:52-4.
Site Only:
OS Name Book, Kincardineshire No. 10 p.93; RCAHMS 1984:19 No.95
PREVIOUS READINGS Southesk
FaDHDONANUITENn
Rh?s 1892
VINONITEDOV
Rh?s 1898
VAMUNONITEDOV
Allen
VUONON i TEDOV
Diack
AVUO ANUNAO UATE DOVENI
Macalister
VUENONITEDOVOR
Padel
VUUNAN iTEDOVOB.B
AUQUHOLLIE - Ogham-inscribed pillar
• *x/ffir:
: • W'
AUQUHOLLIE - Ogham-inscribed pillar (detail)
* ^
////
15 14 13 12
V m
11 10
9 8
5 4 3 2 1
AUQUHOLLIE - Ogham inscription (key to numbering)
55
BAC
MHIC
CONNAIN
DISCOVERY AND CURRENT LOCATION This ogham-inscribed knife-handle was discovered in 1919 in the course of Erskine Beveridge's excavation of a wheelhouse on the tidal islet of Vallay, North Uist [NGR NF 7694 7619]. The Bac Mhic Connain site is located at an altitude of 5m on a sand-covered hillock about 165m north of the bay at Saltam, in Vallay Sound, and about 400m north-west of Vallay House [see maps J. According to Armit [1990a: 108J the wheelhouse was 200m from the western shore of the now drowned machair plain of Vallay Strand which 'it is likely to have overlooked'. The variousfindsfromthe excavation are now in the NMS. The ogham handle was accessioned in December 1921 and has the catalogue number GNB 134.
SITE The six week excavation of the wheelhouse at Bac Mhic Connain was part of a larger campaign of archaeological investigation in the area carried out by Erskine Beveridge after he acquired the estate of Vallay and Griminish. These excavations were so productive that in 1948 Sir Lindsay Scott was able to assert that 'we know more of the prehistory of this estate than of any comparable area in the British Isles' [68]. By the standards of modern scientific archaeology, they were unfortunately limited, and the lack of concern for stratigraphy (there are no sections) is a serious failing [Scott 1948:69]. In the case of Bac Mhic Connain the situation was further complicated by the fact that Dr Beveridge died before publishing his results. The task of publication fell, eventually, to J.G. Callander who had only incomplete notes from which to work [Beveridge & Callander 1932].
The Bac Mhic Connain complex had been extensively reconstructed during its lengthy period of use and quarried in recent times to provide stone for the building of Vallay House, so that all that stands today are 'mutilated remains' [OS record card]. The building sequence was little understood during excavation and the lack of detail in the report, and the little that survives to the present, means that the 'plan should be
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BAC MHIC CONNAIN
56
regarded as tentative' [ibid.]. It is clear, however, that the ruins at Bac Mhic Connain represent successive wheelhouse structures [Armit 1990a: 109] of the later centuries B.C. to the first century A.D. After the wheelhouse was abandoned the site continued in use for a considerable period, apparently as a metalworking workshop [Armit 1990c:63; see fig.].
A large number of artefacts were recovered from the site, including pottery, metal (iron and bronze), stone, and organic items (bone and deerhom). None of these, however, are securely connected to the construction or primary domestic use of the wheelhouse, and may relate to any phase in the post-wheelhouse sequence [ibid.]. Hallén commented that 'most of the artefacts are chronologically insensitive and could date to anywhere in this range [c.200 B.C. - eighth century A.D.]' [1994:193]. Stone casting-moulds and triangular crucibles indicate on-site metal-working which Armit would date to the Roman period [Armit 1990a. 110]. In support of this view is the small fragment of Samian ware of possibly second century A.D. date [Robertson 1970] and the whalebone mirror (NMS GNB 61) which falls into the early part of Warner's EIA style and may, therefore, date to the first or second century AD [Warner 1983:164]. The prominence of metal-working at Bac Mhic Connain may simply reflect the fact that the workshop was built on top of the former house while at other sites in the locality the industrial area was away from the habitation area and therefore not excavated [Scott 1948:79]. The preservation of bone and antler was a notable feature of the excavation of Bac Mhic Connain and neighbouring Foshigarry. The surviving finds have recently been subjected to a detailed re-assessment by Ywonne Hallén [1994].
Post-roman period activity is indicated by the bone die (NMS GNB 65) [Clarke 1970; Mackie 1971] and the ogham-inscribed knife-handle. The latter has been dated provisionally by other authorities to the sixthto eighth-centuries AD [OS record card]. This proposed range is indeed not unlikely, but typologically the Bac Mhic Connain ogham is extremely simple. There are no grounds for assigning it to any particular date except within the broadest range. The possibility that the Pool ogham may be as early as the fifthcentury means we should remain open-minded on the dating of the Uist example. At the very similar
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BAC MHIC CONN AIN
57
neighbouring site of Foshigarry, the bulk of the settlement was similarly pre-Roman, though Hallén adduced pins and composite bone combs as evidence of possibly seventh or eighth century occupation [1994:193, but see Foster 1990 on the difficulties of dating such combs].
The raw materials needed for the bronze-working would have had to have been imported [Scott 1948:79]. The presence at Bac Mhic Connainof Samian ware indicates ultimately Mediterranean contacts, doubtless mediated through sites in the Irish sea province further south. Other finds imply links with Ireland, for instance the mirror handle has parallels with bronze items in Ireland [Raftery 1984:209]. An almost identical bronze version from Lochlee, Ayrshire [MacGregor 1976 No.272], complicates the picture, because on the strength of it MacGregor suggests possible influence from the western shores of northern England [MacGregor 1976.9]. Hallén comments on the possibility that such mirrors were status objects [1994:225].
The settlement at Bac Mhic Connain is fairly typical of the period in being well adapted to the local environment, and having access to both coastal and machair resources. The nature of the bone and antler finds lead Hallén to comment on the 'expert exploitation of locally available resources and considerable craft skills' of the inhabitants [1994:228]. Though the site plan shows a change from wheelhouses to cellular structures, domestic ceramics show the continuity of cultural tradition [Lane 1990]. Subsequent research [Lane 1987] has shown that there is no archaeological evidence for the Hebridean invasion of 'Scotto-Pictish' people proposed by Crawford [Crawford 1974; Crawford & Switsur 1977; see Armit 1990a:267]]. Instead, Armit has suggested that the decline of monumental 'broch' architecture and the rise of cellular structures reflects a more centralized society in the region. Though he emphasizes that power continued to be exercised at a highly localized level [1990a.276] he postulated, tentatively, the establishment of some form of centralized control based in Orkney [278]. In this context parallels between Bac Mhic Connain and the Orkney oghams become significant.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
fi^C
MHIC CONN AIN
58
Nothing is known about the district in the early medieval period. At the end of the seventeenth century, Martin Martin mentions three chapels on Vallay including one dedicated to St. Ultán [1703:67]. There were several saints of this name, though dedications to him in Scotland are rare. Beveridge quotes Forbes [1872] as the source of information of St. Ultán of Kintyre, a silver shrine containing a relic of the saints arm still being in the possession of its hereditary keeper at Sanda c. 1600 [Beveridge 1911:297]. The now lost place-name of Petultin 'Ultán's share', near St Andrews [Watson 1926:409], may reflect a cult of Ultán, but the name, which means 'an Ulsterman' was also in general use in the early medieval period [Ó Corráin & Maguire 1990:174-5]. Beveridge mentions the presence of cross-incised slabs [Beveridge 1911:297-8]. The multi-period site of Coileagan an Udail nearby was occupied in the early medieval period but the results of the excavations in the 1970s remain substantially unpublished. The place-name Vallay is Norse ( < vadill 'shallow water, especially passable on horse-back) [Beveridge 1911:96], Bac Mhic Connain means the '(peat) bank of the son of Connán', the Gaelic bac being a loan-word from Norse, bakki 'bank', 'ridge' [Ordnance Survey 1968:6].
DESCRIPTION OF OBJECT A knife-handle of cetacean bone retaining the rusted tang of an iron blade, incised with ogham letters but otherwise undecorated [see fig.]. Material:
Cetacean bone and iron
Dimensions:
Handle: length 108mm, diameter 16mm; Tang: 6mm
Condition:
All of blade missing except msted tang, bone well preserved, carving clear
The bone has been planed to produce a slight ridge for one line of ogham lettering. There is no other decoration. The handle tapers slightly in a gentle curve to the butt-end where there is a conical hollow. A fracture has occurred near the blade-end, which is damaged. It can be seen from the tapering outline of the bone, however, that very little has been lost.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BAC MHIC CONNAIN
59
This is one of fourteen handles found at Bac Mhic Connain, most made of antler. Both antler and cetacean bone have 'highly desirable properties of resilience and strength' [Hallén 1994:198]. Hallén argues that it is unlikely that whales and related marine mammals were being actively hunted, instead the bones were more likely taken from stranded animals [198-9]. Since only a small part of the bone is used for each artefact and the original surface is rarely present it is usually impossible to identify species [ibid.].
DESCRIPTION OF CARVING
The inscription begins about half-way along the handle and continues for approximately 50mm, almost to the handle's end. There is no trace of anything before the first stroke, and it appears that the inscription is complete. The photograph published by Macalister is excellent, but the one published by Callander in PSAS 66 has had the strokes drawn in, not with total accuracy, and the impression given is incomplete. Marion O'Neill's drawing is misleading in that it suggests the presence of a stem-line, but is otherwise reasonably accurate, though some of the vowel-strokes are marginally longer than she suggests [Hallén 1994:Illus.l2.5]
The longest consonant-strokes are c.6mm long. The lines arefineand appear to have been cut with a sharp blade. The carving is generally clear and well-preserved, the only doubt occurs when the ridge has been worn or chipped and vowel-notches lost. The individual strokes appear to have been carefully spaced, so where notches have been lost through wear, the available space may be used to determine what is missing. The component strokes of certain letters differ somewhat in relative length. Macalister reads too much into these variations which are the result, simply, of straight lines being carved across an unstraight angle with a single knife-stroke. There is clear differentiation of vowels and consonants, and even if the outer edge of a group undulates, on the inner edge, all strokes terminate at the same distance from the ridge/stem. The very slight variation in spacing between components strokes of certain letters is negligible. Each stroke is cut with a single slash, there has been no re-cutting, which perhaps implies that the carving was
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BAC MHIC CONNAIN
60
done with skill and some confidence. All strokes are perpendicular to the stem. Reading from the butt to the blade, left to right, the carving is as follows [see fig. for key to numbering]: 1
One long stroke perpendicularly across the stem - M. The upper and lower portions are clearly visible, the middle section on the ridge has been lost, but there is no doubt that it is all one stroke. There follows a small gap where the ridge has been chipped. The letters in the rest of the inscription are tightly grouped so it is unlikely that the empty space is original. Instead we should postulate a vowel of probably one, or a maximum of two, strokes - A (or possibly O).
2
Five strokes to the left of the stem - Q. The first of the group is slightly shorter than the others.
3
Three clear vowel-strokes on the ridge - U.
4
Five strokes to the right of the stem - N. The first three are clear, the last two, though barely visible, are just discernible. Their presence is confirmed by the spacing of the following letter.
5
Three strokes to the left of the stem - T. Macalister read an extra stroke at the beginning of this group (C). Though a mark is apparent in the appropriate place on Macalister's photograph, I could see no sign of the extra stroke on the handle. Padel referred to it as 'indefinite1. If it did exist it would be almost growing out of the final stroke of the preceding N. Since the rest of the inscription is so well-spaced, such extreme crowding is unlikely.
6
Four vowel-strokes on the ridge, the fourth being slightly less clear than the others - E.
7
Five strokes to the right of the stem - N. The fourth is slightly longer than the others. There is no sign of anything between this letter and the next. Padel described a group of extremely closely packed indentations. These were ignored by Callander, but interpreted as U by Macalister and the NMS. Padel himself interpreted them as 'perhaps a single vowel-stroke (A) rather closer to the N than to the ensuing letter'. Surely there is enough room for only an A.
8
Four strokes to the left of the stem - C. The distal ends of the last two are lost in the ensuing fracture. If there had been a fifth stroke its distal end would have been visible beyond the crack. Ignoring for a moment the break, the gap between the fourth stroke of the C and the first stroke of the following letter is sufficient for about two or three vowel-strokes (O or U). There is no
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BAC MHIC CONN AIN
61
trace of any carving there but otherwise the space is inexplicable. The fracture is undoubtedly subsequent to the carving and even if it had been incipient at the time, it was not avoided by the oghamist: the preceding C runs right into the break and the gap is after it. 9
Three strokes to the left of the stem - T. There is no trace of anything following this letter, and the handle ends shortly after.
Reading towards the blade as above yields: M(a/o)QUNTEN( /a)CoT and in the opposite direction: VoS( /a)QEVQUN(a/o)M
Although the ogham could be read in either direction, the former may make more linguistic sense. The only available precedent is that the Gurness inscription appears to read towards the blade. The Bac Mhic Connain oghamist has been able to fit the text into the available space perfectly starting half-way along, but an experienced writer would have little difficulty in judging how much space to allow, so this need not mitigate against the proposed direction of leading.
FORM OF SCRIPT The text is written in a delicate type I form of the script comprising long, thin consonant-strokes and very short vowel-strokes. There is no stem-line. The constituent strokes of the individual letters are closely and evenly spaced, there are no gaps between letters. Twofeaturesare worth mentioning - there are no double letters in the inscription and the initial letter, M, is upright, not oblique. The lack of slope on the M-stroke is in contrast to the Gurness ogham which has a very pronounced slanting long M. In terms of script typology Bac Mhic Connain is closer to Weeting than Gurness, and, despite the gross difference in scale, has many similarities with Auquhollie.
INTERPRETATION OF TEXT No obvious interpretation of the sequence MAQUNTENaCoT immediately presents itself, but a number
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
g ^ C MHIC CONNAIN
62
of observations can be made. The most conspicuous of these is that the opening three letters MAQ suggest the Irish for 'son'. In the earliest Irish inscriptions this word is written MAQQI (gen.). Since palatal quality is not usually indicated in ogham texts, die post-apocope form, corresponding to MS make, is either MAQ [QIC 7, 9, 10, 55, 112, 145, 219, 220, 233, 248, 364, 409] or MAC [QIC 83, 90, 127?, 256], and die former presumably is what we have here. The same form, MAQ, appears on the Gigha pillar, and at Blackwaterfoot 1 there is the related MEQ, though die reading is difficult. These three examples of single Q contrast with occurrences of the word further north and east where the consonant is always geminate, whether MAQQ or MEQQ. In contrast to the Irish and Welsh oghams the word is always spelled in Scotland with Q(Q) rather than C(C), which, given the date of most of the examples must be some kind of fossil.
In initial position as here, MAQ is most naturally interpreted, not as a patronymic, but as the first element in an uncompounded dithematic name of the form MAQQI-N. (Mac-N.). Such names were very popular among the early Irish, especially after about 600 A.D. [McManus 1991:101, 108-9 §§ 6.2; 6.14]. The dependent genitive in Maqqi-N.-type names is frequently a theonym and such names may have had the original meaning 'devotee of N\ The difficulty with this interpretation of the Bac Mhic Connain text is that it leaves the somewhat baffling Unten(a)cot. Perhaps this is related to the name of the Romano-Celtic deity Antenociticos whose temple was at the fort of Benwell (Condercnm) on Hadrian's Wall. Remains of the temple survive along with three inscriptions [Collingwood & Wright 1965: Nos. 1327-9] and parts of a full-size statue of a youthful god [Webster 1987:pl.l6]. The name has so far defied interpretation [Webster 74] and die cult is restricted to this one site. Neither Untenacot or Untencot may be derived from the form Antenociticos as it stands, but may comefroman unattested alternative vocalization of the same root. If Unten(a)cot is in some way connected with Antenociticos then the Bac Mhic Connain text may consist of a personal name originally meaning 'devotee of Antenociticos' (though the Irish evidence indicates that similar names which presumably had their origin in pagan cults, lost their religious connotations and continued to be used well into the Christian period).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
JJAC M H I C C O N N A I N
63
In addition to Maq Unten(a)coty a number of alternative segmentations are possible, for instance, Maqu Nten(a)coty Maqu n-Ten(a)cou Maq U n-Ten(a)coty or Maqun Ten(a)cot. Taking each in turn, I can find no instances of MAQU on Irish or Welsh oghams. The form MAC(C)V- for MAQ(Q)I-, indicating an intermediate stage in the delabializadon of /k w 7, appears on Roman alphabet inscriptions from the second half of the fifth century [McManus 1991:90, 98 §§ 5.18, 5.32] but this is spelled with V not U. The feet that the letter following the Q on the Gigha pillar may be U rather than I serves only to complicate matters. To interpret MAQU as accusative plural, manuscript maiccu, would require us to overlook the lack of any indication of palatalization, and also to come up with an justification of how accusative plural is appropriate in an epigraphic text of this form. To explain Bac Mhic Connain's MAQU as macu (an attested variant of macculmoccu [DH], would require some kind of confusion or conflation with the unhistorical MAQQ, since the antecedent of macculmoccu was MUCOI, i.e. /muk-/ not /mukw-/ [see Formaston for discussion of this word]. Nor is the N easily explained. We may reject the unlikely looking Ntenacot, perhaps leaving for consideration the single N indicating nasalization. But this is hard to square with the syntax of the text as far as it can be understood. The only form to cause nasalization is the genitive plural, MS mace, and this still leaves us with the U to be explained.
To get round the whole problem we might posit an alternative reading, with the five strokes of 4 as vowelrather than b-fl/c/w^-strokes. This would give I and allow the reading MAQ UI, i.e. Mac UL The formula word Ui 'descendants' is used to introduce the name of a tribe, sept, or kindred, and, in the singular, appears in several ogham inscriptions in its earlier form AVI [McManus 1991:110-2 §§6.17-18, 118-9 §6.27]. The name is usually followed by the name of the historical ancestor of the kindred in the genitive. When the name is that of a remote mythological ancestor, or deity, the name may have had religious rather than genealogical significance [McManus loc. cit.; MacNeill 1909:368-9]. The juxtaposition of the form Ul with the Q in MAQ is acceptable only if the latter is taken as a relic of the earlier spelling fossilized, as it appears to have been on all die Mainland Scottish examples, including ones presumably far later than
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
g ^ C MHIC CONN AIN
64
Bac Mhic Connain. In the end, however, there may be insufficient justification for reading 4 as I rather than N. The group is quite similar to 3 which is more clearly vocalic. While die strokes are slightly longer than j ' s , they are not as long as those of the other N (5). Previous commentators do not appear to have been in any doubt that 4 was to be read N.
Abandoning this tack altogether, an alternative interpretation entails taking the first five letters as a single word - MAQUN. To take this as a form of the personal name Mac-con 'son of a wolf [Ó Corráin & Maguire 1990:127] requires, not only explaining Q for c, but U for o. If we are committed to interpreting the Bac Mhic Connain text as Pictish, and are prepared to transliterate the fifth member of the h-aicme as /p/ (since p and q are cognate in Brittonic and Goidelic), then 1-5 might be taken as a form of mapan, maphan, maban, a Pictish word doubtless containing the cognate of Welsh n\ap 'son' (cf. AU 725=726 Talorcan ntaphan) [Jackson 1955:145].
Turning now to the second element, detaching in turn mac, rnacu or macun, leaves Unten(a)c(o)t, Nten(a)c(o)t, or Ten(a)c(o)t, and other permutations if further sub-segmentation is allowed. Once again, no obvious interpretation suggests itself, but a few observations may be made. The final letters C(o?)T may be compared with die Gaulish name element cot(t) (OCorn. cothy Bret, coz 'old'; from the root meaning 'living', 'lasting'). In proper names this is most common as a first element, but is attested in second position in the personal names Esanekoti and Venicotenius and the name of a (north) British tribe, the Atecotti [Evans 1967:186-7]. For the reading -TENCT, John Koch has brought to my attention the root *tantfi-, reflected in the tribal name Tencteri, in Irish Téchtae 'legal Tightness' [DH], cf. Welsh teithi. Comparison may also be made with the female personal name Tunccetaca which appears on a fifth or sixth century roman alphabet inscription at St Nicholas, Pembroke (CIIC 451). Jackson derives this from Brit. *ToncetBc5i cf. W. twng 'oath' [1953:273]. The name TENACI appears in ogham at Ballrorannig, Corca Dhuibhne, Co. Kerry (CIIC 148). The form consists of ten + the guttural suffix -acand is reflected
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
g ^ C MHIC C O N N AIN
65
in the sept name Ui Thenaich [McManus 1991:108 §6.13].
Since neither archaeology nor palaeography can offer much help in dating the Bac Mhic Connain ogham, we must rely solely on linguistic dating. Precision will not be possible until the text is satisfactorily explained, though the lack of anything after the final T is significant. It is not clear whether the U belongs to what precedes or follows it. McManus ascribes Irish oghams which display the results of apocope but not syncope to the middle or second half of the sixth century [1991:98-97 §5.30]. In these inscriptions, he comments, the word for 'son' tends to be written MAQ [96 §5.28]. Since, however, the formula word appears to have become fossilized in Scotland, any dating based on the form MAQ is suspect.
DISCUSSION The Bac Mhic Connain knife-handle is one of a very small group of ogham-inscribed household objects, and as such provides important evidence for the non-monumental use of the script. That the early medieval occupants of Bac Mhic Connain were engaged in metal-working implies they were people of some standing, but the site is not of especially high status. Thus we have evidence of ogham literacy at a non-church-site, on the part of people of middling social standing. One must always beware, of course, of reading too much into the location of an object as portable as a knife, but two factors combine to suggest that this example is unlikely to have travelled far. Firstly it is not a high-status object, but part of the ordinary domestic tool-kit. Secondly, it fits in closely with the other artefacts from the site. Hallén points out that cetacean bone was used at Bac Mhic Connain and Foshigarry in situations where other materials, such as antler, wood, and bronze, would have been used elsewhere [1994:198; cf. MacGregor 1985:31]. Since cetacean bone would not have been available inland, we may assume, at least, that the knife came from the Atlantic coast.
The calcareous sands of the west coast of Scotland meant that a wealth of antler and bone material was preserved at Bac Mhic Connain, artefacts which would have perished in the less favourable soil conditions
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
g ^ C MHIC CONNAIN
66
of the rest of the country [Hallén 1994:228]. By this chance we are given a glimpse of what might have been a far larger category of evidence - ogham on household objects. Two other ogham-inscribed knifehandles have been preserved, one at Gumess in Orkney [q.v.], the other at Weeting in Norfolk [Clarke 1952; Padel 1972:144-8; Holder 1990:58-65]. Other ogham-inscribed instrumenta domestica include the Buckquoy whorl, and the Hiberno-Norse Dublin comb [unpublished]. It is impossible to gauge how common such items might once have been, but the tiny handful which do survive are an important reminder that the current sample, so heavily weighted to public monuments, may not be a true reflection of the original extent to which ogham was used in informal contexts.
The Bac Mhic Connain knife-handle is unremarkable in all respects except for its inscription. The text appears to be a single male personal name and as such probably records the identity of the owner or maker (or donor if it was given as a gift) [see Buckquoy for discussion of the kind of texts appearing on personal belongings in the period]. Previous commentators have been united in labelling the Bac Mhic Connain ogham as 'Pictish', apparently on no more solid grounds than that it is 'unintelligible* and therefore reflects a supposedly non-Indo-European Pictish language. Wainwright took Bac Mhic Connain as evidence for continuing Pictish control of the Western Isles [1961:7]. This seems cock-eyed logic to me. Though the interpretation remains doubtful, the letters are most naturally read as Irish. Whatever the language of the text, the use of the ogham script indicates ultimately Irish influence.
Very little indeed is known about the linguistic and ethnic circumstances prevailing in the Outer Isles in the early medieval period. According to Irish tradition they were inhabited by the TuathlFir Iboth or Ibdaig, names which preserves Ptolemy's name for the archipelago, Ebudae [O'Rahilly 1946:538; Watson 1926:37-8]. The Dame appears to have non-Indo-European origins, but since the names of islands are often of great antiquity, it need not imply that the Fir Iboth spoke a non-Indo-European language (though they might have). Ptolemy's Dumna later Irish Domon, which Watson takes to refer to the Outer Hebrides [1926:40-41] is clearly Celtic. In Irish tradition the Fir Iboth are linked with the Tuath Ore and presented
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BAC MHIC CONNAIN
67
as piratical raiders [O'Rahilly 1946:377; Watson 1926:60-6]. This is of great interest in the light of Armit's suggestion of growing Orcadian hegemony in the Scottish Atlantic.
Lane has contrasted the aceramic southern Hebrides (from Mull southwards) with a northern zone of Iron Age ceramics stretching from Lewis to Tiree [1990:111.7.7]. If this reflects a more general cultural divide then it pre-figures the later boundary between Pictland and Dál Ríada. It is unclear when Gaelic speech became prevalent in this northern zone, but Richard Cox's work on the place-names of Lewis demonstrates that the area was thoroughly Gaelic-speaking before the first Norse settlements in the ninth century [1991]. The language of the Bac Mhic Connain text has not yet been established definitively. It may be Irish, but even if it is not, the use of the ogham script is sufficient proof of ultimately Irish influence at the site. As noted above the mirror-handle may be further evidence of Irish contact. Such contact may have been direct from Ireland or mediated through Dál Ríada, but it cannot be dissociated from the evidence of Irish influence in Orkney. There is no positive evidence for any ecclesiastical activity at Bac Mhic Connain and the site has only ever been explained in secular terms. While all the historical evidence is, of course, for ecclesiastical links with Ireland, archaeology suggests contact was not solely restricted to the church.
BIBLIOGRAPHY
OS Record card - NF 77 NE 5
n.a. 1922: accession 9178; RCAHMS 1928:89, No. 271 [plan of earth-house, mention only of ogham]; Beveridge & Callander 1932:56, 65, fig. 11; Macalister 1940:216 pl.vi, No. 21 (sketch of inscription), fig.8 (photograph) opp.p.218; Padel 1972:133-7; Ritchie 1987:65; Holder 1990:46-52; Hallén 1994:219 ill. 12.5.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND Site only:
BAC MHIC CONNAIN
68
RCAHMS 1928:89, No.271; Beveridge 1931; Scott 1948:75-76, 77-80; Armit 1990a: 108-110.
Artefacts:
n.a. 1922 [bulk of Beveridge collection]; n.a. 1923 [clay pot]; n.a. 1929 [further bone items]; Tylecote 1962:36-7, 194, 199; Robenson 1970:Table 4 [Samian ware]; Clarke 1970:229, No. 4 [bone die]; Mackie 1971:70 [bone die]; MacGregor, M. 1976:9, 141, 143, 163, No.271 [mirror handle etc.]; Laing & Laing 1987:215 [hanging bowl escutcheon]; Lane 1988:55-6; Hallén 1994 [all bone and antler material].
PREVIOUS READINGS
Callender
MAQUNmDENC(o/u)T
Macalister
BELANCEN UCOTA 'Belanc's knife'
NMS label
MEQUNTENUCOT
Padel
(m/h)..QUN(t/c)EN(a/u)C..T or V..S(a/u)QE(v/s)QUN..(m/b)
Holder
AQU(l/n)(q/c)ENUC T (reading based on MacalisterTs and Callander's photographs)
ILLUS I
Site location map with other wheelhouse sites. I Foshigarry: 2 Bac Mhic Connain: 3 Garry lochdrach; 4 Sollas {Drawn by Marion O'Seil). Based upon the Ordnance Survey map 0 Crown copyright
BAC MHIC CONNAIN - Site Location maps. Above: Bac Mhic Connain and the wheelhouse sites of Vallay Sound [Hallén 1994:190 Illus. 1]. Below: As above showing relief [Scott 1948:69 fig.7|.
t
(i)Machair Leathann ; (2) Eilean Malcit; (3) Cnoc a'Comhdhalnch ; (4) Garry lochdrach ; (5) Foshigarry; (6) Bac Mhic Connain; (7) Dun Thomaidh Land over 200 feet shaded
BAC MHIC CONNAIN
i m
Plans of the structures at Bac Mhic Connain. 1 Hearth. 2 Sink. 3 Stone box. 4 Furnace. (Drawn hx Marion O'Neil (after Beveridge & Callander
l#3l~2n
Plans of the structures at Bac Mhic Connain [Hallén 1994:191 Illus.2]
BAC MHIC CONNAIN Ogham-inscribed knife handle [Hallén 1994:220 Illus. 12.5]
B AC MHIC CONNAIN - Ogham-inscribed knife handle [® Trustees of the National Museums of Scotland, Neg. P9557]
Ill II Ill MM BAC MHIC CONNAIN - Ogham inscription (key to numbering)
in 1 2
3
MM 4
5
6
7
8
9
69
BIRSAY
SITE The Brough of Birsay is a tidal islet, just over 20 hectares in area, off the north-west coast of Mainland Orkney at the head of Birsay Bay (NGR HY 239 285) [see map]. The earliest documentary reference to the site is in chapter 31 of Orkneyinga Saga wherein is recorded the decision of the mighty mid-eleventh century Earl Thorfiiiii to build his minster there [Palsson & Edwards 1978]. The exceedingly fertile soils of the hinterland coupled with the defensive, though not isolated, position of the Brough itself and the good landing-place of Birsay Bay were sufficient to ensure that the site had been occupied for many centuries before that. It is especially relevant to a consideration of the importation of ogham to note that the Bay of Birsay 'is particularly well-situated for boats setting out to the Hebrides or Ireland' [Ritchie 1983:47].
Archaeological investigation of the site began in 1934 when the then Chief Inspector of Ancient Monuments, James Richardson, initiated a programme of repair and partial restoration in Area I, the Norse church and cemetery. In 1935 the church was cleared of debris but, though work continued in this area for a couple of years, there was no actual excavation there. In 1936 Mrs Curie assisted Richardson in beginning excavation of Area II, between the cemetery and the cliffs to the east. In subsequent years the work extended to include the isolated buildings in area 111, to the west of the cemetery. All work on the site was halted in 1939 on the imminent outbreak of the Second World War [Curie 1983:67]. A brief synopsis of the major discoveries of this period were included in the 1946 RCAHMS Inventory of Orkney.
Work at the Brough was not recommenced until 1956 when Stuart Cruden and Ralegh Radford began several seasons of excavation in Area II and House C in Area 111. In 1973-4 Mrs Curie and John Hunter excavated Room 5 in Area II. The finds from the initial forty years of excavation were first published in 1982, almost fifty years after the work had first commenced [Curie 1982]. Work continued throughout the 1970s and 80s, by this time in the broader context of the 'Birsay Bay Project'. The results of the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSA Y ( G E N E R A L )
70
rescue excavations of the period 1974-1982 were published by Hunter [Hunter 1986]. Morris is collecting and editing the earlier excavation archives, such as they are, for publication alongside the results of his own Birsay Bay Project excavations [pers. comm.].
In Ritchie's words 'more speculation has surrounded the archaeological and historical interpretation of the Brough of Birsay than any other Dark Age site in Scotland' [1986:5]. The problem is compounded by the non-publication of the results of the earliest excavations. It is clear that the site was important in both the Pictish and Norse periods [see fig.]. The earliest settlement remains on the Brough may date to the sixth century, but the nature of the early occupation remains controversial. Radford identified a chapel and grave-yard beneath that of the later Norse settlement and an enclosing wall. On the basis of these, the ogham inscriptions, two sculptured slabs, and a bell, he understood the site to be a Celtic monastery. Charles Thomas concurred that the Brough, with its curvilinear vallum, was a 'full monastic foundation', perhaps the principal monastery in Orkney [Thomas 1971:37], Although this interpretation was widely accepted for many years it has been called into question by more recent excavators who feel it is founded on an outdated and increasingly ill-fitting model of monastic sites. As Morris points out, Radford's monastic interpretation is based on Tintagel as a type site [Morris 1989:11], but Tintagel is no longer thought to be a religious site [12] (see Morris 1996 for a final refutation of Radford's model).
Its physical location, the substantial nature of the remains, the evidence of high-status metal-working, and the extant sculpture, lead all to agree on the 'special status' of the Brough site, but whether this 'derived from political considerations or monastic impetus is ... finely balanced according to current evidence' [Hunter 1986:171]. Morris thought the interpretation of the most recently excavated material as monastic to be 'attractive but not compulsive' [Morris 1989:15], but in his most recent publication on the site has expressed the opinion that 'the balance is now firmly tipping towards the political' [1996:61]. The small church and graveyard if they do indeed date to the Pictish period, need not be monastic [Morris 1982:81]. The bell, of a type usually diagnostic of Irish monastic influence, came, in feet, from Norse period layers
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND [14].
BIRS A Y (GENERAL)
71
Ogham in Scotland has no obvious monastic connection.
The date and circumstances of the earliest Norse occupation of the site are not clear either, though the 'considerable number of unmistakably Pictish objects' in the Lower Norse Horizon [Curie 1983:78] implies an element of Pictish continuity or assimilation. In the Middle Norse Horizon, which lacked Pictish material, there are a number of rooms with shared walls and separate entrances which have been interpreted as monastic cells. Curie thought this unlikely because of the large number of spindle-whorls found in the area [1983:78]. It has been the traditional view that Earl Thorfinn erected a minster, 'Christ Church', the first Norse Christian Church in Orkney, on the site of a Celtic Chapel on the Brough, though the alternative site in Birsay village on the mainland has perhaps a better claim. By this time Birsay was the centre of the Norse Earldom of Orkney. In the early twelfth century there was a cathedral at Birsay, dedicated, according to tradition, to St. Peter. Though it was superseded by the erection of the cathedral in Kirkwall in the mid-twelfth century, it continued as a place of pilgrimage until the reformation.
EARLY CHRISTIAN SCULPTURE FROM THE BROUGH OF BIRSAY In addition to the ogham- and rune-inscribed stones, three slabs bearing sculpture have been recovered [see figs.]. Cross-slabs The two cross-inscribed stones are the only explicitly Christian items from the site. One, not mentioned by Curie, is a small slab incised with a simple outline cross with a circle at the intersection of the anus. It currently stands upright near the facsimile of the symbol stone, but was found near the church lying horizontal, covering a grave [Ritchie 1986:12, ill.]. The other, smaller stone, was a surface find in Area II. It is a fragment of sandstone flag, 115x220mm, dressed and smoothed on both sides. The double outline latin cross has expanded rectangular arms and a rectangular base [Curie 1982 No.607,111.45 p. 70, discussed p.92. See fig.].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY (GENERAL)
72
Symbol Stone The famous Birsay slab was discovered in fragments in 1935 close to the side of the cemetery. It is now in the NMS, Cat. No. IB 243 (a rather poor facsimile stands on the Brough). When complete it would have been about 2.0m x 0.8 m, yet it is only 25mm thick, leading most commentators to postulate that the back, presumably carved with a cross, is missing [Curie 1982:97]. As Ritchie has explained, this is a misconception, and the famous association with the triple grave 'is the result of wishful thinking' [1986:12; 1989:53]. The extant face is incised with two symbol pairs - a 'mirror case' with crescent-and-V-rod, above a Pictísh beast with eagle - a panel containing, in false low-relief, three warriors bearing shields and spears. The slab is discussed in some detail by Curie [1982:91-2, 97-100; see also RCAHMS 1946:4-5 No.l; Close-Brooks & Stevenson 1982:30; A. Ritchie 1983:52].
RUNIC INSCRIPTIONS FROM THE BROUGH OF BIRSAY Four rune-inscribed objects, bearing between them perhaps six separate inscriptions, have been recovered from the Brough of Birsay [see figs.].
Since it cannot be ruled out that the two pre-War ogham
inscriptions date from the Norse period any discussion of them must take account of their runic counterparts.
The first inscription came to light at Birsay in 1921 when it was spotted by Hugh Marwick built into the outer face of the north chancel-wall of the Norse church about a foot from the ground [RCAHMS 1946:35 No. 120]. The runes were on its upper surface and for this reason must have been carved before the stone was incorporated in the wall. The 'longish, narrow stone, seemingly dressed' [RCAHMS 1946:35], designated Birsay 1 (Barnes OR 6), is 2'10" x 6" x 4" (approx. 0.86 x 0.15 x 0.10m). The runes occupy 25" (c.640mm) of the narrow surface and can be divided into three sections. The runes in the left-hand section are almost completely worn away, the ones in the middle are better preserved and the ones to the right are 'quite distinct' [RCAHMS loc. cit.]. From what remains it can be seen that the three groups are carved in different 'hands' which, combined with the differential wear pattern of the three sections,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY (GENERAL)
73
suggests they are of different dates. Only the right section has been read - -filibusranru. This has been interpreted as Philippus r(eist) enn rú(nar) 'Phillippus carved (these) nines' [Olsen 1954:164], though Barnes characterized this as 'a very uncertain guess' [1993:35]. The form of the nines dates to the tenth or eleventh century [NMS label]. The text(s) on this stone are of particular interest in comparison with the Birsay oghams because like them, they are best interpreted as informal graffiti. The stone was doubtless dressed for inclusion in a building as a lintel or similar. Whatever its precise location it must have been sufficiently accessible for passers-by to carve their names. When the building became obsolete the valuable ashlar would have been put to good use in the construction of the new chapel. This scenario, which implies a degree of casual runic literacy at the Brough, is very similar to the ones put forward below for the ogham-slabs.
The two other rune-inscribed stones, which came to light in 1934, are more formal in content and form. During the clearing-out of the cemetery area by the Office of Works, fragments of runic inscriptions were discovered including two which had been split longitudinally and re-used as building material in the Norse chapel [Radford 1959:18]. In both cases only the lower pan of the inscription survives. The extant text of Birsay 2 (Barnes OR 8) consists of the lower halves of about a dozen nines and two dividers in the form of xs. Barnes would go no further than to identify one rune as U or R [13], Liestol attempted a full reconstruction in the belief that the letters were the beginning of the common formula -reisti stein henna eftir X raised (this stone after Y)' [Liestol 1984:227 fig.73]. Even less of Birsay 3 (Barnes OR 9) survives, little more than the bottom third of the ten or so characters. Again Barnes would venture no more than that one rune was 'probably R* [1993:40], while Liestol thought that 'a fairly safe reconstruction of most of the nines' yielded -Pina after w-, ie. pan of the same formula 'X raised (this stone after Y)' [Liestol 1984:226fig.72].
The fourth runic text from Birsay (Birsay 4, Barnes OR 11) is the famous nine-inscribed seal's toothpendant [Curie 1982:59-60 No.253, ills.37, 38. See fig.]. The text consists of the first six letters
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY (GENERAL)
74
of the futhark very lightly incised and, since the carving is invisible at a distance of more than a few centimetres, Curie interpreted it as amuletic rather than ornamental. This item too was recovered in the pre-War excavations conducted by the Office of Works and, according to Curie, was found in the paved passage 1 in Area II (the same context as the allegedly ogham-inscribed pebble). The form of the runes cannot be dated any more closely than 'pre-ttOO' [Curie 1982:60].
Birsay 1 was presented to the NMS by the island's proprietor shortly after its discovery and is currently on display in the museum [NMS 191]. Birsay 2 and 3 were left behind in the Birsay site museum when the other finds were moved to the Tankerness House Museum in the 1990s. Birsay 4 is on display in the
NMS.
SELECT BIBLIOGRAPHY (Site Only)
OS Record card - HY 22 NW RCAHMS 1946: Radford 1959; Cniden 1965; Morris 1981; Curie 1982; Curie 1983: Ritchie 1983; Thomson 1983; Ritchie 1985; Hunter 1986; Ritchie 1989:52-55 Morris 1991.
OGHAM INSCRIPTIONS FROM THE BROUGH OF BIRSAY DISCOVERY Three ogham-inscribed building stones have been recovered in the course of archaeological excavations on the Brough of Birsay. The most recent, and best documented of these discoveries, occurred in the summer of 1980 during excavation by the University of Durham under the supervision of Christopher Morris [Morris 1981:36]. The earlier two were discovered in the excavations before the Second World War, but, as Ritchie has pointed out [1985:192] there is some contusion over their precise find-spots. Cniden
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY (OGHAM)
75
describes one, which Padel labelled Birsay 1, as having been re-used as a building stone in Thorfínn's Palace* [1965:25] (perhaps confusing it with the rune stone found in 1921 built into the chapel wall?). In the official guide to the site [Radford 1959:5], Radford stated that one ogham came from the churchyard (i.e. the same area as Thorfinn's Palace1), but in a later publication [Radford 1962:174] refers to only one ogham from Birsay and says it was 'picked up on the beach below the Brough'. Full publication of the site was delayed for over 40 years and in her catalogue of finds Curie makes no mention of the first two ogham stones, she does however note the discovery of Birsay 3 [Curie 1982:86]. In addition to the three genuinely ogham-inscribed slabs a small pebble recovered from the Lower Norse horizon is incised with lines which Curie suggests may be ogham [1982:120, see Appendix - Dubia for details].
The existence of each of the various inscriptions has been mentioned in passing in a number of works [see bibliography for each stone] but all are otherwise unpublished. Padel covered Birsay 1 and 2 in his unpublished MLitt thesis. At the time of writing, Birsay 3 is still in the Department of Archaeology of the University of Glasgow. The current whereabouts of Birsay 1 and 2 are unknown. They were in Edinburgh in 1971, when Oliver Padel examined them in Argyle House, Lady Lawson Street, the then headquarters of the Dept. of the Environment [Padel, pers.comm.]. This is the last record of their existence. Historic Scotland, the successor to the Dept. of the Environment, has been unable to find them, or the originals of the drawing and rubbing which constitute the only extant visual record of the inscriptions, the photographs of which are reproduced here [Richard Welander, Inspector of Ancient Monuments, pers.comm. after protracted correspondence with author]. I have made extensive enquiries to museums and individuals and have been unable to gain any further information, except to rule out the National Museums of Scotland, the Tankerness House Museum, Kirkwall, and the site museum at Birsay. It is, of course, deplorable that such important items should have been lost so recently while in State care. Padel*s account, the only detailed description, is thus of the greatest importance. My discussion is based on his thesis and private notes. I am most grateful to Dr Padel for making the latter available to me.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY
BIRSAY i
76
1
As discussed above, this stone, dubbed the 'Wall Stone' by Padel, appears to have been found in a secondary context, reused in the eleventh century structure known as 'Thorfinn's Palace', though there is some doubt over the precise circumstances of its recovery. Specifically, it is not clear if the identification of the context as 'secondary' is based on the archaeological fact of the stone's position, i.e. if the inscription was turned to the interior of the wall it would have been inaccessible, therefore the ogham would have had to have been carved before the stone was put in position, or, if it is based, solely on the assumption that an ogham must date to the Pictish phases and therefore an earlier use of the stone. Whilst the later may have been in line with then current thinking on ogham, it can no longer be supported. The bi-lingual ogham- and rune-inscribed slab from Killaloe, Tipperary [CIIC 54], has been dated to the second half of the eleventh century and there is no a priori reason to reject the possibility that Birsay 1 is also the product of a similarly mixed Celtic/Scandinavian milieu of the same period. The following account is based on Padel [1972:56-58, and private notes] and the Department of the Environment photograph A 2894-2 [see fig.].
DESCRIPTION OF OBJECT A squarish flag-stone inscribed along the narrow face ('the only really flat surface' [Padel 1972:56]) with an ogham inscription. Otherwise undecorated. Stone:
Orkney flag
Dimensions:
24tew x 12" x 4V£W (i.e. approx. 0.63 x 0.31 x 0.12m)
Condition:
Fragmentary, 'much chipped' [Padel MS notes]. According to Padel the stone has been broken and the two pieces cemented together.
This, the largest of the three Birsay ogham-stones, is a building slab, most probably carved along its only exposed surface when in situ in a wall, rather than erected specifically to bear an inscription. Unless further information comes to light it is not possible to determine whether the carving occurred while the stone was part of Thorfinn's Palace' or, rather while the stone was part of a much earlier construction.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY 1
77
CARVING TECHNIQUE The photograph supplied by Padel is of a rubbing of the stone, rather than the surface of the stone itself [see fig.], therefore it is difficult to discuss the technique or condition of the carving in detail. Padel commented that the technique was very similar to Lunnasting and Formaston [MS notes]. The inscription starts several inches in from the left edge and runs parallel with the long edges to within an inch or two of the right edge. Thus it is more or less centred vertically but is more extensive to the right on the horizontal axis. The stem is \9}á" (c.495mm) long though the inscription occupies only Wá" (c.370mm). Padel interpreted this as indicating a lack of planning. The lettering, however, is placed centrally on the stem, so the ample length of clear stem at either end may instead be a stylistic feature. The far end of the stem is not definite, but seemed to Padel to end before the edge of the stone [MS notes]. Padel comments on the air of 'hurried casualness' of the carving of Birsay 1 [MS notes].
DESCRIPTION OF OGHAM INSCRIPTION Three features combine to give an unambiguous indication of the direction of reading - the forward slope of b- and h-aicme consonants, the hammer-head A (8) and several angled vowels. The letters are generally fairly tightly spaced, though successive letters of the same aicme are clearly differentiated, i.e. 314, 6/7, 10/1L The spacing to either side of the hammer-head A (8) is more generous, perhaps to give room for the cross-stroke. The gap after the first letter is wider than might be expected, unless Padel's 'false start' stroke is allowed. The first few inches of stem appear to lack lettering. Padel rejects two strokes either side of the cemented crack even though 'they appear to be of the same technique as the inscription' but 'cannot be fitted into any rational scheme, as far as I can see' [1972:58]. The inscription continues as follows [see fig. for key to numbering]: 1
A single stroke across the stem. The upper (left) half is perpendicular to the stem but the lower (right) slopes forward at an angle of about 45°. The two meet, not exactly on the stem, but slightly below it. The other m-aicme letters are straight across the stem so the interpretation M is not appealing. Padel suggests a 'feather-mark', i.e. a non-phonetic directional indicator —>—,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRS A Y 1
78
but there are no extant examples of this asymmetrical form. It could be that the carving represents an A, begun as a simple perpendicular cross-stroke but modified into an angled vowel. Without examining the stone it is impossible to offer a definitive interpretation. Padel mentions a second oblique stroke to the right of the stem just beyond this letter which he says is artificial, though not visible on the photograph. This he interprets as a possibly false start on the next letter since, in his opinion, it cannot itself be a letter. It would be hard to account for the larger than expected gap between this letter and the next unless there was some such problem. 2
Two angled vowel strokes across the stem pointing in the direction of reading - O. The second of these wavers to the right of the stem. According to Padel this part of the stroke is carved double implying the carver tried to improve it.
3
Five strokes to the right of the stem sloping forwards - N. The first of these has been recut and the second continues across the stem at a more perpendicular angle, in other words this stroke closely resembles L Padel thought the protrudence probably accidental, which does not seem to have been the case with 7. The spacing of the individual strokes suggests they are to be taken together as a single group. The strokes are fairly evenly spaced and parallel (to the right of the stem) though each is successively longer than its predecessor. If the stroke on the far side of the stem is not accidental it might have been added to rectify an earlier mistake, either to insen a missing letter (H?, A?) or to separate the five strokes into 2 + 3 .
4
As above, though not quite so clear - N. The first stroke of the group is at a slightly more acute angle than the rest, which may, as Padel suggests, be a deliberate attempt to differentiate this group from the previous one. Padel says the third stroke does not quite reach the stem though this is not clear from the rubbing, nonetheless the interpretation is not in doubt.
5
Two angled vowel-strokes across the stem pointing in the direction of reading - O. Padel described this letter as careless, since, rather than meet at a point on the stem, the two pans of the first stroke overlap in a little cross just below it. Though feint in the photograph, Padel says the letter is not in doubt.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY 1
79
6
Five long oblique strokes across the stem, closely spaced - R.
7
As above, with the first stroke having been re-cut - R. The two letters are very clearly spaced.
8
A single stroke perpendicularly across the stem with a horizontal cross stroke towards the right distal end. Even though the cross-stroke is not at the tip of the stroke there can be no doubt that a 'hammer-head A' was intended, if carelessly carved.
9
Six strokes to the right of the stem sloping forward. The first stroke has been recut. According to Padel the third is feint and does not quite reach the stem, though again this is scarcely discemable from the rubbing. If the stroke were discounted a reading LV would be possible. Padel suggested that the carver may have begun to carve the V too close to the L, but realized his mistake before completing the stroke, hence its faintness. If one were not prepared to discount this third stroke one could observe that the third and fourth strokes are very slightly more generously spaced than the rest, which could be taken to imply 3+3. i.e. VV. A third possibility is that the oghamist has simply miscounted, that the sixth stroke is superfluous and that the letter is N. There are thus three possible readings - N, VV, LV - and it would be unwise to privilege one above the others without examining the stone. According to Padel the elongation of the fourth and fifth strokes, apparent on the rubbing, is an illusion.
10
Five long oblique strokes across the stem, sloping forwards - R. Padel read the middle stroke of this group as having 'bent tips', in other words being the character //J7/, found on Birsay 2 and Altyre. The alleged flaps are not immediately apparent on the rubbing. In his notes he describes the third stroke as 'very curved', but thought it 'very possibly not intentional'.
11
Five long strokes across the stem, sloping forwards - R. More generously spaced than previous letters but clearly a unitary group.
The stem continues for a further inch or two, apparently without any further letters. Padel identified a 'very indefinite' stroke to the left of the stem and continuing across it for a shon distance. Since there is nothing either immediately before or after it Padel was happy to reject this stroke. Which gives the following reading:
(m/h/a)ONNORRA(n/vv/lv)RR
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRS A Y 1
80
FORM OF SCRIPT The form of the script used is a simple though slightly developed form of Type II, i.e. lib. The more developed features being that all the letters slope forward, the vowels are angled, and there is one forfid, the hammer-head A. More simple features are the lack of marked spacing between letters and the lack of bind-strokes, apart from the seriffed A. The lack of word division may not be significant if, as appears likely, the text is a single word, a personal name. The letters are 'squarish' in outline, in other words a five letter stroke has a similar vertical and horizontal length.
INTERPRETATION OF TEXT The text appears to be complete, unless the strokes Padel identifies either side of the cemented crack and at the very end turn out to be ogham letters. Since the site was inhabited for such a long time and the archaeological context of this stone is so vague, the date of this inscription could lie, potentially, any where from the early Pictish period through to the later Norse, say sixth to twelfth centuries. Because of this long date range and the known cultural affiliations of the site, the text could be in Irish, Pictish (i.e. Brittonic) or Norse.
The (m/h/a)ONNORRAvvRR
might represent
(m/h/a)onnorrqffrr or
(m/h/a)onnorrauurr. The final R(R) is reminiscent of nominative ending of a Norse male personal but value /v/ for V seems unlikely, thus the Norse male personal names Ónarr and Ávarr [Lind 1905 s.n.n.] are unlikely to offer a solution. Nonetheless, Padel suggested the text might represent Old Norse and offered ON Norrcenn 'Norse, Norwegian', though, as he himself admitted the interpretation is unlikely, not least because the adjective would not end in -r. Likewise his suggestion that 0 is ON 6 'from' is not very likely.
If the text is read (m/h/a)ONNOR RANRR or even (m/h/a)ONNO RRANRR it might be compared with the runic filibus ranru of Birsay 1, with the gemination of the R merely a hang-over from traditional ogham orthography. Regardless of whether or not Olsen's expansion of ranru to r(eist) enn rit(nar) 'carved these runes' [1954:164] is accepted (and as explained above, this is by no means certain), it is not inconceivable
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY 1
81
that the ogham text may bean attempt to render that sameformulawith ogham letters. This would leave (m/h/a)ONNO(R) as a personal name. Unfortunately, none, either Celtic or Norse, immediately suggest themselves.
DISCUSSION See general discussion of the three Birsay ogham-slabsfollowingthe entry for Birsay 3.
BIBLIOGRAPHY Radford 1959:5 (mention only); Cruden 1965:25 (mention only); Padel 1972:2, 10-11, 16, 55.
PREVIOUS READINGS Padel
BIRSAY
mONNORRA(lv/n)RR
2
As discussed above, this stone, dubbed the 'Shore Stone' by Padel, was 'picked up on the beach below the Brough' [Radford 1962:174]. Padel raises the possibility that the stone is water worn [MS notes] which may indicate that it had lain on the shore for some time before discovery. Presumably this was a secondary context and the stone had tumbled down from its place in some structure on the Brough proper. Unfortunately the precise find-spot seems to have gone unrecorded so we cannot begin to investigate from which structure it might have fallen. This leaves us without any indication as to the date of the inscription. As with the larger of the two early discoveries the ogham could lie anywhere from the sixth to twelfth centuries and relate to an Irish, Pictish (Brittonic), or Norse-speaking milieu. The following account is based on Padel [1972:56-58] and the Department of the Environment photograph A 2894-2 [see fig.].
DESCRIPTION OF OBJECT Described by Padel as a Very irregular slab', on its flat edge survives a weatheredfragmentof ogham text.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRS A Y 2
Stone:
Local flag-stone
Dimensions:
14#" x 9lá" x 2" (approx. 0.37 x 0.24 x 0.05 m)
Condition:
Badly weathered, text incomplete.
82
Padel does not illustrate the stone nor give any indication of the relative position of the inscription on the stone, other than to say it is on the 'flat edge1. Though he describes the inscription as very weathered, Padel does not discuss the appearance of the stone itself, nor whether or not it is intact. It is probable that, like Birsay 1 and 3, it is a building slab inscribed in situ along the only visible surface. However, until a photograph can be obtained or the stone examined in person, I cannot rule out the less likely possibility that it is the remains of a small slab, either recumbent or upright, erected specifically to carry the ogham text.
DESCRIPTION OF CARVING The photograph provided by Padel is not of the stone itself, but rather of a drawing, probably based on a rubbing. For this reason it is not clear which marks are strokes and which the result of casual damage. Also the relative weighting of strokes is unclear making it uncertain how much importance is to be attached to small additional lines etc. The stem is visible for approximately 3'4" (c.80mm), but, according to Padel, die stone is too badly weathered to indicate how much has been lost. The scale of the letters and the reading, as best as it can be discerned, imply that this is only a short section of a longer text. There is no clue however, as to where in the sequence this portion occurred. The stem wavers and appears to be more finely incised than the letter-strokes. It begins at the point of hypothetical intersection with the first stroke of 7, but it is unclear if this is a clear terminus or just where the text emerges from wear. Padel says that the stem 'seems to continue crookedly beyond lower edge?1 without specifying how far this is from the patch of wear. The direction of reading is indicated by the slope of 2. The extant portion is as follows [see fig. for key to numbering]: 1
The remains of a single oblique stroke below the stem, sloping forward. The stroke does not quite meet the stem, presumably because of wear. As it stands it should be read as be B. Padel
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRS AY 2
83
notes, however, that the surface opposite this stroke is worn away. If the stroke originally continued at the same angle across the stem it would be M, though if produced for an equal distance on the opposite side it would come close to intersecting the first stroke of 2. Padel makes the interesting suggestion that the stroke could be the lower half of an angled A (->*). Angled vowels appear on both the other Birsay oghams and the objection that letter 3 is a straight-stroke vowel is not compelling since the two forms co-exist on Bressay, Burrian, Cunningsburgh 3, Formaston, and Lunnasting. If strokes have been lost from the beginning, it is, of course, possible that this stroke is thefinalone of any letter in either the b-, m% or angled vowel-aicmi. There follows a small nick parallel with the first stroke, also not touching the stem, which Padel thought 'probably due to weathering'. If it is the remains of a letter it could be a second stroke similar to the first. The above arguments would apply equally, though it is even less likely to be an m-aicme letter since it would intersect with the first stroke of 2 almost as soon as it crossed the stem. Though this second stroke appears to be at an angle, such a short section survives that it is just possible that it is the remains of a short straight stroke across the stem - A. This would facilitate the reading MAQI, but would strain the bounds of probability given the limited space. Padel suggests it may be a feather mark [MS notes] 2
Five strokes above the stem, sloping forwards - Q. Only the pan of the first stroke nearest the stem survives, though Padel notes that it is not as short as appears from the DOE photograph and the reading is not in doubt. The strokes are neither entirely straight nor parallel and each overshoots the stem slightly. This gives the impression of an inscription carved quickly and with confidence.
3
Five long strokes across the stem - 1 . These are not balanced on the stem but extend further below than above. The middle stroke appears slightly curved, Padel says it is carved double. It may be just a botched stroke, or alternatively the 'flapped* or 'angled S-shape' stroke which appears as a singleton on Latheron, and at Lochgoilhead and possibly Altyre in the middle of a group of five on Birsay 1. There follows a dot close to the stem. Padel thought it might be one
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY 2
84
of a pair of word-dividers, like those on Bressay, and noted that there was a mark opposite, though he conceded it could be just weathering. Another possible interpretation is that the dot is intended to differentiate between two successive letters of die same aicme. The analogy for this is Lunnasting character 16, though in die Shetland example the dot has been interpreted as having been added later to clarify two strokes carved too close together. The Birsay letters are already well-spaced, the gap between them is already about two and a half times that between 2 and 3. Without seeing the actual stone it cannot be ruled out that the dot is merely die result of casual damage. 4
A single stroke across the stem - A. Since the h-aicme strokes (2) are clearly sloped it is unlikely that m-aicme consonants would be perpendicular to the stem, hence the reading M is rejected. From the photograph it appears that three short, thin, oblique strokes are appended to the left of the lower tip of this stroke. If they are genuinely part of the inscription they are without parallel in the epigraphic ogham corpus. A suggested interpretation is that the oghamist has inserted a three stroke letter of the h-aicme (T) which was omitted in the initial carving. Padel simply says the letter is carved with several strokes, so perhaps the photograph is misleading.
5
After a gap at least as long as that between 3 and 4 there is a single oblique stroke below, but not quite touching the stem, sloping forward. Padel notes that the surface opposite is worn away and that the stroke may have continued across. Like the first stroke, it could be pan of any letter of the b-, m-, or angled-vowel-a/c/w. Clearly, some text has been lost from this end, but there is no indication of how much.
The following tentative reading is possible: -(b + /a + /m + )(a/ )QI(:/ )(t/ )A(b + /a + /m + )-
FORM OF SCRIPT Again, since the extant inscription is so short, there is little to say about the form of the script other than that it is simple Type II, broadly similar in aspect to that of Birsay 1. The flanged letter, if that is what
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRS A Y 2
85
it is, is an obvious connection between the two, but they also share similar over-all proportions in lettershape and spacing. If the dot between 3 and 4 is accepted as a word-divider then this would imply a leter date, possibly in the Norse period (see Introduction for discussion of word-dividing dots).
INTERPRETATION OF TEXT Since the text is so shon and doubtful there is little that can be sensibly said about it. Certainly it can neither be proved or disproved to be in Irish, Pictish, or Norse. The aQI could be Olr. MAQI, and if so would be the only Scottish example of this word spelled with single Q.
DISCUSSION See general discussion of the three Birsay ogham-slabs following the entry for Birsay 3.
BIBLIOGRAPHY Radford 1962:174 (mention only); Padel 1972:55-61.
PREVIOUS READINGS Padel
BIRSAY
-(b/a)QI(:/ )A(b/a)-
3
The third ogham inscribed flag-stone from the Brough (BB 80, layer 543 ref.2761) was unearthed during the Summer of 1980 in Site IV S. The stone had been re-used in an area paved with large sandstone flags. No other finds were recovered from this context which was sealed 'by one of the rubble spreads between Sites E and S' [Morris 1981:36]. Calibrated radiocarbon dates obtained for this context were in the range 560-769, centering on c.648. Thus the secondary context of the ogham can be dated approximately to the seventh century, plus or minus 50 years. I examined the stone in the Department of Archaeology, University of Glasgow in March 1994, and most gratefully acknowledge the information and assistance
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRS A Y 3
86
provided by Prof. Christopher Morris on that occasion. Prof. Morris also kindly supplied me with contact prints of the stone [see figs.].
DESCRIPTION OF OBJECT A parallelogram-shaped building stone, undecorated apart from a short inscription in bind oghams [see fig.]. Stone:
Local - Stromness flag series (orange brown calcereous siltstone, ferric rich)
Dimensions:
0.31 x 0.35 x 0.065 m
Condition:
Flag intact. Only middle section of inscription survives. Some individual strokes badly worn, top of third letter lost in spall.
The inscription is incised along the flattest of die narrow feces. This is not a continuous smooth surface but rather disjointed. There is a marked disparity in the level of the surface immediately after the final stroke of 4 which evidently pre-existed the carving of the inscription. Other spalls and fractures post-date the carving which has suffered much. Immediately following the single stroke of 7 the surface has flaked off taking with it the outer tip of the stroke and any subsequent letters. Also gone is the distal portion, and presumably the bind stroke of the last three strokes of 4. Surface abrasion has worn away parts of all letters except for 3 and 5, so a measure of doubt surrounds all but these.
DESCRIPTION OF CARVING The stem is visible for 190mm, but may have continued further in both directions. It runs roughly parallel with the long edge, wavering slightly. The maximum breadth of ogham-inscribed face is 50mm and the ogham stem is placed not medially, but only about a third of the way up the narrow edge. The strokes have been sharply and fairly deeply incised with a blade, the stem a little more substantially carved than the letter strokes. The direction of reading is not in doubt, since the slope of all letters but the first is diagnostic. The inscription is as follows [see fig. for key to numbering]: 1
The extant portion of the ogham begins with what appears to be a single long oblique stroke
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY 3
87
across the stem. The right hand portion is worn. There is some doubt over this stroke because die stone has suffered much damage in this area. There is nothing left of the surface immediately before this stroke and it is possible that some ogham has been lost from the beginning. Possibly M. 2
Between 1 and 3 there is a gap of 45mm. After the first 30mm there is the remains of what appears to be a stroke. The only portion to survive is the bit on the stem itself, though there may be a faint stroke above. If this is a letter it may be A. Half-way between this stroke and J is a faint stroke to the right of the stem. The gap is less than that between the other letters and this slight stroke is highly doubtful. If extended across the stem it would intersect with the first stroke of J, for this reason it cannot be interpreted as the remains of a vowel stroke. Possibly B.
3
Five strokes to left of the stem, roughly parallel, sloping backwards, bound - Q. Each stroke is slightly longer than the last (in the range 15-30mm) giving the whole a wedge-shaped appearance. The first stroke protrudes slightly beyond the bind stroke, which itself over-shoots the first stroke to form a tiny cross at the intersection. This lack of precision is reminiscent of the Burrian ogham and both it and Birsay 3 have the air of being carved slightly carelessly, but with confidence and speed.
4
Five strokes to the left of die stem. A substantial chunk of the surface has been lost at this point, taking with it between a third and a half of the distal ends of the third, fourth, and fifth strokes. On die published photograph [Morris 1983 fig.40] there appears to be a bind stroke along the tops of the last three strokes, in fact this is just the shadow cast by the broken edge and there is no bind stroke on the stone at this point. From a casual glance one gets the impression that the first two strokes are parallel with each other but not with the final three, suggesting the reading 2 + 3 DT. A closer examination, however, will reveal that the first and fifth are exactly parallel and it is the intervening strokes which are at various angles. A look back at 3 shows that the oghamist took no particular care to keep the strokes entirely straight or parallel so, though 4 is even less regular than 3 it is not unacceptably so. Since the bind-stroke continues after the second stroke
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY 3
88
as far as the break there seems no justification in not taking the five strokes as a single group Q. After the fifth stroke there is a break in the surface of the stone which clearly pre-existed the carving of the ogham. The inscriptions continues at the lower level. 5
A single stroke to the left of the stem, sloping backwards - H. This score is rather deep and not in any doubt.
6
Two pairs of strokes either side of the stem meeting to form a forward pointing arrow-head (> >•) - O. The first two strokes do not quite meet, the right hand stroke intersecting with the stem slightly beyond the left one, but the discrepancy is slight, certainly not enough to warrant interpreting them as separate letters (HB), and further suggests the rapid and facile execution of the inscription.
7
After a gap of 18mm a single stroke to the right of the stem. The surface is fractured immediately after this stroke and any subsequent carving has been lost. There is a faint shadow to the left of the stem opposite this stroke but as it stands it is insufficient to warrant explanation as the upper half of an angled vowel-stroke. This letter is best interpreted as B or any subsequent letter of the b-aicme (i.e. L, F/V, S, or N).
It is unlikely that we have the entire original inscription. Approximately a further 60mm of stone would have been available originally, sufficient for two to three letters. Even that would be surprisingly, though not improbably, short. However, if the ogham was indeed carved when the stone was in situ in a wall it is possible that the text continued across the join and onto the neighbouring stone. The tentative reading is as follows:
-m(a/ )(b/ )QQHO(b+)-
FORM OF SCRIPT Despite the brevity of the surviving inscription Birsay 3 displays a number of significant features, the bindstrokes on the consonants and the angled vowels. Both these features would normally be considered to be late in the typological sequence of ogham development. The comparatively early radio-carbon dates obtained for this item force a re-examination of the absolute dating of the typology. The Birsay 3 dates
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY 3
89
centre in the mid-seventh century, though range from the mid-sixth to mid-eighth, requiring that the typological sequence be stretched back in time. That this is necessary underlines the fact that script typology can be a guide to terminus post quern dating only.
INTERPRETATION OF TEXT With a text as fragmentary and doubtful as this one it is difficult to make clear assertions as to meaning. The two most likely readings are -mabQQHO(b+)- and -maQQHO(b+)-. The first holds out the intriguing possibility that is starts MAB, i.e. Brit, mab < map-son*. That maplmab was also the Pictish word for 'son* is suggested by the annalistic reference to the death of Talorcan maphan [AU 725]. Thurneysen took this as a form of maban with ph for pp corresponding to the cc of Olr. mace, the gemination due to affective intensification [Thurneysen 1946:92-3 § 150]. A. O. Anderson's alternative explanation is that maphan is for map-hoen 'son of Owen* [1922:222 n.7]. Either way the word attests to the use of the Common Celtic word for 'son* among the Picts. To read MAB at the beginning of Birsay 3, however, would leave the awkward cluster QQHO. In any case the reading is highly doubtful and the more obvious interpretation -maQQHO(b+)- is to be preferred. MAQQ can be taken as a fonn of Irish MAQQI > mac 'son', the same spelling of this word being attested elsewhere in Scotland at Latheron (Dunnodnatt Maqq Net-) and probably at Formaston (Maqq[?]Talluorrh). As far as I am aware, however, MAQQ is nowhere attested in Ireland, the post-apocope forms of MAQQI being MAQ and MAC not MAQQ and MACC [McManus 1991:124 §6.30]. This suggests that Birsay 3 is not a purely Irish text, like its near neighbour and younger contemporary Buckquoy, but rather is in some way assimilated to the local language (presumably Pictish).
If there is a portion of the lettering lost from the beginning, then Birsay 3 may be an example of the 'X MAQQI Y' formula so familiar from Ireland, and found in Scotland at Latheron and elsewhere. Possibilities for the Y slot, personal names beginning 0(b+)n are Ok, Olchobur, Olchu, Ollam, Ossán,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
Ossine, and Onchu, to name a few listed by O'Brien [CGHisrm.].
g IRS A Y 3
90
This, however, would leave 5
unaccounted for. There is considerable uncertainty over the correct transliteration of the first letter of the h-aicme. McManus flavours a hypothetical /j/ lost before the Old Irish period [1986:26,1991:36-7], SimsWilliams favours a Latin-derived h [1993:162-70]. Though the character never appears on classical Irish ogham pillars, in the later, say ninth century and later, oghams, which are merely ciphers for standard Old Irish manuscript spelling, it is used for the letter h. This might explain the presence of H in Birsay 3, a mute letter before a vowel [Thuraeysen 1946:19-20 §25]. In Scotland, however, ogham 'H' occurs in situations, such as NEHHT- and TALUORH where one might expect a spirant, either fXJ or ///. This explanation is less easy to tally with the evidence of Birsay 3 since we would not expect lenition in this position.
One problem with interpreting Birsay 3 as a 'X MAQQI Y' type text is that there is scarcely enough room for even two additional letters before 1. For this reason, unless one is prepared to imagine the inscription beginning on a previous stone and continuing onto this one, the (X) MAQQ HOb* interpretation has to be abandoned. Rather than a patronymic phrase, we may have a solitary compound personal name of the common Irish type Mac-N. The correct transliteration of H is still as much of a problem but there is a saint of the Ui Bairrche called Mac-Onchon recorded in the genealogies of the Lagin in the Book of Leinster [CGHA6, §121 a 42] and the same name may be behind Birsay 3.
Birsay 3 displays a number of features typical of Type lib oghams. The sloping of consonants, clear spacing between letters, and, of course, bind-strokes are all widespread aids to legibility. The angling of vowels may be a further example of attempts to enhance legibility, by adding another feature to the list differentiating vowels and consonants, but it is possible that they represent nuances of vowel sounds not covered by the standard straight vowels. Nothing can be read into the lack offorfeda since the fragmentary sample is so short. The prime significance of the script of Birsay 3 is in demonstrating that bind-strokes
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRSAY 3
91
and angled vowels were in use already by the mid-seventh century or thereabouts.
DISCUSSION See general discussion of the three Birsay ogham-slabsfollowingthis entry.
BIBLIOGRAPHY Morris 1981:36,40 (mention only); Morris 1982:81 pl.41 (ill. upside down); Curie 1982:86 (mention only); Morris 1996:54-5 (quoting author) fig. 3.8 (right way up).
PREVIOUS READINGS None
GENERAL DISCUSSION OF BIRSAY OGHAMS The size and shape of the three Birsay slabs and the position of the text on each stone suggests that the Birsay oghams were carved while the stones were in place as constructional elements in walls or other structures. The inscriptions are most easily classed as informal graffiti. Birsay 1 appears more or less complete, but Birsay 3, despite covering most of the available surface, is so shon that it is at least possible that the text began and/or continued on a neighbouring slab.
The three texts are carved in different hands and though Birsay 3 is firmly dated, there is no reason to assume that all are contemporary. Nonetheless three oghams from one site is a remarkable concentration, especially if one adds to the total the whorl from nearby Buckquoy. These four oghams constitute twothirds of the total of Orkney oghams, but whether this accumulation reflects greater Irish influence at Birsay, or higher levels of literacy, or merely the intense archaeological attention lavished on Birsay for over fifty years is unclear.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BIRS AY 3
92
Since none of the Birsay slabs were found in their primary position it is impossible to know if the oghams ran horizontally or vertically. Given the shape of the stone we might presume die former. Carving ogham horizontally represents a major departure from the classical norms of ogham inscriptions, but if the inhabitants of Birsay were familiar with non-lapidary ogham, as, for instance, the circular Buckquoy ogham or the inscribed knife-handle from Gurness, then the convention may no longer have been that strong. Lack of comparable Irish evidence prevents an evaluation of whether such features were innovative and specific to Scotland or had always been part of an Irish chattel-ogham tradition parallel with the classic pillars.
ORKNEY ISLANDS
LOCATION MAP - Orkney with inset Brough of Birsay and Point of Buckquoy [After Morris 1990:70 fig.3]
BIRSAY - Plan of main Pictish and Norse settlement [Curie 1982 ill. 3]
BIRSAY - Sculptured slab with Pictish symbols and figures [Curie 1982 ill. 4]
BIRSAY - Cross-incised stones: Large [Ritchie 1986:12] Small [Curie 1982 ill.45 (detail - no.607)]
BIRSAY - Rune-inscribed fragments, Birsay II (Barnes OR 8) [photo. Historic Scotland, Neg. A 995-1 (1938), Crown Copyright]
BIRSAY - Rune-inscribed fragment, Birsay III (Barnes OR 9) [photo. Historic Scotland, Neg. A 995-3 (1938), Crown Copyright]
0
50 mm
BIRSAY - Rune-inscribed seal's tooth, Birsay IV (Barnes OR 11) [Curie 1982:ill.37]
BIRSAY 1 - Rubbing of ogham inscription [Photo. Department of the Environment (Historic Scotland) No. A 2894-2, June 1971 (Crown Copyright)]
^Méw^'^L'W'íptiP^i BIRSAY 1 -Ogham inscription [Padel 1972:56]
<£*•**
1 2
3
4
5
6
7
8
10
////
11
*
C/t'/
BIRSAY 1 - Ogham inscription (key to numbering)
BIRSAY 2 - Rubbing of ogham inscription [Photo. Department of the Environment (Historic Scotland) No. A 2894-1, June 1971 (Crown Copyright)]
/^MW^ BIRSAY 2 - Ogham inscription (key to numbering)
BIRSAY 3 - Ogham-inscribed slab, front and back faces [photo, supplied by Prof» C. D. Morris, Dept. of Archaeology, University of Glasgow, Reg. BB80]
BIRSAY 3 - Ogham-inscribed slab, upper and lower faces [photo* supplied by Prof. C. D. Morris, Dept. of Archaeology, University of Glasgow, Reg. BB80]
BIRSAY 3 - Ogham inscription
N BIRSAY 3 - Schematic representation of ogham inscription (key to numbering)
93
BLACKWATERFOOT
1-2
SITE The King's Cave at Drumadoon, in the parish of Kilmory, Arran [NGR 8844 3092], is the largest of a string of caves in the sandstone cliff about 3km north of Blackwaterfoot. The name Dmmadoon (Druim an Dúin, 'ridge of the ton') comes from the large structure on the rock above. Mania however, refers to it as fDruim-cruey' [1703:219]. Situated on the western shore of the island, the cave has a commanding view of the coast of Kintyre about 10km away. The name 'King's Cave' derives from the legend that Robert the Bruce encountered the spider here [Mackenzie 1914:31, 113]. The tradition is, however, comparatively recent: to Martin and Pennant this was Fingal's or Fin mac Cool's Cave. Inside it is large and irregularly-shaped, being about 37m by 9m and about 15m in greatest height. Towards the back, before the cave narrows to a point, a rock buttress divides it into two aisles [see rig.].
The walls of this scheduled monument are decorated with a number of incised carvings which date from perhaps as early as the Bronze Age to modern times [OS record card]. There are no early carvings in the north aisle, but in the south there are horses, deer, and concentric circles. The buttress bears a large and a small cross and the figure of a man with raised hands holding a curved object above his head. Towards the entrance, on the south wall there are two triangles and on the nonh a group of serpent-like spirals and two ogham inscriptions. In addition to the ancient carving there is a great deal of modern graffiti, some of which may obscure older work, such as Martin's 'two-handed sword', which is no longer visible. Simple carving of this kind is very difficult to date. Some of the animals may be of Iron Age date [Thomas 1961:23], a proportion of the rest is doubtless from the early Christian period, but much of it is modern. In addition to the carving, there are opposing rows of small holes cut into the rock about 1 V£-2m above the present floor. These have been interpreted as sockets for small bars of wood. Martin records the local tradition that the legendary hero Fin mac Cool and his band used them for 'holding big Trees, on which the Caldrons hang for boyling their Beef and venisons' [219].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BLACKWATERFOOT
94
J. A. Balfour conducted a brief excavation of parts of the cave in 1909 [see fig.] but, apart from a fragment of bronze, found nothing more than ashes, some shells, and a few pieces of animal bone. The bronze item [current whereabouts unknown; ill. Balfour pl.XXXIII.2] is described as being engraved on one side, and though no measurements are given, was small enough for Balfour to suggests it was 'possibly a fragment of a fibula'. The simple interlace pattern with which it is incised suggests a Dark Age date, though the significance of this is uncertain. The fragment could have been swept into the cave at any time, or introduced in the material used to build the bank at the entrance in the later half of the nineteenth century [Balfour]. The cave has been much used over the centuries and for a variety of purposes. Its numerous recorded uses include being the school of the district for 'a considerable number of years' in the midnineteenth century, and the frequent venue for meetings of the Kilmory Kirk Session in the eighteenth [Balfour 216]. Martin records that it was the wont of the minister to preach in a cave a few miles to the south 'in regard of it being more Centrical than the Parish Church' [219], and it is not unlikely that church services were also held in the King's Cave. Today the cave is a seasonal habitation for Travellers.
DISCOVERY The animal and geometric carvings have been known since at least the end of the seventeenth century, but the presence of ogham was not noticed until about 1968. Jackson [1973a] says that he was told of the existence of the ogham 'a couple of years ago' by Kenneth Steer of the RCAHMS, and that they visited the cave together in September 1971 to examine, measure, and draw the inscription. Though only a short distance away, the second ogham was not discovered until 1992 when Ian Fisher of the RCAHMS was carrying out further survey work on the cave [pers.comm.]. That the second ogham went unnoticed is surprising, but should not necessarily cast doubts on its authenticity. The cave is full of natural fissures and copious graffiti and the weathering of the inscription is sufficient to prove that it is ancient.
The two confirmed ogham inscriptions are carved on the left-hand wall of the cave, about 2m above the present floor, the first about 9m and the second about 1 lm from the entrance gate [see fig.]. In the same
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BLACKWATERFOOT
95
area is a panel of snake-like, loose spiral interlace graffiti. That the two oghams are contemporary with one another or with any other of the carvings in the cave cannot be proved. Jackson's photograph shows what looks like another ogham to the right of the first and parallel with it, but he says it is a piece of modern vandalism, not carved into the rock but drawn in pencil or ink. He does not mention what appears in the photograph to be farther ogham-like marking to the left These may, however, be natural fissures in the rock face with only a coincidental resemblance to ogham. I have not examined the inscriptions personally, the following is based on Jackson's published account of the first ogham, and on RCAHMS photographs of the second.
DESCRIPTION OF CARVING - Ogham Inscription 1 The stem line runs vertically between two patches of wear, the lower 260mm long, the upper 84mm. Immediately after the end of the upper patch there is a clear cross-stroke where the stem appears to end. According to Jackson there is a similar stroke below the lower patch but this is not absolutely clear from his photograph. These two end-strokes are about 500mm apart. Nothing is visible in the upper patch of wear, in the lower there may be traces of occasional strokes, but these are very faint. In the central portion, between the areas of wear there are a series of fairly clear strokes between 30 and 40mm long. The visible section is as follows [see fig. for key to numbering]: 1
Jackson read four strokes across and perpendicular to the stem, with the part of the first stroke to the right of the stem very faint (E), though from the photograph it looks as if there could be five (I). Then a space.
2
Two strokes, wider apart, with aflakemissing between them giving the impression of a hollow O.
3
One long oblique cross stroke, diagonallyforwards- M. Jackson is doubtless correct to dismiss a little mark to the left of the stem, joining it to the M stroke.
4
Four cross-strokes perpendicular to the stem. According to Jackson the part of the first stroke to the right is 'faint but certain' - E.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND 5
BLACKWATERFOOT
96
Five strokes to the left of the stem - Q. The space between the second and third strokes is wider than between the others. Jackson thought it 'perhaps more probable' that these represented two letters (D T). From the photograph there appears to be a slight knob on the rock between the two strokes which could have influenced the spacing, but Jackson does not mention it. It is generally the case that when two letters from the same aicrne directly follow one another, and are therefore open to confusion, they are more carefully spaced than successive letters from different aicmi, which are less likely to be misread. Yet die space here is less than that between die first stroke of this letter and die last of the preceding one. For this reason I would prefer die reading Q, but since so little of this inscription survives it is difficult to judge how carefully it was carved and therefore to know how much to make of the spacing.
6
Four cross strokes perpendicular to the stem - E. From the photograph there appears to be a small parallel line after the last of these, which could be die remains of a fifth stroke (I) but it is impossible to be sure.
As Jackson pointed out, the condition of the stone does not rule out the possibility that letters have been lost from the beginning and the end of this inscription. Only a few would fit in the space at the end, but diere is room for at least half-a-dozen at the beginning. Jackson read: EOMEDTE or EOMEQE (or in the unlikely event of the text being read downwards EVLEMOE or ENEMOE). I would prefer. H(i/e)OMEQ(i/e)H
INTERPRETATION OF TEXT If die inscription is complete as it stands then it is difficult to interpret it in terms of a Celtic, or any other known language. If, as seems likely, letters have been lost from both ends, it can be interpreted as Goidelic: '([—]eol[—]io) Meqi ([—])', or '([—]eol[—]io) Meg (/[--]/£[—])\ in which case, it appears that we have something like die traditional formula 'X MAQQI Y \ with only the last two letters of the first name, and at most, die first letter of the second. The MEQ can be compared with the MEQQ of
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BLACKWATERFOOT
97
Bressay and St Ninian's Isle which may represent Olr, make (gen. of mace 'son'). As Jackson pointed out [1973:54], if the inscription is as late as the eighth century, the -eo could be the genitival ending of an /stem personal name, eg. Fedlimid, Fedlimtheo. Alternatively, it could be the end of the nominative of a name such as Reoy Búaidbéo or one of the compounds with second element Nia/Nio, such as Flaithnia and Maicnio.
DESCRIPTION OF CARVING - Ogham Inscription 2 The second ogham inscription is at a similar height to the first, about 2m further into the cave. It too has a near-vertical stem-line but is carved in a different 'hand'. Its strokes and letters are more widely spaced, less carefully parallel and of more varied length. In over-all impression, it is less neat and compact than the first. The strokes appear V-shaped in section and are slashed across the surface of the rock. Some strokes, especially to the right of the stem, seem exceptionally long, appearing to tail off into natural fissures. Until I have examined the inscription in person, the following remains extremely tentative.
From the photograph, approximately 29cm of stem-line is visible. It is unlikely to have continued any higher because of a protuberance on the rock face. Though it may have begun lower down, in the rough patch immediately before the start of the visible section, no traces survive. Reading from the bottom up there are [see fig. for key to numbering]: 1
Three strokes to the right - V. A shadow in the photograph exaggerates the length and breadth of these and the stem at this point. The strokes appear to run into natural fissures which give them the appearance of being very long indeed. The hollowness in and around the stem may be the result of water wear or chipping. The first and third strokes are perpendicular to the stem but the second appears to run obliquely. At its extreme right end, where it blends into a fissure, it is close to and parallel with the third stroke, but they diverge half-way along and the second stroke joins the stem equidistant from the two on its flanks. This raises the possibility that the group is to be read as 1 + 2 (B L).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND 2
BL ACK W ATERFOOT
98
Three cross strokes perpendicular to stem. They are of irregular length, the first being particularly long on the right-hand side, but it is difficult to distinguish stroke from shadow and fissure. Surely these are to be taken as a vowel - U (but if of the m-aicme - Gw).
3
There appears to be a short nick on the stem but this is very faint. It is marginally closer to the preceding group, so could be the remains of a fourth stroke. If it is to be taken separately it would be the vowel A, but this would imply that the two groups on either side were from the maicme, creating an unpronounceable sequence of consonants. On balance the nick is probably not a letter, the sizeable gap between letters 2 and 4 serving to clarify two successive letters of the same aicme.
4
There appear to be four cross-strokes perpendicular to the stem, each beginning at the same distance to the left of the stem but continuing greater, irregular, distances to the right. Probably a vowel - E. In the unlikely event that this is a four stroke consonant of the m-aicme, it would be one of only two examples in an epigraphic context from either Scotland or Ireland, the other being St Ninian's Isle (phonetic value uncertain, possibly /sV, see McManus 1991:38).
5
Two strokes to the left of the stem - D.
6
Two strokes to the right of the stem - L.
7
One stroke to therightof the stem, probably continuing across - A. The stem appears to continue for a very short distance beyond this stroke, but there are no further letters.
Precedent is finnly against a top-down reading of the inscription, which would produce ADLEUV. Funhermore, the stem of this ogham leans slightly to the right as if the upper part was getting beyond the reach of the carver. This can be taken in support of a right-hand stance and to imply a bottom-up reading. Thus the most likely reading of this ogham is: VUEDLA, less likely, but possible, is BLUEDLA.
INTERPRETATION OF TEXT For the transliteration of V (/w/ or IfI) see Introduction. The name Vuedla calls to mind the iddlua, 'wise
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BLACKWATERFOOT
99
woman, witch' (< *wid- 'to see, to know'), of the famous Gaulish inscriptionfromLarzac [Meid 1992:406; Lejeune 1985]. The Irish reflex (< *widlmay John Koch, pers. comm.) appears as the name of the seeress Fedelm, in Táin Bó Cuailgne, and of a number of famous early Irish women, legendary and historical. The very popular male name derived from Fedelm, Fedelmidl Fddlimid, is mentioned above (Blackwater 1) as the kind of /-stem name which would yield a genitive ending -eo. The name appears on two Irish ogham inscriptions, CIIC 206 and Pilsworth, Kilkenny [McManus 1991:73-4]. The name Fedach isfroma different root and perhaps means 'dweller in the woods' [for all three» see Ó Corráin & Maguire 1981]. Since 2, (U) indicates labial quality the most plausible explanation for this text is in Brittonic rather than Goidelic terms, though the spelling is problematic, cf. AuquhoUie for initial VU (for MS lUu). [For Uuedla cf. ?Nechtla (Formaston)].
FORM OF SCRIPT The Blackwaterfoot oghams are written in different 'hands', but are both from the same stage in the development of the script (Type Ha). Both have dráwn-in stem-lines, and vowel strokes rather than nicks, and Blackwaterfoot 1 exhibits the concomitant slope of m-aicme consonants. Thus they represent examples of simple Type II inscriptions and, since they lack all 'late' features, such asforfeda, bind-strokes, and word division, typologically, they should be placed near the beginning of the Type II series. At best this would offer a relative date, but since simple Type II ogham continued in use alongside more ambitious, formal varieties of the script, inscriptions using this form cannot be closely dated. They could be as early as the Pool ogham (sixth century) or later than Dunadd (late eighth century?).
The Pool inscription provides the closest parallel in Scotland, to the script of Blackwater 1 and 2, which may lend support to an earlier rather than a later dating. Pool is another informal ogham, which, like Blackwaterfoot 2, has long thin strokes and irregularly-spaced letters. Three other Scottish oghams are similar except for having doubled letters (a feature which may have significance for dating); Brandsbutt,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BLACKWATERFOOT
100
Scoonie, and St. Ninian's Isle* The first two are formal, public monuments, exhibiting high quality carving in addition to the lettering, and even the undecorated slab from St. Ninian's Isle is carved in a formal style.
DISCUSSION The two inscriptions from Blackwaterfoot provide valuable evidence of informal, non-monumental use of ogham, a category otherwise poorly attested. Though Blackwaterfoot 1, in particular, is carved carefully, neatly, and with an eye for balance, the two are clearly casual graffiti. Outside Scotland there are so few extant examples of ogham in contexts other than on die classic memorial pillar, that there is very little indication of the kind of text that might be committed to ogham in other circumstances. The recent discovery of ogham graffitti at Knowth, Co. Meath [Prof. George Eogan UCD, pers. comm.], bolsters the view that the script was more widely known and used than the pillars suggest. Almost all the extant ogham inscriptions, in all contexts, record personal names, and while there are a number of interesting exceptions, when tackling new texts it is reasonable to start with the expectation that they too are most likely to consist of personal names. It would be particularly surprising if graffiti like the Blackwaterfoot oghams were anything else. It might be expected that names in such a context would appear in the nominative case, but the standard ogham formula may have exerted such influence that ogham names were habitually written in the genitive, even if that case were not strictly appropriate.
BIBLIOGRAPHY
OS Record card - NR 83 SE 10 NMR
MS 1968 R. W. B. Morris
Inscription 1:
K. H. Jackson 1973a ill.
Inscription 2:
Unpublished
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
Site only:
BLACKWATERFOOT
101
MacKenzie 1914:31,113; Balfour 1910:213-18 ill.; Martin 1703:219; Pennant 1774:1812; Stuart 1867:xciii pl.36; Thomas 1961:23.
PREVIOUS READINGS
Inscription 1:
Jackson
Inscription 2:
None
EOMEDTE or EOMEQE (EVLEMOE or ENEMOE).
BLACKWATERFOOT KING'S CAVE
Exterior of King's Cave [Balfour 1910 pl.XXXIII.1]
i ,% J.U j.^j^á^L, $ |« I*
Fit:, f*—riati of King's Case. {The stuufol portion lejircsciiti* llic sections dug.)
Plan of interior showing area of Balfour's excavation [Balfour 1910.214]
1
i
"J?mfíi
Interior of cave showing position of ogham inscriptions [®RCAHMS Neg. B69005]
-B=
N \ * \*
Schematic representation of ogham inscription (key to numbering)
BLACKWATERFOOT 1 - Ogham inscription [Jackson 1971 pl.K]
sí-^tS/'isL-
*£&S *
v k
i«*ri
•ÍSuSCí £tf£Í2 :
i'l^?i-5-
BLACKWATERFOOT 2 - Ogham inscription [©RCAHMS Neg. B69011 ]
7 6 5
Schematic representation (key to numbering)
BLACKWATERFOOT 2 - Ogham inscription
102
BRANDSBUTT
DISCOVERY The Brandsbutt Stone, a scheduled monument, was marked as a 'sculptured stone' on Sheet 54 of the First Edition Ordnance Survey Map but it is unclear exactly when the carving was first recognized (it is not included in Stuart's Sculptured Stones of Scotland). According to Davidson, Aberdeenshire was surveyed in the period 1865-71, and a search through the relevant Ordnance notebooks in the Scottish Record Office might yield a more precise date [J. L. Davidson 1986:15]. Passing reference was made to the sculptured stone in 1878 [J. Davidson 1878:5], but it appears that F. R. Coles, Joseph Anderson's assistant at the Antiquities Museum in Edinburgh, was the first to identify the ogham inscription as such. Coles describes how a visit to the site in September 1901 to check the position of the stone met with success when Frank Dey, his driver on that occasion, 'detected marks upon the stone half buried'. On cleaning, these were seen to be part of a symbol pair and ogham inscription [Coles 1901:230]. The megalith had been broken up» doubtless by the traditional method of lighting a fire against it and shattering it with a dowsing of cold water, and built into a dyke near the farm of Brandsbutt about 1 mile (1.61km) north-west of Inverurie [NGR NJ 7599 2240]. Following Coles's discovery the dyke was dismantled and searched, and funher pieces of the stone were recovered. The photograph published in ECMS shows four fragments of the sculptured face, these are reproduced in Macalister's drawing [1940:196 fig.5]. At some date subsequent to Macalister's visit a fifth fragment was recovered, and incorporated into the restored structure, which is how it stands today. Given the extent of subsequent digging activity in the area it seems a vain hope that the remaining missing pieces will yet come to light.
SITE When Padel visited the stone in the early 1970s it was still in a small Ministry of Works enclosure just to the west of Brandsbutt farm. In the meantime a housing development has grown up around it to form a suburb of Inverurie [see fig.]. In 1983, prior to the commencement of the building work, Ian Shepherd,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
103
Grampian Region Archaeologist, conducted an exploratory exercise to confirm F. R. Coles's identification of a large stone circle at the site [Shepherd 1983]. Shepherd was able to identify twelve or thirteen stoneholes comprising a circle 25m in diameter, centred on NGR NJ 7601 2239. Two of the stones remain, one relocated in a dyke. The larger of these, known locally as the 'Douping Stone' was used as recently as October 1907 in the ceremony of the riding of the Burgh Marches and the creation of a burgess [Coles 1901:230; OS Record Card]. As far as can be ascertained, the ogham-inscribed stone is the third surviving constituent member of this Bronze Age monument. There is no shortage of precedents for this early Medieval re-use of prehistoric monuments. One of the stones of the remarkable complex at Broomend of Crichie, just south of Inverurie, was graced with a pair of symbols [Coles 1901:219-25; ECMS 160] and only slightly further afield are the ogham-inscribed standing stones at Logie Elphinstone and Newton [see entry on Logie Elphinstone for further discussion of this phenomenon]. As John Davidson points out, Brandsbutt is situated by the old road north from Aberdeenshire into Moray (the modern A96), specifically on the route to the hill-fort of Dunnideer (as is the Class II cross-slab, the 'Maiden Stone') [1878:5]. [See entry on Logie Elphinstone for further discussion of the archaeological and historical associations of the Lower Garioch area]. The name of the farm, pronounced /branz'but/, is from the surname Brand and the Old Scots butt meaning 'a piece of ground less than a complete rig' [Alexander 1952:187].
DESCRIPTION OF OBJECT The stone is a huge mass, of roughly triangular section, broken into several pieces. It is incised on its smoothest face with two Picdsh symbols (crescent-and-V-rod, snake-and-Z-rod) and, parallel with the left edge, an ogham inscription [see fig.]. Stone:
Dark whinstone [ECMS]
Dimensions:
In lower part: c. 1.07m x 1.27m x 0.91m (3'6ff x 4'2" x 3')
Condition:
Stone fragmentary, but extant carving very well preserved.
The dimensions given above are those listed by Romilly Allen in ECMS, with approximate metric equivalents. I measured the maximum width between the parallel edges of the sides to be 1.40m, but the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND difference is not significant.
BRANDSBUTT
104
Since the stone is of irregular shape it is difficult to attain precise
measurements. In its current position the stone lies about 10° from vertical and its highest point (almost the top right corner) is 1.42m above ground. The bottom right corner is missing, but other than that the side edges are intact. Portions have been lost from both top and bottom as indicated by the fractures of the upper surface and the incompleteness of the lower terminal of the Z-rod. Our only guide to the possible original dimensions of the stone, is the observation that symbols tend to be medially placed on large standing stones (e.g. Newton, Edderton) when these are re-used as Qass I monuments. If this is applicable to Brandsbutt then it is likely that only minor amounts have been lost from the upper and lower extremities. A fairly modest triangular wedge at the base is all that is required to bring the stone back to the vertical. Though the ogham traverses three fragments these fit neatly together and none of the text is missing before the end. The stem continues to the break, see below for discussion of possible continuation. The stone is very hard and has suffered virtually no ill-effects from weathering. The inscription is, as Diack said, 'as legible as the day it was cut' [1944].
DESCRIPTION OF CARVING Though currently on a diagonal axis, the carving is in tact arranged vertically relative to the intact sides of the stone. In the following discussion it will be assumed that the original orientation of the stone was with side edges vertical. The lower extremity of both ogham and symbols is roughly level (on this new axis) and the symbols are well over to the right to allow for the ogham, with the tip of the right-hand end of the V-rod curving round the edge of the stone slightly. This unity of design, and the identity of the carving technique used for both, indicates that the two are contemporary. Henderson cites Brandsbutt as a 'particularly impressive' example of 'the Pictish artist's capacity for composition' [1972a: 171].
The
strength of the over-all design, the quality of carving, especially of the serpent, and the general consistency, thickness, and smoothness of line, combine to make this an expert piece of work. Padel thought these features gave an impression of lateness, but any such assessment is necessarily subjective.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
105
In the case of Brandsbutt we are fominate in being able to benefit from C. A. Gordon's detailed analysis of the carving technique of certain of the North-East symbol stones [Gordon 1956]. Gordon demonstrated that all the Class I stones which he examined in the North-East were carved using the pock-and-smooth technique, though they differed widely in the amount of laborious smoothing which had been done. Of die stones in his sample, he mentioned Brandsbutt as being the most 'expertly smoothed' though he could still discern 'faint traces of the original pocking' [1956:44]. Though Brandsbutt is an exceptionally fine example of the technique it is very much part of the local tradition of incised sculpture. Gordon noted that the same technique was used on several Class I stones in the vicinity, including, most interesting for our purposes, Logie Elphinstone.
The scores are broad and shallow in section. The stem is 30mm across at die surface, but only 10mm at the bottom of the channel. The letter scores are slighter, an average 25mm at the surface, 5mm at die base. The comfortably finger-sized grooves are 'smooth and regular, contrasting clearly with the flat but roughish surface of the boulder' [Gordon]. The length of the strokes varies, the tendency being towards diminution, i.e. starting large and trailing off. The three As are respectively 99, 95, and 83mm long; the two Rs are similar, but die I and the E are 126x129mm and 103x116mm. Whether or not this is deliberate, and if so whether for aesthetic or practical reasons is not clear. It may be die result of trying to squeeze the required text into a finite space without detailed planning. This diminution however is revealed only by close analysis and is not immediately obvious to the naked eye. It does not in any way mar die over-all impression, and one cannot help but agree with Macalister's judgement that 'every score is perfect' [1940:196]. Romilly Allen highlights the lack of rigid symmetrically in Pictish sculpture and contrasts it with 'the monotony of die absolutely "cast-iron" repetition' of modern work [ECMS 47 n.2]. Certainly when it came to lettering, the aesthetic was not one of strict regularity.
DESCRIPTION OF CARVING - DECORATION Leaving the ogham aside for a moment, Brandsbutt is typical of Class I Pictish Symbol stones in being
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
106
incised with a single 'statement' consisting of a pair of symbols one above the other, placed close together almost to the point of touching [Thomas 1963; Forsyth 1996b]. The upper symbol, a 'crescent-and V-rod', i s b y far the most common o n Class I monuments. Elizabeth Alcock lists 52 occurrences on Gass I stones, roughly twice that of the next three most frequent symbols [1989:10-11]. It is widely distributed, being found in all of Alcock's zones, though with a clear nonhem weighting (north of Mounth), and particular concentrations in her zones B (Caithness and Sutherland), D and E (roughly modern Grampian region). The crescent-and-V-rod is twice as likely to appear in upper as in lower position [ibid 15].
The lower of the two Brandsbutt symbols is the much less common 'serpent-and-Z-rod' which occurs on only five other Class I stones (Dnimbuie 1INV, Insch ABD, Inverurie 1 ABD, Newton ABD, and Dairy Park, Dunrobin SUT]. This particular pairing is unique to Brandsbutt, though both symbols appear on nearby Inverurie 1 separated by a 'circular-disc-and-rectangle' and with the addition of a 'double-disc-andZ-rod'. If this quartet is to be read as two 'statements' on one stone, then the crescent and die serpent are in different 'statements'. Serpents-and-Z-rod appear on six Class II slabs (Tarbat 1 ROS, Balluderon FOR, St. Vigeans 2 FOR, Gask PER, Logierait PER, Meigle 1 PER), but only once with a crescent-and-V-rod (Tarbet 1, with the insertion of a 'tuning-fork' to form a trio). This relative 'overabundance' in Class II led Gordon Murray to conclude that the snake symbol appeared comparatively late in the Class I sequence [1986:251]. Furthermore, it is interesting to note that, as reponed in ECMS, the Class I and the Class II distributions of this symbol (ABD/INV and ROS/FOR/PER respectively) are mutually exclusive.
Looking more closely at the form of the Brandsbutt symbols, the decoration on the crescent-and-V-rod is an example of Stevenson's 'Dome and Wing' design, a type restricted almost exclusively to Class I, and particularly well represented in the North-East [Stevenson 1955:102-6, fig. 15, map 4]. Stevenson thought the dome-and-wing type crescent were later than the pelta-type, Murray came to slightly different conclusions [see 1986:226-35 where the form and distribution of the crescent-and-V-rod is discussed at length]. As Murray noted, the 'classical' form of the V-rod appears to have been a fish-tail terminal to
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
107
the left and a lentoid terminal to the right, each with a pair of simple, inward-curving curls at its base, the whole being reminiscent of the head and flight of an arrow or spear. The more devolved forms fail to maintain the differentiation of the terminals. The Brandsbutt crescent departs from the prototype in two regards. Firstly, in place of the fish-tail, it has a closed 'onion-dome1 terminal, a form which appears to be unique. Secondly, in place of the simple curls, Brandsbutt has more elaborate S-shaped scrolls. On the left terminal these join at the top and flow down in a heart shape, while on the right they are attached at their base and curl up like a lyre. This formation is more like the typical Z-rod termination and it seems probable that there has been some confusion or conflation between the two conventions [Murray 1986:235]. Rhynie 6 has a right hand terminal almost identical to Brandsbutt's, though the left one is of more typical fish-tail-and-curl form.
The Brandsbutt serpent is missing its third quarter but in what remains, resembles all but one of its Class I fellows in having its head to the left and its body crossing over, under, over the middle section of the rod. The exception is Newton (ABD) which points to the right and goes under, over, over. One feature which Brandsbutt has in common with Newton, however, is that both have schematic representations of the scales on the animal's body. Newton's is the more complex design, incorporating a circle-and-dot motif and occasional sub-division of scales, but it could be argued that the simple scalloping on the Brandsbutt creature suggests volume more effectively.
Both the terminals of the Brandsbutt Z-rod are simple peltas, at Newton and Drumbuie INV the terminals are elongated peltas with the terminal nearer the head significantly longer than the other. The connection between the Newton and Drumbuie symbols is further strengthened by the fact that both are paired with a rodless double-disc (though in reverse order). The serpent geographically closest to Brandsbutt, that on Inverurie 1, is exceptional among Class I serpents in lying under a straight rod. Of the Class II examples, at Meigle, Balluderon, and Gask the snake lies over the Z-rod, at Logierait it has a grotesque head and is wrapped four times round a straight, horizontal rod. At St. Vigeans the symbol is turned 90°, as it is at
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
108
Tarbat where, furthermore, there appears to be confusion over the terminals. These Class II serpents-andZ-rod seem much devolved, the relative scale of the two components is different and the balance between them is lost. They appear to be a separate, probably later, development, and are thus not directly relevant to a discussion of the Class I examples.
The symbol stones are so poorly understood that we are a still a very long way from understanding the meaning and purpose of monuments like Brandsbutt. What we can say, on the basis of the above observations, is that it is fully typical of its class. In terms of both design and carving technique it is closely linked with its North-East neighbours, particularly Inverurie, Logie Elphinstone, and Newton, but also with stones further afield. Like the Newton Stone, Brandsbutt is a very accomplished piece of sculpture, the product of a skilled artist-craftsman.
Any more specific comments depend on an adherence to the so-called declining symbol theory. According to this hypothesis, which was first set out with reference to the crescent-and-V-rod by Stevenson [1955], one can observe a process of simplification and decline in the quality of symbol decoration from a classical or prototypical form. On the basis of this, one can form a relative chronology, which inter alia confirms the general lateness of Class II symbols. Isabel Henderson developed this idea, applying it to V- and Zrods, notched rectangle, and Pictish beast [1958]. Her argument was that the geographical concentration of the best examples of a particular symbol indicate its origin centre. This analysis led her to conclude a nonhem origin for the whole system, centring on Easter Ross. Gordon Murray pursued this line of argument and, on the basis of a detailed statistical analysis of the most commonly occurring symbols, concluded that, while the system as a whole is northern, individual symbols originated in different areas [1986].
In theory, this approach is of great importance for understanding, and especially dating, the Pictish symbol stones. In practice, however, there are a number of methodological problems [Murray 1986]. The main
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
109
stumbling block is the subjectivity of the art historical analysis. The divergence of opinion between Stevenson and Murray over the decoration of the crescent is a case in point. Another problem is that decline need not be uni-directional, that is to say, one could invert certain typologies to exemplify rationalization rather than progressive confusion. For instance, it is widely held that the V- and Z-rods were originally arrows and/or spears, but as this meaning was forgotten their terminals ceased to be differentiated. From an alternative perspective, however, the arrow and spear interpretation could have been imposed at a later point within the Class I sequence in order to 'make sense' of an abstract design. The undifferentiated Class II terminals could thus be, either a continuation of the original design, or declined versions of the arrow/spear forms.
The third problem is our ignorance of the rate of decline. We need not assume that it was constant, either within a local area, or between regions. So, even if relative sequences of decline can be agreed on, there is, as Murray put it, always the 'possibility of conservatism in a centre of excellence' [1986:224]. He goes on to explain, '[a] design in northern Scotland close to the supposed classical form could be contemporary with a devolved design in another area, for instance Aberdeen' [ibid.]. We have no way of assessing the depth of time involved in this process, but furthermore, since the typology is not firmly anchored at any point we have no idea of absolute chronology. So, even though Brandsbutt exhibits a number of stylistic features which might be termed 'late' - the use of the serpent symbol, the dome-and-wing design on the crescent (but note Murray's contrary opinion), the anomalous terminals on both symbols - all that can really be said is that is 'later', but how much, and than when, we cannot tell.
DESCRIPTION OF CARVING - OGHAM INSCRIPTION The stem-line is arranged vertically on the face, roughly parallel with the left-hand edge. It is more or less straight, though wavers slightly, especially in the second half of its length. The stem starts cleanly at the first stroke, several centimetres above the current base of the stone. It rises for roughly 1.18m (3'IOW) until it meets the fractured edge just beyond the last visible stroke. As Macalister pointed out, any
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
110
subsequent strokes must have belonged to the h-aicme since the b-surface survives almost intact for several centimetres more. The individual letter-strokes are very evenly and closely spaced, each equidistant from its neighbours, the only exceptions being two instances where the spacing serves to clarify successive letters of the same aicme, i.e. between 6 and 7, and between 12 and 13. The gap between the final stroke and the break is of approximately the same size as that between the previous two letters, the carefully differentiated 12 and 13. On the basis of this, one might have expected a further letter of the b-aicme, but this cannot have been the case for the reason given above. Such spacing is slightly puzzling. If the arrangement of the rest of the letters is a guide, any H-stroke would be expected to follow hard on the end of the final stroke of the S. Since it obviously does not, the possibility opens up that the S was in fact the final letter and that the stem continued for only the shonest distance beyond the break. While it seems to have been the convention to place the first stroke of a formal inscription at the very beginning of the stem (e.g. Ackergill), since the end of the text is so often lost or damaged, it is unclear if texts always finished as neatly. The stem of the Scoonie ogham projects very slightly beyond the final stroke, though not as far as here at Brandsbutt. The total length of the Brandsbutt ogham, relative to the size of the symbols, may be a further indication that little, if anything, has been lost. Having begun parallel with the lowest point of the symbol pair, the ogham has, at the break, already over-shot the upper end of the V-rod by several centimetres.
To have continued much further would, to our eyes, have upset the balance of the
composition, but whether the Pictish anist would have been similarly disturbed we cannot know.
One would expect, of course, that the text read upwards, though the intractability of the resulting reading might tempt one to read down, citing the precedent of nearby Newton. There is however a slight slope on the letters of the h-aicme, which can most easily be discerned on comparing them with the neighbouring vowels, e.g. 314 or 516. This slope confirms the direction of reading as up. A quarter of the way up the largest fragment a very firm beginning is made with [see fig. for key to numbering]: 1
Five long strokes perpendicularly across the stem - 1 . These strokes are the longest of all the vowel-strokes (125-129mm).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND 2
BRANDSBUTT
111
Five long strokes obliquely across the stem - R. The very definite forward slope of this group clearly differentiates it from the preceding vowel. The strokes are not balanced on the stem but are slightly shorter to the left. The first stroke stops a little short of the others to the right of the stem, thus avoiding intersection with the last stroke of the I. It is a mere 150mm in length, the others range 159x175mm.
3
A single long stroke perpendicularly across the stem - A. It is wedged quite tightly between the opposing slopes of the flanking strokes.
4
Three strokes to the left of, and perpendicular to, the stem - T. There is a very slight slope on this letter, its proximal end forward, relative to the direction of reading.
5
A single long stroke perpendicularly across the stem - A. Slightly shorter than the preceding A.
6
Two strokes to the left of, and perpendicular to, the stem - D. There is a slope on this letter similar to, though slightly more pronounced than, that of 4.
7
As above - D. These two pairs are very carefully spaced to clarify that they are to be read separately. A gap sufficient to take a stroke has been left between them.
8
Two long strokes perpendicularly across the stem - O. A fracture runs across this pair diagonally from left to right downwards, but nothing is lost and they are quite clear. The text continues on the second fragment.
9
A single long stroke perpendicularly across the stem - A. The shortest of the three As.
10
Five long strokes obliquely across the stem, sloping forward - R.
11
Four long strokes perpendicularly across the stem - E. This group is disturbed by a fracture running diagonally from left to right up, but nothing is lost and they are quite clear. The text continues on the third fragment.
12
Five strokes to the right of, and perpendicular to, the stem - N. The stem wobbles slightly at this point.
13
Four strokes to the right of, and perpendicular to, the stem - S. Again two successive letters of the same aicme have been carefully spaced to avoid conflation, with a gap sufficient for an
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
112
additional stroke left between them. The stem continues the short distance to the final fracture. There is no trace of any further strokes. This gives a reading of: IRATADDOARENS-
FORM OF SCRIPT The form of the script is a very simple one. Features to note are: the lack of any of the supplementary letters, the slope of the letters (m-aicme well sloped, but only the very slightest of slope on other consonants); the relative length of strokes in the various aicmi (vowel strokes the same length as strokes of the b- and h-aicmi, therefore occupying only the middle two quarters of the ogham band, as do the longer diagonal m-aicme consonants); the tight spacing of strokes with nothing extra between letter groups. All of these features point to a place early in the Type II typological sequence, though this is not necessarily indicative of an early absolute date. The design aspect of the Brandsbutt ogham is the strongest of all the Scottish oghams. It is the one farthest away from the casual, tachygraphic inscriptions of Gurness or Burrian. But the artist has retained the concern for legibility of real writing. This may be a display script, but it could still be read easily.
INTERPRETATION OF TEXT There seems to be a law of Scottish ogham that the more legible the inscription, the more inscrutable the text. If so, then Brandsbutt is the limit case. The text is utterly clear and unambiguous. There is no disagreement over the transliteration required, but most authorities have thrown their hands up in despair at the text. Only Macalister made a determined effort to interpret it, and the result is not at all convincing. If, on the analogy of Newton, one were to read the text from top down, the reading would be CQERAOLLAVARI. The final I looks like an Irish genitive singular ending, but the rest is no better than the other reading, and in any case, the slope of 7 is against it. IRATADDOARENS-.
So we shall have to stick with
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
113
The thirteen letters of the text may be complete as they stand, or we may be missing one or two more from die end. If die text continued it is unlikely to have done so for very long. The most likely interpretation of a text of this length is a single personal name. Brandsbutt is a little longer than Ackergiirs eight letters and Scoonie's ten, but thirteen letters is not inconceivably long for a single word. Six syllables does, however, seem rather too many for that, and sub-division into two or more words is more likely. Where such a division or divisions might come we can only guess. No indication of word division would be expected, and the exaggerated spacing between 6 and 7, and between 12 and 13 serves a different purpose. One piece of guidance comes from die tendency of Scottish ogham inscriptions to avoid double consonants in word-initial position. We can therefore take the division IRATA DDOARENS- as unlikely. The juxtaposition of OA is a difficulty if the text is to be interpreted as Celtic, perhaps we might, therefore, segment, IRATADDO ARENS-. Unless, that is, that OA is a form of the verb 'to be' in Pictish, cognate with Breton oa 'was* [Jackson 1953:350].
If the text is not a single name, it may consist of name plus patronymic, or name plus epithet or title. None of the Irish formula words, MAQI, AVI, NETTA, are here. If the word is not a name then we have no precedents to guide us. We can pick out a couple of short segments vaguely reminiscent of Celtic words or elements, but unless the whole text can be plausibly accounted for, diese partial interpretations are worthless. The -REN- recalls the VORRENN of Newton, but how, then, to account for the final -S (though cf. final -S in Cunningsburgh 3 -CONMORS) ? Alternatively, might we have some form of Welsh rhén 'lord* [J.T. Koch, pers.com.]. Macalister compared the IR- to the i/r- prefix of the early stretches of the King-List [1940:197]. Alternatively, we might have the Brittonic article ir but this seems unlikely for an inscription.
Thomas Clancy has suggested to me one way of cutting through the Gordian knot. He proposes to segment the Brandsbutt text I RAT ADDOAREN S-, and interpret it as Old Irish / rath Aduarén s~. According to this interpretation, / is the nasalizing preposition 'in' (followed by dative or accusative case). Since vowel
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
114
length is not indictated in ogham inscriptions, RAT could be either rath or rath, both of which have several meanings. DH quotes Thurneysen as identifying rath as the verbal noun of ernaid 'bestows, grants'. It is used to mean a grace (especially in a religious sense), a virtue or gift (e.g. 'gift of prophecy*). It can mean 'granting of a favour' or 'mark of a favour or boon*, 'fortune*, 'prosperity'. Related to this is rath meaning 'goods, chattels, property', 'especially as bestowed in bounty or given as a due or stipend*. In a legal context rath means 'fief, dolr-rath and soir-rath referring to stock given by patron to client. There are two separate words rath, one meaning 'surety, guarantor', the other, frequent in place-names, means 'earthen rampart', and by extension a chiefs residence, fort, or burial ground surrounded by such a rampart [DE]. From limited evidence, for instance Buckquoy's -DDACT for -dacht, it appears that in Scottish ogham inscriptions certain lenited consonants are shown simplex, and unlenited doubled. Thus the single T of Brandsbutt may be taken to represent -th. As Dr. Clancy points out to me, Aduar is an attested male personal name [Annals of Innisfallen 810, CGH], of which Aduarán or Aduarén would be 'plausible diminutives' [pers. comm.] The diminutive -én [Thurneysen 1946:174 §272] does not appear to decline in names. Uhlich interprets Adúar as a compound of the intensifying prefix ad, and iiar 'cold', meaning 'very cold'. He compares it with MW. adoer 'sad, without feeling, cruel, chilling' (GPC 33) [1993:144-5]. The final S is a little problematic. The form of the extant carving shows that it cannot be followed by a vowel, thus the very plausible interpretations^*?, 'this', and sund, 'here', are ruled out (there would almost certainly be insufficient space for the latter in any case). Comparison with the final S of Cunningsburgh 1 's CONMORS opens the possibility that in both cases we may have some reduced form of so. The whole text could thus be interpreted as 'in rath/ráth of Adúarén (here)', i.e. 'this is the rath/ráth of Adúarén'. Given the archaeological context of the stone, the meaning 'earthen rampart', 'fort', is unlikely. While rath usually refers to moveable wealth, 'property', 'fief, may be the meaning here, or perhaps rath 'gift', 'something bestowed*. If the stone had been an explicitly Christian one it could have been compared with other stones marking the donation of land to the church (CIIC 946 Kilnasaggart, Armagh, 993 Llanfihangel Ystrad, Carmarthen). Thus the text would mean 'this is the gift
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
115
of Adúarén', either given by the magnate Adúarén, or, since Adúar is the name of a saint [Ó Riáin 1985 §§186, 277], given to a church in honour of its patron. Since the monument is, however, a symbol stone, a secular meaning is perhaps more likely, 'this is the property of Adúarén'. Dr. Clancy reminds me of the provision made in Irish law for exactly this kind of marking of land ownership, int ogam isin gallon ... gebid greim tuinide, 'the ogham on the pillar stone has the force of ownership' [Kelly 1988:204]. It might be wondered why an Irishman was in a position to give or receive land in the Garioch, but see entry on Logie Elphinstone for discussion of Irish connections with Aberdeenshire.
DISCUSSION When trying to make sense of the monument as a whole, it should be noted that the ogham is on the same scale as the symbols and is given equal prominence. The two have been demonstrated to be contemporary, but whether their messages are separate, equivalent or complimentary remains to be established.
On
analogy with the bi-lingual oghams of Wales and Dumnonia, it is possible that the two components duplicate the message, with the ogham intended to give access to those unskilled in reading the symbols. Alternatively the ogham might communicate something which could not be expressed within the native symbol system, a foreign personal name, for instance. Faced with the possibility that Brandsbutt and Ackergill might be Pictish Rosetta stones, it is all the more frustrating that so little can be made of their texts. The juxtaposition of 'Irish' ogham and 'Pictish' symbols is interesting, one wonders if the symbols did have the overwhelming ethnic connotation which modem commentators and enthusiasts appear to assume.
For the re-use of Bronze Age monuments in the Pictish period, see discussion of Logie
Elphinstone ogham pillar.
BIBLIOGRAPHY
OS Record card - N J 7 2 S E 2 3
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRANDSBUTT
116
Coles 1901:230; ECMS 506-7, fig.551; Browne 1921 pl.xli.2; Macalister 1940:196-8 fig.5; Diack 1944:61-2; K. Jackson 1955:141, pl.12; Gordon 1956:43-4, pl.VIII.2; Padel 1972:61-2; Shepherd & Ralston 1979:30.
Site only:
Shepherd 1983.
PREVIOUS READINGS
ECMS & Padel
IRATADDOARENS-
Macalister
IRATADDOAR ENS-
1*5
BRANDSBUTT - Ogham-inscribed Class I Pictish Symbol Stone [^University of Aberdeen, Anthropologcial Museum]
\
I
\
Cw' >
V
'
\
BRANDSBUTT - Ogham-inscribed Class I Pictish Symbol Stone
13 12
11 10
BRANDSBUTT - Ogham inscription (key to numbering)
117
BRESSAY
DISCOVERY
This cross-slab was first brought to public attention by Edward Charlton, the Newcastle-on-Tyne Antiquary, who, in 1852, had been informed by W. H. Fothenngham of Kirkwall of the existence of a 'runic' inscription in the minister's garden at the manse on the Island of Bressay, Shetland. The slab had been found by a labourer digging a piece of waste land near the old churchyard of Cullingsburgh (formerly known as Culbinsgarth) on the east side of the island [NGR area centred HU 521 423] sometime before [Charlton 1855b]. It was taken first to Gourdie House the home of Captain Cameron Mowat of Garth, and was later brought to the churchyard of Bressay by the minister Rev. Dr Zachary Macaulay Hamilton. Dr. Charlton visited Shetland that summer and, seeing the stone, recognized the inscription as ogham. Dr Hamiton lent him the stone which he took back with him to Newcastle. A gutta percha cast of the oghams and the stone sent to Dublin for the inspection of Charles Graves who was then writing his work on Irish ogham. These casts were presented to the Royal Irish Academy by Albert Way [Charlton 1855b]. Having waited six months without a reply from Graves, Charlton went ahead and exhibited the slab at a meeting of the Archaeological Institute of Great Britain in Newcastle-upon-Tyne [Charlton 1855a]. A cast was deposited in the Institute's Museum in Newcastle [note by Atkinson in Brash 1879:358]. The stone was sent back to Shetland where it remained until February 1864 when Dr Hamilton presented it to the National Antiquities Museum where it is currently on display (NMS Cat. no. IB 109) [n.a. 1856:240]. In the meantime, Graves had presented his interpretation of the text to the Archaeological Institute [n.a. 1855, 1856] and to the Royal Irish Academy [1858]. At a later date, Stuart arranged for another cast to be sent to Ireland for the analysis of Brash [Brash 1879:355]. The stone was first illustrated by Charlton [reproduced n.a. 1856; see fig.], subsequently by Stuart [1856], Wilson [1863:240], and by Romilly Allen [ECMS].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESS A Y
118
SITE The ruined church of St. Mary's at Cullingsburgh is surrounded by a very old burying-ground [RCAHMS 1946:1-2 No. 1083; HU 53 SE 5, HU 54 SW 5]. MacDonald & Laing did not list it in their survey of early ecclesiastical sites [1968], but the discovery of the cross-slab surely suggests some kind of Christian foundation there in the early medieval period. Charlton notes that the building is remarkable among the 'humble religious edifices of Shetland1 in being a cross-kirk with transepts [1855:156], which indicates the importance of the site at a later period. Very little is known of pre-Norse Bressay. The old churchyard is immediately adjacent to a very ruinous broch which has not been excavated [HU 522 423; RCAHMS 1946:4 No. 1086]. At Gungstie on the small neighbouring island of Noss is the site of an early chapel and burial ground, the probable source of an early cross-incised slab found locally. Within a few hundred yards is Papil Geo, (<Papabyl 'priests dwelling') a place-name which incorporates the Norse word for Celtic ecclesiastics, papar [MacDonald & Laing 1968:129; MacDonald 1977].
St. Mary's chapel lies near the head of the large sheltered bay, Cullingsburgh Voe, on the north-east coast of Bressay. The west of the islandfrontsBressay Sound, a kilometre-wide stretch of sea between Bressay and a peninsula of mainland Shetland, which provides ideal shelter for a large number of ships. The curve of the peninsula is now occupied by Lerwick, the Shetland capital, and in recent centuries the area has been of great importance. As Anna Ritchie points out, however, this prominence is comparatively modern. The low fertility of the land in this area made it unattractive for early settlement, Lerwick is from Norse leir vik 'mud bay'. The first capital of Shetland was Scalloway, on the west coast, opposite Lerwick [Ritchie 1985:40]. That the early ecclesiastical focus was also on the west coast is reflected in the sculptural remains from Papil (Burra) and St. Ninian's Isle.
DESCRIPTION OF OBJECT A tall, very thin cross-slab, tapering to the bottom. Sculpted on its broad faces with low relief decoration and incised up the length of both flat narrow feces with ogham letters.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
Stone:
BRESSAY
119
Chlorite schist [RCAHMS 1946:2]. A slab of the local stone, soft, with a rough and rather wood-like texture.
Dimensions:
c. 1.15m tall x c.0.4 - c.0.3 m wide x c.0.05m thick
Condition:
Good, complete. Inscription intact, clear, and well preserved.
At a little underfourfeettall, this is a rather diminutive monument (the related Papil slab is about a metre taller), though almost exactly the same size as the Lunnasting slab. Whereas the later is carved only with a line of ogham, and a single linear cross, the Bressay slab is highly decorated on both sides. It is conventional to label the cross face of a Pictish slab the 'front'. At Bressay both sides have crosses but the more developed side, the one with the larger cross, the single lion, and the single horseman, is clearly die 'front'. The carving continues lower down the front of the slab than the back. The RCAHMS state that it is possible that the oghams might have been cut 'subsequent to the first erection' of monument [1946:3]. This need be no more meaningful than that a house's roof was built subsequent to the erection of its walls. It seems to me that there is no way of telling whether or not the inscriptions were pan of the original monument. What is clear is that the lettering is a well-carved monumental text, not by any means graffiti. Since the ogham does not impinge on the rest of the design their contemporaneity can neither be proved nor disproved but, since the inscription is a formal one, aesthetically well-integrated, it seems preferable, in the absence of positive evidence of re-use, to accept it as pan of the original schema.
DESCRIPTION OF CARVING - DECORATION Both faces of this rather wedge-shaped slab are completely covered with carving in low and false relief. The top of the slab is slightly shaped to follow the profile of the two opposed beast's heads which frame the top of the front of the slab.
'Front' Two beasts with massive heads and tiny bodies face each other across the top of the slab. They hold between them in their jaws a small human figure, who may be intended to represent Jonah. They are of
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESS A Y
120
a type seen elsewhere in Pictland with a fore-limb and a curling tail (see Brodie for fiinher discussion of such creatures). They sit on top of a disc containing a large ring-and-knot interlaced cross-of-arcs. The petal-shaped interstices between the expanded arms are filled with fiinher interlace and there is a circle of interlace at the centre of the cross head. The cross sits on a plain shaft containing a small profile figure on horseback above a band of simple interlace. Either side of the cross-shaft is a large profile figure wearing a long hooded garment, carrying a crozier and satchel over the shoulder - the conventional attributes of an ecclesiastic. In front of the right-hand figure is a figure-of-eight twist of interlace which overlays pan of the shaft, above his head is a ring attached to a loop in the margin. In front of the lefthand figure is a simple incised cross. To the left, die cross reaches straight to the edge of the slab, on the right there is a small gap, which is filled with a band of vertical fret-pattern. The above carving occupies the upper two-thirds of the slab. The lower third is almost completely occupied by a lion in profile facing left. This creature has been compared to manuscript versions of the lion symbol of Mark the Evangelist. In the gap between his tail and the feet of the right-hand ecclesiastic there is a horizontal band of twostrand interlace. At the very bottom, in shallower relief, is a smaller creature like a bear, a boar, or a 'fat pig' [ECMS] in profile facing right.
'Back' This fiace is divided into three unequal panels framed by plain narrow bands. The top panel occupies half the face and contains a cross within a ring all created out of a single band of interlace. The narrow middle panel, which takes up about a sixth of the fiace, consists of a pair of opposing lions facing each other with open jaws. The bottom panel contains two profile ecclesiastics like the ones on the front. Around the cross is an interlace of two-strand ring-twist of a distinctive pattern 'typical of Norse-influenced sculpture of Clydeside, south-west Scotland, Wales and elsewhere' [Stevenson 1955:28]. The only Pictish example of ring-twist occurs on the late cross-slab St. Vigeans No. 10 [op. cit. (1980 reprint footnote *)]. Stevenson identified the ring-twist as having 'split-bands' [1955:28], a late eighth-century Norse invention that was common in the Viking Period, the Bressay examples being comparable to Jellinge style Norse an of c.900
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSAY
121
[1981:285]. Along with a number of other demonstrably late features, this enabled Stevenson to date the Bressay slab to the ninth or early tenth century.
Lamb is perhaps a little harsh in describing Bressay as 'an unimpressive effort' [ 1993b:267], but he is quite right to compare it unfavourably with Pictish sculpture of the previous century.
Romilly Allen
characterized the interlace as 'not of the best kind either in design or execution* and pointed out the absence of key-patterns and spirals as an indication that it was 'not of the finest period* either [ECMS 9]. Stevenson discussed Bressay in a section on the 'dregs of Pictish tradition' [1955:128]. Romilly Allen pointed out the close connection between Bressay and another Shetland cross-slab, the much earlier and 'infinitely superior' Papil [ECMS 12]. Stevenson put forward a detailed case that Bressay is a late ninthor tenth-century copy and elaboration of Papil 'characterised by a haphazard scatter of decoration and a marked clumsiness of drawing' [1955:128; 1981:284-5]. I cannot agree with Laing that this view 'is simplistic' [ 1993b: 31]. That Papil was still standing, providing sculptural inspiration, in the ninth-century and later is, as Stevenson pointed out, evidence of Christian continuity in early Norse Shetland [1981:289]. That the patrons of the Bressay cross chose to model it so closely on an ancient standing monument is an interesting insight into their perception of their Christian heritage. So many of the features of the Bressay slab are derived from Papil that it would be wrong to stress, say, close similarities between the Bressay lion and the evangelist symbol in the Book of Durrow, since the connection is not independent but mediated through the older slab. The Irish influence which Curie detected at Bressay [1940:78] may be entirely derivative of the Irish-influenced [Laing 1993b] Papil slab. Stevenson argued from the entire corpus of early Christian sculpture in Shetland that 'artistic, and so probably also ecclesiastical, links with [mainland] Pictland were first maintained or renewed, but later ... [from the second half of the ninth century], were supplemented or possibly replaced by contact both with Iona and with Anglo-Scandinavian Northumbria' [1981:289]. The interpretation of the Bressay ogham proposed below similarly implies an underlying Pictish culture with an admixture of Gaelic and Norse or Hiberao-Norse.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESS A Y
122
DESCRIPTION OF OGHAM INSCRIPTION The two lines of ogham occupy almost the full length of the narrow edges of the slab. There is no evidence of any lettering along the shaped upper surface. The longer of the two lines, the right-hand side, consists of twenty-nine letters and starts parallel with thefeetof the lion. The shoner side, to the left and comprising nineteen letters, begins a little higher, level with the mouth of the lion, and ends a little lower. The surfiace is broken above left-/9, but, if the text is not complete as it stands, there would have been room for no more than another letter or two. The letter-strokes take up the whole width of the narrow edge, i.e. the ogham-band stretches from edge to edge. All the strokes are sharp, narrow, V-section, chisel-cut scores. Some are quite deep, all are well-preserved (only right-50 and left-i, 4, and 6 are badly worn). The stem has been re-cut after the cross-strokes and is deeper and broader than them. Occasionally strokes cross the stem slightly, but never far or deeply, and these over-shoots are clearly just a slip of the blade.
Right Edge:
The stem starts cleanly a centimetre or so before the first stroke so we can be sure the
start of the text is intact. All the b- and h-aicme letters slope, to a greater or lesser degree, in the direction of reading, so there can be no doubt that the ogham readsfromthe bottom up, as follows [seefig.forkey to numbering]: 1
Four short strokes to the left of the stem, sloping slightly backward - C
2
Five long oblique strokes across the stem, sloping forward - R.
3
As above - R. The groups 2 and 3 are carefully spaced.
4
Two long strokes, perpendicularly across the stem,flankedto the right of the stem by two short strokes a and b. 4a slopes slightly forward, parallel with the strokes of 5 and 5, 4b is more or less perpendicular to the stem. The proximal tip of 4a intersects with and underlies thefirstcrossstroke at the stem, the distal tip of 4b intersects with and underlies the distal tip of the first stroke of 5. Southesk and Macalister interpreted this group as a unique forfid, some kind of modified
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSAY
123
/o/ sound. More likely is that it is a simple error, which is how all other authorities appear to have interpreted it. Perhaps the carver began to cut the first stroke of 5 (an angled stroke to the right) before realizing that 4 had been omitted. He or she may have corrected the error by cutting a perpendicular cross-stroke on top, continued with a second, then began 5 again, but, mistakenly with a perpendicular stroke (parallel with the previous) rather than the sloping stroke required. The first stroke of 5 was cut again on top. Given the form and texture of the stone it would be difficult to 'erase' such errors, so the carving was left well alone. Strokes 4a and 4b are less prominent than their neighbours, squeezed in between the overlying strokes, and it seems unlikely that they were pan of the original plan. Another possible explanation is that the oghamist meant to carve CRROSSCC but omitted an S in error. The strokes 4a and 4b might then be intended to show that the right side of the cross-strokes of 4 are to be read twice, once as half of O, and again as the middle two strokes of a four-stroke b-consonant, S. There is, however, no precedent for this kind of strategy. 5
Four short strokes to the right of the stem, sloping forward - S.
6
Four short strokes to the left of the stem, sloping backward - C. Each stroke is less sloped than the one which precedes it.
7
As above - C. These strokes are almost perpendicular.
8
A pair of dots flanking the stem - word boundary.
9
Five short strokes to the right of the stem, sloping slightly forward - N. Letters 7 and 9 are rather more widely spaced than would be expected if there were no dots at £, thus the word division was part of the original layout and not a leter clarification.
10
A single long stroke perpendicularly across the stem - A.
11
A single short stroke to the left of the stem, sloping slightly backward - H.
12
As above - H. These two, 11 and 12 are carefully spaced to avoid ambiguity.
13
Three short strokes to the left of the stem, sloping slightly backward - T.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESS A Y
124
14
Three shon strokes to the right of the stem, sloping forward - V.
15
As above - V.
16
Two shon strokes to the left of the stem, sloping backward - D.
17
As above - D.
18
A single stroke perpendicularly across the stem - A.
19
Two shon strokes to die left of and perpendicular to the stem having distal tips which curl outwards away from one another. The first of the two curls back almost to the stem-line. QL ,
20
As above, though longer and straighter with little hooks on the end rather than open arches. The difference in form between 19 and 20 is scarcely significant. For possible value see below.
21
Four shon strokes to the right of the stem, sloping forward - S.
22
A pair of dots flanking the stem - word boundary.
23
Two short strokes to the left of the stem, sloping backward - D.
24
A single stroke perpendicularly across the stem - A.
25
Three short strokes to the left of the stem, sloping slightly backward - T.
26
As above - T. The two, 25 and 26 are well spaced.
27
Five long oblique strokes across the stem, sloping forward - R.
28
As above - R. Again, well spaced.
29
A pair of dots flanking the stem - word boundary.
30
A single long stroke across the stem, angled at the mid-point to point up, ->-- some kind of A (see below).
31
Five short strokes to the right of the stem, sloping forward - N.
32
As above - N. Padel read another single long stroke across the stem, following this (A) though he claimed the portion of the stroke nearest the stem to the left was missing. I could see no trace of it.
Which gives a reading: CRRO( /s)SCC : NAHHTWDDAddS : DATTRR : aNN(-)
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND Left Edge:
BRESSAY
125
Macalister inserted dots between 6/7 and 819 but I could not make these out. Padel gave
only the second pair. More of the stem is visible before the first letter but, once again, we may be sure the beginning is intact. The slope of b- and h-aicme consonants indicates the direction of reading is from the bottom up, as expected [see fig. for key to numbering]: i
A single short stroke to the right of die stem, sloping forward, rather faint, but not in doubt - B.
2
Four long strokes, perpendicularly across the stem - E.
3
Five short strokes to the right of the stem, sloping forward - N. The strokes of this letter cross the stem very slightly, but not enough to warrant reading R.
4
As above - N. This letter is very worn, but not in doubt.
5
A unique character. Five long strokes undulating across the stem, sloping slightly backwards some modified form of I ? (see below).
6
Four short strokes to the right of the stem, sloping slightly forward - S. This letter is very worn, but not in doubt.
7
Four long strokes perpendicularly across the stem - E.
8
Four short strokes to the right of the stem, sloping forward - S.
9
A single long oblique stroke across the stem, sloping forwards - M.
10
Four long strokes perpendicularly across the stem - E.
11
Five short strokes to the left of the stem, sloping backwards - Q.
12
As above - Q.
13
Two short strokes to the left of the stem, sloping forwards - D.
14
As above - D. These four successive letters of the h-aicme (11-14) are unambiguously spaced.
15
Five long oblique strokes sloping forward across the stem, overlain with five identical strokes sloping backwards - the cross-hatched HRforfid, very carefully carved (see below).
16
A diamond-shaped lozenge sitting on the stem - an angular version of ihe-frforjfd (see below).
17
A single long stroke perpendicularly across the stem - A.
18
Four short strokes to the right of the stem, sloping forward - N.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
19
BRESSAY
126
As above-N. Thereafter the surface has worn away. If further letters are lost there cannot have been more than one or two. There need not have been any.
Which gives a reading: BENiSESMEQQDDRoANN(-)
FORM OF SCRIPT Both lines of the Bressay ogham are written in the same hand, a clear and regular Type lie form of the script with word-division dots and a handful of forfeda, two of which are unique. The stem is more substantial than the letter strokes probably, in pan, because it runs along the grain of the stone. The long, straight vowel-strokes take up to the full width of the ogham band. There is variation in the degree to which b- and h-aicme letters slope, some are almost perpendicular while others are sharply oblique, but all slope in the anticipated direction. Component strokes are parallel within letters and evenly spaced. All letters are clearly spaced, often with as much as a stroke's worth between them. Successive letters of the same aicme, for instance right-////2//J, are arranged to prevent ambiguity.
The letters are evenly
arranged along the entire length of the text and, like Lunnasting, there is no bunching towards the end, indicating a degree of planning absent, for instance, at Formaston, Newton, and Buckquoy. The form of right-4 suggests that an error may have been made in its initial carving.
The ratio of stroke-length to stroke-spacing gives the Bressay letters a profile closer to the mean than the rather squat letters of some of the later inscriptions like Burrian, Cunningsburgh 3, and Lunnasting. Like the three aforementioned oghams, Bressay smells of the scriptorium. The proportion of the strokes and the technique with which they were cut strengthens the manuscript-like appearance of the script (though, in contrast to Burrian, Bressay is carved on a monumental scale). The curves of letter left-5 are more suited to the pen than the chisel, and the unattached voids created by letters left-76 and, especially, left-75 are vulnerable to becoming detached and are therefore rather unsuitable for epigraphic purposes (though the carver here knew the stone - none of the enclosed lozenges have been lost). In general appearance Bressay is, perhaps, most similar to Formaston with which it shares angled vowels and the-bforfid, but
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSAY
127
not word division. This last feature is found at Lunnasting, which has bound letters, angled vowels, but no other developed features in common with Bressay. A little closer is Burrian which has bound letters and lacks word-division, but does share with Bressay angled vowels and the RR forfid.
There are fiveforfeda at Bressay, one of which is repeated. Three of these are paralleled on other Scottish monuments and also in Irish manuscripts, but the other two are unique to Bressay:
Character Right 4 (,| | ( ) : This is probably not a forfid but rather a corrected error in carving (see discussion above).
Character Right 20, 21: This rabbit-eared character is unique to Bressay. Since it consists of two strokes to the left of the stem previous authorities have taken it as some kind of modified D, presumably the voiced spirant /Ó/. This seems by far the most likely explanation (and if it is not we can only guess at this character's value), but implies a breakdown in the old convention DD for /d/, D for /Ó/ (presumably the initial D in the following word has the value /d/ because of the practice of avoiding geminate consonant symbols in word-inital position [Harvey 1987]).
Character Right 28 (->): Angled vowels are discussed in the entry on Formaston where it is suggested that they may indicate long vowels. Whatever its value, we may take it as some form of A, contrasting with the standard A-character occurring in the previous two words (/0, 18, 23).
Character Left 5: (12ZZZ) This unique character, five undulating strokes sloping backwards across the stem» is presumably a vowel, since it occurs between the second and third of a group of three Ns. As a fivestroke vowel we may take it as some form of I, rather than R. Angled-stroke I (?/t/) does not occur at Bressay, but since angled-stroke A does (?/5/), we may have warrant for rejecting /f/ as the value of this character. Though it is really rather different from them, it may be useful to compare úúsforfeda with
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSA Y
128
instances elsewhere in Scottish oghams of cross-strokes with flaps, e.g. Birsay 1-2, Latheron, Lochgoilhead, though most of these face in the opposite direction.
Character Left 15:
A character consisting of/awr-stroke cross-hatching ()$$) appears in the late
eighth or ninth century Bern ogham alphabet and syllabary under a label which has previously been read as RR. Sims-Williams has argued convincingly that instead this is SS [1992:38-39 fig.5], but rather than remove the manuscript key to the Bressay character's value, he has in fact demonstrated the precedent (with the letters BB and LL) which shows that the five-stroke character here C$$K) represents RR. SimsWilliams suggests that monograms for LL and RR were generated by 'late but keen-eared ogamists' to represent the geminate or long liquids /11/ and /rr/ [1992:72]. The context of this character at Bressay, after DD and before OA, fits Sims-Williams's interpretation, but if the forfid is /rr/, what then is meant by the doubling of the ordinary R character on the other line of text (right 213 and 26127) ? As Macalister points out, thought it may be easier to read, this character is not easier to carve, and, since there was ample space for two characters, we can only assume some nuance of phonology/orthography drove its inclusion. The only other instance of this character in practical use is at Burrian. It appears in a cross-hatched alphabet in the Book of Ballymote [No.64, Calder 1917 Aur. 6085], but there all the letters are doubled and the device has no phonetic significance.
Character Left / 6 : ( - £ - ) This character is discussed in the entry on Formaston [see also Sims-Williams 1992:58-60]. Its earliest vocalic value seems to have been /D:/, though it later came to be used for the diphthong oífóe, and ultimately the digraph oi.
Word-division: The use of pairs of dots to indicate word division is discussed in the Introduction where it is argued that, while a late feature, the device cannot be taken as diagnostic of a date in the Norse period (though we know on other grounds that Bressay certainly does date to the ninth century or later). The suggestion that word-division dots were a feature borrowed from the runic tradition was first made by
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSA Y
129
Graves, in connection with this stone [1855:248-9].
INTERPRETATION OF TEXT The transliteration of H and V is discussed in the Introduction. As for the forfeda, we have some notion of the likely values of right-25, left-75 and -76, but, as for right-79 and -20, and left-5 we are only guessing. For all the problems of both transliteration and interpretation, the use of these supplementary letters suggests an attempt at orthographical precision. Despite its hybrid character, the Bressay inscription is far from slap-dash. Word-division is indicated on the right edge, so we may not introduce further divisions, but, pace Macalister, it is not indicated on the left edge and must be supplied. A further question is the order in which the two sides are to be read, if indeed they are to be taken sequentially. In the earlier oghams it might have been natural to give the left side precedence, but in some later examples this old convention appears to be breaking down (cf. Brodie, Formaston, Golspie). We might take the longer line first, but the allocation of letters might simply have followed word-boundaries* in which case this is not a sure guide. It could be that the two are separate sense units and it does not matter which goes first, certainly no indication is given on the inscription itself.
The interpretation of the text is greatly helped by the word-division on the right edge: CRROSCC : NAHHTVVDDAddS : DATTRR : aNN
CRROSCC All have recognized this as meaning 'cross' (Latin < crux, see entry on Burrian). What is of particular interest is that the form here /krosk/ is virtually identical with the crosg an attested variant of the more common Modern Scottish Gaelic crasg, genitive croisg (as opposed to the Old Irish cross. Modern Irish cms). The final g has not been satisfactorily explained but is standard [Padel 1972:31]. Watson gives a sole example oicrasg meaning a carved stone cross, in a place-name recorded in the seventeenth century of a monument near Dornoch [Watson 1929:486 n. 1].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSAY
130
NAHHTWDDAddS The final S of NAHHTWDDAddS looks suspiciously like a Germanic genitive case ending, doubly so since the word appears to be in syntactically genitive position - '(the) cross of Nahhtuuddadd\ Following Graves, Brash suggested [1879:355] that the second word at Bressay might commemorate the Naddodd 'of the Faroes' who, according to later Norse sources accidentally discovered Iceland when blown off course en route to Orkney in 861 [Landnamabok 4, ch.4; 26, ch.63; OlafTryggvisson's Saga ch.113 (ch.199 in Flateyiarbok 1.248); cited by Anderson 1922:338]. possibility is intriguing.
We are, of course, one syllable short, but the
The name Naddodd is presumably composed of naddr 'arrow' + oddr
'spearhead', and though not otherwise attested, is nonetheless a regularly formed Norse name [Lind 1905 s.v. Naddoddr, I owe this reference to Anne Haavaldsen, University of Bergen, and thank her for her generous advice on Norse nomenclature]. The genitive form would be Naddodds, Lind gives the spelling Naddodz which may support the reading of characters 19 and 20 as /5/.
If this name is correctly identified, the spelling appears to have become contaminated by ogham spellings of names related to Nechtani'Nation, NEHT-, NEHHT-, NETU (see entry on Ackergill). The similarity to the Norse name may, however, be fortuitous, and Bressay's NAHHTVVDDAdd(S) may be a Pictish name after all. NAHHTVV- is most closely paralleled by Fonnaston's NE4HHTV-.
In both cases a
Gaelicizrog transliteration F(F) seems less plausible than the more Pictish /w/, which renders unlikely the interpretation of DDAdd as the Old Irish diminutive ending -that, which, in any case, 'is found only with nouns denoting inanimate objects' [Thurneysen 1946:175 §274.4]. On a suggestion of Stokes [1890:40], Rh?s proposed taking the last eight characters, VVDDAddS, as Unddaóó, a less Gaelicized spelling of the name which appears elsewhere as Fothad (Fothud, Fothaid) [1892:297]. The name is well attested in Ireland as a personal name [CGH] and was borne by two Scottish bishops, Fothad son of Bran scriba and bishop of the Hebrides who died in 963, according to the Annals of the Four Masters [quoted by Anderson 1922:471-2], and Fothud ardepscop Alban, bishop of St. Andrews, who died in 1093 (AU). According
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSAY
131
to Irish genealogical tradition, the ancestors of the Uithne of Nonh Munster were Na Tri FotMd, the three Fothads, who were reckoned to have come from Scotland and were linked with the Tuatha
Orcandlbdig,
the tribes of Orkney and the Outer Isles [CGH 155 b 26 -156 a 4 (pp. 264-5)]. We may take it that Fothad < Uotados, the pseudo-eponym of the Uotadini > Gododdin. Bressay 's DD would suggests a non-lenited first consonant, Nait-Uudadd perhaps ?
DATTRR Previous authorities have taken this as a form of the Old Norse dottir 'daughter', while querying the likelihood of such a loan. Macalister maintained that, as a 'Bronze Age' population, the Shetlanders had no word for 'daughter' until the arrival of the Norse, and acquired mac from the Irish only when the adoption of Christianity necessitated an equivalent of Latin filias [1940:203]. This outlandish assertion, founded on a belief in non-Indo-European Picts practising matriliny, may be dismissed as a fantasy. There is no need to posit the borrowing of vocabulary for such a fundamental human relationship when it is possible that in this context die word could be a quasi-proper noun. If the person in question was habitually referred to not by their given name, but by the patronymic Naddodsdottir, then the dottir might come to be felt to be pan of their name, and thus untranslatable, rather than a common noun in a descriptive phrase. The spelling DATTRR is some way from dottir, but might reflect a Gaelic-speaker's attempt to render a word which had been heard but not seen written. Certainly, if it is not dottir. a convincing Celtic alternative is hard to find.
ANN In the light of the standard interpretation of the previous word, this has been taken as the female personal name Anna (cf. Modern Welsh Ann). Biblical names for both men and women appear to have made little impact in pre-Norman Scotland and Ireland and, if this is a Gaelic name, it is perhaps more plausibly linked to Ann the name of a female Irish deity and saint, or the more popular Áine [[Ó Corráin & Maguire
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSAY
132
1990:s.v.v.]. There is no evidence, however, for a final vowel [pace Diack and Padel]. But perhaps we have been misled in thinking the name must be female, An(n)y gen. Ans, was a Norse male personal name, ('a very, very old name1 [Lind 1905]), borne, for instance, by a man (and his maternal grandfather) mentioned in Landnamabok as plundering and marrying in Ireland before setting off for Iceland [Landnamabok ch. 107.p.45, cited by Anderson 1922:336]. The Old Norse female name On means 'hope' [Lind].
It is noteworthy that both lines of ogham at Bressay end in -ANN, if this is more than coincidence it may indicate that we are not, in fact, dealing with a name at all. In Modern Irish and Scottish Gaelic the adverb ann 'there' is used with the substantive verb to denote existence. No such verb is present. We could take an elided tha as read at the head of the sentence \thd) crosc Nahtudads dattrann\ but to state the obvious in this way seems an unlikely text. For reasons of syntax we may reject any form of the definite article (Goidelic or Brittonic, cf. Cornish and Old Breton unstressed definite article an [Jackson 1953:656-7].
Turning to the left line of ogham, there is not indication of word-division, which must be supplied. The sequence MEQQ in the middle immediately suggests the familiar *X mac Y? formula, i.e. BENiSES MEQQ DDROANN.
MEQQ Presumably this is Irish mac 'son'. The same form is encountered elsewhere on Shetland at St Ninian's Isle and, if not a northern dialectal variant of the MAQQ encountered further south, may be a fairly straightforward rendition of genitive meicc (with the anachronistic Q a fossilized relic of the old spelling reserved for ogham only, as standard in Scottish oghams).
Macalister's ingenious suggestion that
BENISES-MEQQ against DUNNODNNAT-MAQQ, reflects vowel affection is not refuted by St. Ninian's Isle's -ES-MEQQ, but seems highly unlikely.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESS A Y
133
BENiSES The precise value of 5 is not known, though we may guess at some kind of I. Brash pointed out that Naddod is recorded as having a grandson Beinir [1879:255]. Beinir means 'to straighten out', and appears in Scaldic poetry meaning 'helper, giver' [Lind 1905]. The genitive form would be Beinis, which, once again, seems one syllable short of what we have here at Bressay.
Ferguson calculated that if the
identification of this Benir were correct the inscription would have to be mid-tenth century, which need not be too late [Ferguson 1887:134-5]. Another Beinir appears as the great-great-grandson of Erp, son of Maelduin 'an earl in Scotland', given his freedom by Aud the deepminded and who sailed with her to Iceland [Landnamabok chs.82-3, Sturla's version ch. 103 p. 158, cited by Anderson 1922:384]. He cannot be the person commemorated, but further attests the use of the name. The Anglo-Saxon Chronicle of 911 records the death in a battle between Anglo-Saxons and Norse, of one Benesing [cited by Anderson 1922:401]. One looks in vain for a Celtic name of this form.
It is possible, however, that further segmentation is required, and that we should separated off the initial BEN, which could be Old Irish ben 'woman', 'wife'. Could we have 'the wife of Ises son of Droan'. Ises does not bring to mind any Celtic personal names. The name / (//) appears in a Bemitraige genealogy [CGH 111 b 50], but if Ises is the genitive, this is not Celtic. Ben- occurs in composition with other nouns to denote a female, or something pertaining to the female (the more common composition form is ban-, e.g. ban-ech 'mare', ban-liaig 'female physician' [DIL]), but what ises in ben-ises could mean is unclear. It seems that taking Ises as a form of Jesus, invariably fs(s)uffsa in Irish, may be going a bit far, tempting though it might be to interpret ben-Ises as 'wife of Christ', i.e. 'nun'.
DDRoANN(The value of character 76 remains doubtful, the correct transliteration might be D(d)ro(an(n)y D(d)róean(n)> or D(d)rdan(n)y and even the segmentation is not secure: it may be that we should separate off the final
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESS A Y
134
ANN (D(d)roi I D(d)roe I D(d)roi an(n)). The geminate initial consonant is surprising and perhaps we are to take MEQQDDRo(ANN) as a single unit, not mac Dro(an), but Mac-Dro(an). Drón, genitive Drain, is well attested as an Irish male personal name [CGH 154 d 49, LL 316 b 59; 317 b 13; Ó Riáin 1985:704.79], cf. ogham DROGNO (CIIC 167) > Drón, gen. Dróna [McManus 1991:107], hence the tribe of Ul Dróna who gave their name to the Baronies of Idrone, Co. Carlow [McManus 1991:179 n.34]. For females there is Dron [CGH 157 42] (both names perhaps related to the common noun dron 'solid, firm, substantial, vigorous' ? [DH]). If the name were Drón or gen. Drain, why then, Bressay's A ? The ANN can scarcely be the diminutive -en (the diminutive is Drónán [Ó Riáin 1985:84, 662.48]). Unless we are to posit a rather tortuous spelling, we must look elsewhere.
If Bressay's characters 13-16 are to be transliterated Drol than we may indeed have a form of Irish dnd (later drai and droí) 'druid', 'magus', as Graves suggested, though syntactically we would expect, rather, the genitive druád [DILJ], Brash compared MEQQDDRO(i)ANN to the Ere mocu Druid7, a thief from Coll, named by Adomnán (VC I 41).
In this connection, Rhys drew parallels with ogham MAQI-
DROATA (CIIC 503, Isle of Man). He may have been closer the mark in citing the runic TRUIAN on a rune-stone from Bride, Isle of Man (D is not used in Manx runes) [Kermode 1907:169 No.92 pl.xlii]. Kermode pointed out that the Druian of the inscription may be commemorated in the name of an adjoining quarterland, Glentruan [ibid.]
CRROSCC: NAHHTWDDAddS : DATTRR : aNN | BENiSESMEQQDDRoANN Macalister put forward an elaborate explanation of the Bressay text, replete with suffixes, based on his theory that Pictish was a semi-agglutinative non-Indo-European language [1940:204-5]. This can be safely dismissed. All scholars, Macalister included, have identified MEQQ and DATTR as loanwords, from Old Irish and Old
Norse
respectively,
the identification
of
CROSSC is
also
fairly
secure.
NAHHTVVDDAdd(S) is surely a personal name, similarly (BEN)ISES (if BEN is not 'woman' /'wife') and
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSAY
135
DRo(ANN), and possibly the ANN of the right edge. How these words fit together, however, remains to be explained.
Previous authorities have interpreted the two lines as '(the) cross of Naddodd's daughter, Ann. (Of) Benises son of Droan'. Some or all of die proper names are Norse, but the syntax is not wholly Norse. Some or all of the linking common nouns appear to be Irish (crosg, meic, lann), but the morphology and syntax are not wholly Gaelic. As Southesk pointed out, the text conforms to neither Irish nor Norse epigraphic formulae. An example of an appropriate Norse syntax is seen in the runic text on the cross from CilleBharra (Kilbar), Barra (Outer Hebrides): Eftir Porgerdu Steinars ?dóttur es kors sjá reistr. The woman commemorated is / orgerdu Steinars dóttitr, 'Porgerdu daughter of Steinar'. If Bressay commemorates a Norse woman named Ann, then we would expect her to be referred to as ANN NAHHTVVDDAddS DATTRR, 'Ann. Naddodd's daughter'. If NAHHTVVDDAddS and ANN are names, then the syntax NAHHTVVDDAddS DATTRR ANN is more reminiscent of the Irish formula X mac Y / X ingen Y. Following this model, NAHHTVVDDAddS DATTRR ANN would be 'Nechtudad, daughter of An', an interpretation which at least has the advantage of one attested name, and it of the appropriate gender for the person bearing it. If the S is a genitive ending then it may relate to possession of the cross not the daughter, 'the cross of Nechtudad, daughter of An', but why Norse morphology in an otherwise Gaelic or Gaelicized text ? If the text on the other side is ben Ises meicDroan, it may give further information about Nechtudad, 'and wife of Ises son of Droan'. If this is rejected, Benises meqq Droann may be the name of the man who erected the cross (her husband or son, perhaps), or the man who made it. If instead it is Benises Meqq-ddroann* we might have a simple list of the names of the two men involved in the erection of the cross. Taking both sides together we might even have 'The cross of Nectudad's daughter. An, Benises, MacDroan', the honorand named solely by her relationship to a man, commemorated by three men (her sons?). Jesch has discussed the well-established practice of wealthy widows in Sweden and Norway of the eleventh century and later erecting rune-stones to the memory of their dead husbands. To give just one examples, she cites the stone at Alstad with its partly metrical text: 'Jorunn raised this stone
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESS A Y
136
after..., who had her to wife and she brought the stone from Ringerike, out of Ulvoya | and the stone will honour them both' [1991:71]. Turning the tables at Bressay, the text might mean 'the cross of Nectudad, daughter of An (in memory of her husband) Benises son of Droan', or, if Ann is a woman, 'the cross of Nectudad's daughter An, (daughter of) Benises son of Droan', cf. the rune-stone from Dynna, Norway, in which a Christian mother commemorates her dead daughter [ibid.].
DISCUSSION It is a striking feature of early Irish inscriptions (ogham and non-ogham) that they seldom, if ever, commemorate women. Norse runic inscriptions, in marked contrast, were frequently commissioned by or for women [Jesch 1991:48-74], This is equally true of the mixed Celto-Norse milieu of Norse colonies in the British Isles. The cross from Cille-Bharra, which commemorates a woman in an entirely Norse text on a Celtic form of monument, has already been mentioned. A remarkably high proportion of the twentysix extant runic inscriptions from the Isle of Man commemorate dead women, as mothers, fostermothers, daughters, and wives [Kermode 1907; Jesch 1991:74], Of particular interest is the so-called Mai Lumkin cross, from Kirkmichael which, in addition to its runic text has two ogham inscriptions, one an alphabet, the other, though very worn and not yet satisfactorily interpreted, is probably personal names, [Kermode 1907:100-1; Macalister 1940 No.20; Padel 1972:110-111]. The runic text on this possibly Pictish-inspired cross-slab well illustrates the mixed cultural milieu of contemporary Man [Kermode 1907:195-9]: MAL : LUMKUN : RAISTI: KRUS : ThENA : EFTER : MAL : MURU : FUSTRA : SINE : TOTTR TUFKALS : KONA: i s : AThlSL: ATI x 'Mael-Lomchon raised this cross in memory of his foster mother MaelMuire, the daughter of Dufgal, the wife of Auógisl' BETRA : ES : LAIFA : FUSTRA : KUThAN : ThAN : SON : ILAN X 'It is better to leave a good foster son
than a bad son.' Not only does everyone here, except the honorand's husband, bear a Celtic name, but there are 'several glaring grammatical errors' in the Norse [Kermode 1907:198].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSAY
137
As Page has said, the Manx runic inscriptions 'apply a common Norse formula of commemoration, with a common Celtic variation of wording' [Page quoted by Jesch 1991:73]. A similarly hybrid epigraphic tradition appears to be reflected in the Bressay ogham. The very fact that it commemorates a woman may be taken as a Norse symptom. Bressay affords an insight into ninth- or early tenth-century Shetland society, commemorating a family of Christians created by inter-marriage between Norse and Celtic-speaker. We see an underlying Pictish stratum, but with heavy Irish influence brought in either long ago through the church [Laing 1993b], or more recently by Hiberao-Norse Vikings. This picture contrasts with posteighth century Orkney, where the tradition of Christian Pictish sculpture ceases with the arrival of the Norse [Stevenson 1981]. It may have been in Shetland that the Norse developed an interest in ogham, the only post-ninth century epigraphic oghams survive from Celto-Norse contexts (Killaloe slab, Dublin comb, later Manx oghams).
BIBLIOGRAPHY
OS Record card - HU 54 SW 12
Charlton 1855a ill., 1855b; Graves 1855, 1856:248; 1858 ; Smart 1856:30 pl.94, 95; Wilson 1863:240 ill.; n.a. 1865 pi. 11-12; Goudie 1878:20; Brash 1879:354-8; Southesk 1884:186-98; Ferguson 1887:134-5; Rh$s 1892:296-302, 1893:411; 1898:377-80; ECMS 5-10 fig.4 a-d; Goudie 1904:35-8; Nicholson 1896 L; Macalister 1940:202-6, pl.IIIf; RCAHMS 1946:2-3, No. 1084, fig.462,476-7; Diack 1944 xiii; Jackson 1955:140-2; Padel 1972:62-7; Stevenson 1981:284-5; Close-Brooks & Stevenson 1982:35 ill.; Crawford 1987:169, 170, 171 ill. fig.60; Laing 1993:31.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BRESSAY
138
PREVIOUS READINGS
Graves
CROSC NAHDFDADS DATR ANN | BENRES MECCUDROI ANN 'the cross of Natdodd's daughter here. Benres of the sons of the Druid here*
Southesk
BERNISESSJ: MEOODDRROIANN CRRQESCC : NAHHTFFDDADDS : DATTR.ANNB['the body of Bemis, son of Dru, rests here. The body of Krusa, Natdod's daughter, rests here, wife of....'
Ferguson
CRRObSCC : NAHHTFFDDADDS : DATTRR : aNN BENNRpES : MEQQDDRROI ANN
Rhys
CRROSCC : NAHHTVVDDADDS : DATTR : ANN BENNISES : MEQQDDRROANN 'Cross of Great-Vudda88's daughter, wife of Mac-Drroanf
ECMS
CRROSCC : NAHHTVVDDADDS : DATTR : ANN BE(nn/rr)ISEF : MEQQDDRROANN
Macalister
CRROSCC : NAHHTVVDDA(dd)S : DATTRR : ANN BENNIS : ES : MEQQDDrOANN
Diack
CRROSCC NAHHTVVDDACS DATTRR ANNA BERRISESt. MEQQDDRROANN
Padel
CRROSCC : NAHHTVVDDA(dd)S : DATTRR : ANN(a) BENNISES : MEQQDDRROANN
•Back'
'Front'
Brooks & Stevenson 1982:35] .inscribed cross-slab IClose BBESSAY-Ogham
<*>!
I
&
Iff!
i
'-LJ^Jy
BRESSA Y - Ogham-inscribed cross-slab [n.a. 1865, pl.XII, based on Stuart 1856, pl.XCIIII-XCV]
Cross-slabfromPapil, Shetland ["Trustees of the National Museums of Scotland, Reg. IB 46]
BRESS AY - Ogham inscription on right edge
32 31 30 29 28 27
26 25 24 23 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7
A fc
I
i
6 5 4 3
2 [trustees of the National Museums of Scotland, Reg. IB 119]
a
f
(key to numbering)
BRESSA Y - Ogham inscription on left edge
19 18 17 16 15
i4
12 10 9 8
q
i 1
7 6
2 1 [•^Trustees of the National Museums of Scotland, Reg. IB 119]
(key to numbering)
139
BRODIE
DISCOVERY This sculptured cross-slab, also known as 'Rodney's Stone', was uneanhed in 1781 during excavations for the foundations of a new parish church at Dyke, Moray (formerly Elginshire).
The new building was
constructed behind its pre-Reformation predecessor and the stone evidently came from the graveyard of the old church (NGR NH 990584). It was re-erected in Dyke village the following year in commemoration of Admiral Rodney's victory over the Count de Grasse at the Battle of the Saints, hence it's name. An alternative tradition concerning how die stone got its name is preserved in a note made by J. Grahame Callander in the margin of the copy of ECMS in the library of the Society of Antiquaries of Scotland in Edinburgh. Citing the authority of the Rev. John MacEwan the minister of Dyke, he states that the stone was dug up by a gravedigger 'locally known as Rotteny', and that it was after him, not the Admiral, that the stone was so called [OS Record card]. Black lists no such name in his definitive survey of Scottish surnames, the closest is Rothney an uncommon surname derived from the place in Premnay parish ABD [1946:701]. For all that it is now beyond proof, the admiral seems the more likely eponym than the gravedigger [pace Shepherd & Ralston 1979:31]. George Brydges Rodney (1719-1792) enjoyed an illustrious naval career, during which he captured Martinique, St. Lucia, and Grenada from the French and relieved the Spanish siege of Gibraltar. His victory off Dominica in 1782 was no less than a crushing of the French fleet and earned him Parliament's thanks and a peerage [DNB]. It seems entirely plausible that popular enthusiasm for the first Baron Rodney should cause his name to be attached to the stone discovered at the time of his greatest victory.
Some years before 1842 Rodney's Stone was moved to the grounds of Brodie Castle [NSA Elginshire 221], which is now the property of the National Trust for Scotland. The stone still stands there beneath a wooden shelter at the side of the castle drive (NGR NH 9842 5872) [at the time of writing plans are afoot to move the stone indoors at Brodie Castle, not before time]. As it currently stands, the broken lower edge
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
140
rests on a modern base and the slab is held upright by four wrought-iron struts which partially obscure sections of the inscription. At some point in its history, perhaps in the seventeenth or early eighteenth century the cross-slab was inverted and re-used as a tombstone. On the 'front' face two recessed panels were cut into the interlace designs either side of the cross-shaft, and two pairs of initials incised. The back was unharmed except for a broad recessed band cut along the top edge which carryied away the end of both lines of ogham on that side. The bottom of the slab may have been lost at this time if the stone was deliberately cut down to fit a particular space. But if the slab had already fallen or been pulled won the botom might have long since sheered off. The greater weathering of the cross-face, in comparison with the comparatively better preservation of the reverse, is to be expected if the stone lay as a grave-cover with the cross uppermost.
The ogham inscriptions on the margins of both front and back faces, though clearly visible, went unrecognized for almost a century after the discovery of the slab. Stuarts's plates give only a vague indication of horizontal and oblique strokes. Writing in 1886, Southesk records how the first intimation that these marks were ogham arrived 'some months ago' in a letter addressed to Joseph Anderson, Keeper of the National Museum of Antiquities 'by one of his more distant correspondents' [1886:20]. Southesk visited the slab and published the first account of it in the Proceedings of the Society of Antiquaries of Scotland.
LOCALITY Dyke is situated on the fertile coastal plain of the southern shore of the Moray Firth, a little west of Forres. As Ross [1983] and Jones [Jones et al. 1983] emphasize, modern topography gives a misleading impression of Dyke's original setting. It is now well back from the sea, from which it is separated by the thriving forest of Culbin. Old photographs of this 28 sq.km. area of blown sand between the mouths of
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
141
the rivers Findhorn and Nairn, show that up to the turn of the century it was a 'desert-like wilderness' with dunes up to 30m high [Ross 1983:187]. The extent of the sand has varied over the years and it is difficult to reconstruct its range in the early medieval period. Aerial photography reveals the extent to which the area immediately south of the sands was favoured for settlement in prehistory [Jones et al. 1983; see map following entry on Altyre]. Over the years a number of items, including shards of domestic pottery, beads, and jet rings, dating to the Pictish period have been found amongst the dunes, not to mention the famous bronze snake armlet of the first or second centuries A.D. [Ralston & Inglis 1984:37-8, 43, 45].
The place-name Dyke has been interpreted as simply the English 'dike' [Macauley 1976:260], but Nicolaisen comments that 'one would not expect an English name of such importance to have been recorded in the twelfth century in this pan of the country. Is it possible', he wonders, 'that an English or Scandinavian loan-word in Gaelic was employed when the name was coined ?' [1993:262 n. 18]. In the central Middle Ages Dyke was a thanage [Grant 1993:72; McNeill & Nicholson 1975 map 21], and the Anglo-Saxon influences detectable in the thanage system in Scotland are well-known [Muir 1975:26; Barrow 1973]. The names of the seventy or so thanages appear not to be new coinings, but are often very ancient (taken from specific places rather than districts [Muir 28]). Many have names of Pictish origin. There is no direct evidence for the standing of Dyke in the early medieval period, but Driscoll's emphasis on the thanage's roots in Pictish social institutions [1991:108] raises the possibility that Dyke was a centre of Pictish secular administration. Dyke church sits on a large mound within an oval enclosure, two features usually taken as diagnostic of an early medieval antecendence. The additional evidence of the cross-slab may indicate that here was the principle ecclesiastical site in the putative Pictish shire, maybe a monastery, or perhaps a minster-like foundation.
DESCRIPTION OF OBJECT A large Class II symbol-inscribed cross-slab, sculptured in relief on both sides with oghams incised up three of the vertical edges [see fig.].
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
142
Stone:
Grey sandstone
Dimensions:
Height: 1.875m (left) 1.905m (right); Width: 0.95m (top) 1.05m (bottom); Thickness: 0.1 m
Condition:
Broken at bottom, portion missing. Minor defacement due to modern re-use. Face badly weathered, carving on reverse well preserved. Ogham severely damaged and legible only in short snatches.
Rodney's Stone is a fairly typical Class II Pictish cross-slab with a relief cross on one face and symbols on the reverse. It is slightly broader and thicker at its base than its top which, paradoxically, gives the visual impression that the sides are parallel, and has the added advantage of enhancing stability. The top of the slab is weathered but substantially intact (the surface has been cut back in the area immediately below the top). The lower edge, however, is broken, and from the design on the back it is clear that a portion is missing. At least a further 100-150mm would be required to accommodate the lower arm of the Z-rod. The stone does not appear to have been deliberately defaced beyond the modifications described above made when it was re-used as a grave-slab \pace Macalister 1940:200]. A deliberate toppling at the time of the Reformation or Covenanters may account for the bottom having been snapped off, but equally the elements and straightforward neglect may be to blame. The face is considerably more weathered than the reverse which is in keeping with its having lain face up when in use as a grave-cover. The upper third is the most severely weathered on both sides, and the lower third, the least affected. The condition of the friable sandstone is not helped by the very damp conditions of its current location.
DESCRIPTION OF CARVING - DECORATION The thin Brodie slab has been carefully dressed and, apart from the incised ogham, all its carvings are in fairly high relief. Both these features point to a date later in the Pictish period [Shepherd & Ralston 1979:31]. The whole of the front face is taken up with a cross with small rounded hollows at its armpits (Romilly Allen's type No. 101a, Cramp 1984 All). This is the most popular cross-shape on Pictish slabs.
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
143
The whole of the cross is outlined in a thick roll-moulding, which is also found up the left-hand edge. On the right edge is the broad, flat, raised band which carries the ogham inscription (Allen thought there was ogham on the left also, but if there ever was, there is no trace of it now). The cross is decorated with accomplished, complex, and delicate interlacing, unfortunately obliterated in places due to lamination. The pattern consists of a single continuous band arranged in a series of knots. These have been analyzed by Romilly Allen as follows: three circular knots (No.433) arranged vertically in the centre of the cross head, above and below it; in the upper arm and again at the top of the shaft knot No. 714 repeated twice in each of two vertical rows (this knot is also found at Collieburn SUT, Glamis 2 FOR, and Gattonside ROX); the same knot, but without the circular band around it, recurs in die narrow parts of the horizontal arms; these arms have different patterns, to the left Stafford knots interwoven with extra bands placed facing in opposite directions (this knot, No.613, is found also at Nigg ROS, Glamis 2 FOR, and Meigle 4 PER), to the left is interlace composed of Nos. 265 and 431; towards the bottom of the shaft are double Stafford knots with extra bands repeated twice, above and below that, knot No.613; the whole of the shaft is arranged in two columns. It is interesting to note that Brodie should share two of its interlace patterns with Glamis 2, but it is difficult to know what to make of it. There is a certain geometric logic to such patterns and we need not assume borrowing or a common model. In each of the four background quadrants created by the arms of the cross there are interlaced beasts, in pairs and singly. The top right quadrant is almost entirely obliterated, and each of the others has suffered severe weathering, but glimpses of the original designs remain. Towards the bottom of the two lower panels squarish recesses have been cut in the modern period and the initials AC (right) and KD (left) incised, these are inverted relative to the cross.
The back is dominated by a huge pair of sea-monsters which occupy the full width of the slab and almost half its length. Below that are two massive symbols, a Pictish 'beast' and a rdouble-disc-and-Z-rod\ All are highly decorated in relief. The monsters have eyes and large jaws with rows of teeth. At their gills they have spirals and their necks and faces are decorated with triquetra knots and spirals. Their bodies are peppered with tiny bosses representing scales, a feature Stevenson diagnosed as late [1955:121]. Allen
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
144
calls these fish monsters, Stevenson refers to them as 'hippocampi [ibid.]. Since they lack both limbs and fish tails neither label is entirely accurate: Chambers' dictionary defines hippocampus as a 'fish-tailed horse-like sea-monster'. A monster of the Brodie type occurs opposite a creature with forelimbs and a fish tail on Meigle 1 PER, so we may take it that the two creatures were distinct, if related types (see Aberlemno 2 for perhaps the finest example of the hippocamp type). These here at Brodie could as easily be land serpents as water-snakes. The motif of squat, affronted, limbless creatures with large jaws and spiral tails appears on the cross-slabs from Bressay, Skinnet CAT, Mortlach BNF, on the Maiden Stone ABD, at Kettins FOR and Meigle 1. A single such creature, much debased, appears on Menmuir 1 FOR. Allen considered these creatures to be symbols, a view with which Stevenson concurred. While they may have had a symbolism analogous to the kind of symbolism supported by the stereotyped single horseman, I think it is wrong to classify this motif along with the Pictish symbols proper (the double disc, the 'flower', the serpent and Z-rod, etc.). For one thing hippocamps/sea serpents never occur on Class I stones, and they occur on stones which have no other symbols; but, crucially, they never combine with other canonical symbols to form symbol pairs and thus fail to match my definition of a 'proper' symbol [Forsyth 1996b].
The other two motifs at Brodie, the Pictish beast and the double-disc, are definitely proper symbols, in fact they are two of the three most common symbols. Their appearance here as a pair is typical. This exact combination occurs on two Class I stones from Aberdeenshire, Dyce and Kintore 2 (front), on a Class II slab (Eassie FOR), and in an anomalous triple on the St. Madoes 1 cross-slab PER. In all except the St. Madoes example, the beast is the upper symbol. The combination beast above rodless double-disc appears once on the cross-slab at Shandwick ROS. The decoration of double-discs is usually too simple to facilitate stylistic dating. The Brodie example has each disc filled with, according to Romilly Allen, eight double spirals with C-shaped connections (No. 1108), and small pellets in the spaces between them. There are further pellets in the angle of the Z-rod. The form of the beast (long thin body, small head, straight
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
145
squared 'beak') and its internal ornamentation (merely decorative knot-work - interlace spiral knots No. 1108 and a Stafford knot at the head - rather than the integral scrolls of the finer earlier examples) point unequivocally to a later date [Stevenson 1955:121]
Between the pair of opposed monsters there is an assortment of curvilinear motifs, Allen's 'curious circular objects or ornaments' [ECMS 133]. In the centre is a large spiral disc decorated with pellets, above is an object which looks a little like a penanular brooch with expanded terminals, to the right is a crescent with pellets, and to the left a circular disc decorated with a triskele, at the bottom between the monsters' tails is another small disc decorated with a triskele. Each item is distinct and individual suggesting that they are intended to represent specific items, rather than merely fill the space with abstract ornament (thought they perform this function also). If the upper item is indeed a brooch, then the crescent may also be a piece of metalwork, like the crescent-shaped silver plaque from Laws, Monifieth FOR [ECMS 280-1]. If the other items also represent metalwork then perhaps we have here two monsters guarding a hoard of treasure.
In his discussion of the development of the 'Boss Style' in Pictish an, Stevenson cites Brodie as an example of the simplification which followed the great elaboration of the later Boss Style exemplified by the great cross-slabs of Nigg, Shandwick, and Rosemarkie. Emphasizing the predominantly northern distribution of later Boss Style, Stevenson dates one of the Rosemarkie stones to the third quarter of the ninth century [1955:121]. We might have deduced from this that he would have placed the Brodie stone in the third or fourth quarters of the ninth century. He may, however, have revised his dating, in a later article he offers a date 'perhaps around 850' for Brodie [1981:287].
DESCRIPTION OF OGHAM INSCRIPTION The entire length of the right margin (front and back) and the left margin (back only) is taken up with a raised band approximately 50mm wide. A stem-line has been incised up the centre of each band with
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
146
strokes to right and left covering its full width. On the reverse face traces of ogham letters are visible along the entire extant length of the right margin. On the left margin letters are visible at least as far as the curl of the monster's tail [see fig.]. Thereafter the surface is too worn to tell and the final quarter of its length has flaked away entirely. The total extant length of the ogham text on the reverse is therefore approximately 3m, but in its complete original state it may have extended up to 3.8m on this side. On the front of the slab ogham is visible on the right margin only [see fig.]. About one third of the way up the slab the margin has laminated away and it is impossible to discern how much further, if at all, the lettering extended on this side. Ferguson describes this ogham as having a 'definite termination' at the top [1887:143], a view echoed by Macalister, but the carving is no longer clear. The arms of the cross intersect with the side margins but, though parts of the margin are still visible at the ends of the arms, these are too worn to see if they might have been carved. The portion of margin above the arm does not appear to have been inscribed, though insufficient remains to be conclusive. Even at a conservative estimate the total length of ogham inscription on this monument originally may have been as much as 4m. The left-hand edge of the front face is decorated not with a broad band but with a narrower moulding similar to that which outlines the cross. This would seem to prove that the three bands were intended to bear ogham from the outset, and thus that the inscriptions were pan of the original design, rather than added later, not that there could have been any real doubt that the three texts were contemporary with one another and with the rest of the carving.
Ferguson maintained that the front-right ogham should be read upwards, but that the other two went down. Macalister thought the oghams on the reverse read up the right and down the left. Both were mistaken. The direction of the angled-vowels and the slope of b- and h-aicme consonants indicates unequivocally that all three are to be read vertically upwards with the reader's head inclined to the left, exactly as one would expect. The three sections may be part of one continuous text, or represent independent sense units. Either way we do not know in which order they should be read. If it could have been proven that the
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
147
back-left inscription did not cover the entire length of the slab, this would suggest that the back-right was to be read first. In the earlier oghams it is usually the left-hand edge which is primary, but later Scottish oghams such as Golspie and Scoonie run up the right-hand edge. The fact that on the front it is the righthand margin which is used in preference to the left implies that at Brodie too, the old convention was breaking down (cf. Formaston for a more problematic example of a right-side ogham). If the ogham on the front is complete as it stands then either it was the final section, or it is separate from the main text. The question of the order of the three sections is, however, largely academic since the carving is so poorly preserved that only one short sequence can be interpreted with any certainty.
The nature of the ogham script is such that a letter damaged, even only partially, is virtually meaningless. Unlike the roman alphabet, it is impossible to reconstruct ogham letters from isolated strokes or pans of strokes, hence the great difficulty of extracting a usable reading from the weathered carving at Brodie. Detailed measurement and analysis might permit a partial reconstruction, but such work is impossible with the stone in its current position, out in the open, beside the road, in a dark wooded corner.
Front left This side is too damaged for certainty, but there does not appear to have been any ogham on it. I follow Padel in labelling the remaining three sides A B C as follows:
A
Front right
The inscription has already started at the fracture and it is possible that letters have been lost from the beginning. The slope of 8 confirms that the direction of reading is from the bottom up, as follows [see fig. for key to numbering]: 1
Three short strokes to the right of, and perpendicular to, the stem - V. The surface to the left of the stem is lost, so Padel's suggestion of AL is not impossible, but, given the spacing is unlikely.
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
148
This appears to be a complete letter rather than the end of a longer one, but the fracture is such that certainty is impossible. 2
Two long vowel-strokes perpendicularly across the stem - O.
3
Five short strokes to the right of, and perpendicular to, the stem - N.
4
There follows an illegible gap of 70 or 80mm, sufficient for about five strokes. Roughly a third of the way along a single long stroke is discernible, perpendicularly across the stem, suggesting either a five stroke vowel, or, since 5 is a vowel, and a double vowel is unlikely, a two or threestroke vowel (O/U) followed by a two or three stroke consonant (L/D/G, V/T).
5
Four long vowel-strokes perpendicularly across the stem - E. It is just possible that a fifth stroke has been lost in the gap - 1 . There are slight traces of carving to the right of the stem beyond this, which might comprise the three strokes of a V, but this would be rather cramped and the rest of the inscription is well-spaced.
6
Four short strokes to the left of the stem, sloping very slightly backwards - C. Macalister though he saw at this point a uniqueforfid similar to the pair of modified Ds on Bressay. No evidence for such an interpretation remains, and it is not in keeping with the general simplicity of the letter forms used elsewhere on this stone.
7
As above - C.
8
Two long strokes across the stem, angled upwards at their mid-point - O. The second of these is approximately level with the top of the recessed panel containing the modern letters AC.
9
After a short gap there is one oblique stroke to the right of the stem, sloping forwards, and then one to the left sloping backwards. These could be the remains of further strokes of the same form as 8. The gap would be sufficient to take one more to make a total of five - 1 . Several other interpretations, however, are equally possible (B+H, a second O, etc.). After this the surface has laminated away. Though the inscription probably continued at least some way beyond this point it is extremely difficult to gauge how far it reached originally. Towards the top, above the cross-
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
149
arm, sections of the right-hand border without ogham can be traced, suggesting the text did not occupy the entire length of the slab, but so little survives that the evidence is inconclusive. Which gives the following very tentative reading: -](v/v + )ON(a + /a + m + /a + b + /a + h + )(e/i)( /v)CC(i/obh/oo)[~
B
Back left
As with the other two sections, the inscription has already begun before the fracture and a section of unknown length has been lost from the beginning. From the break onwards, the ogham reads as follows [see fig. for key to numbering]: 1
Five long strokes obliquely across die stem - R. The first is feint, which, as Padel points out, accounts for Macalister's Z (he was reading down).
2
Two long strokes across the stem, sloping very slightly forward. The first is close to perpendicular (certainly less sloped than the surrounding strokes) and the second is a little faint. They seem too close to be read separately as AM. Padel preferred G, Macalister read O, really there is little to choose between them except perhaps that elsewhere in the inscription vowels tend to slope a little backwards if there is any possibility of doubt.
3
Five strokes to the right of and perpendicular to the stem. The surface of the stone immediately to the left of the stem is lost opposite the first four strokes, the fifth continues across. Since the spacing is so regular, these are unlikely to have been subdivided and therefore we ought to follow Macalister in reading I rather than SA
4
Five strokes to the right of the stem. Padel reported the surface as lost opposite die last two strokes, it has now gone opposite them all. Macalister read R suggesting there was some doubt even then - N or R.
5
Three long strokes obliquely across the stem sloping forward. Thefirststroke slopes a little less than the other two and is a little further spaced from them, whether this warrants the reading AG or MG rather than Gw is unclear. Macalister took the first stroke as a vowel (A) but its slope is
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
150
perhaps too great for that. 6
There follows a very worn section then three or four h-strokes. Since the b-surface is more or less intact we may take it that the patch of wear contained another h-stroke or two. Macalister read two groups of three, i.e. TT, Padel thought C Very likely'.
7
After a short gap which appears clear, the h-surface has laminated off for a distance sufficient for perhaps six or seven strokes. The b-surface is intact for a little over half of this, implying that the first few strokes must have been of the h-aicme.
8
Just below the metal bracket there is a group of perhaps five h-strokes - Q. Padel thought these Very dubious1.
9
Two long strokes perpendicularly across the stem - O.
10
Two (Padel) or three strokes to the left of, and perpendicular to the stem - D or T.
11
Three strokes to the left of the stem sloping slightly backwards - T. The difference in angle may be to differentiate two successive letters of the same aicme.
12
Two angled vowel-strokes pointing in the direction of reading (-^->-) - O.
13
The next section is extremely worn, any strokes are likely to have been to the right of the stem. Padel read three and then a fourth at a slightly greater distance, all perpendicular to the stem - S.
The inscription continues to a point a little higher than the base of the monster's tail, but the rest of the carving is so worn that very little can be said of it. It is over twenty years since Padel made his observations and it may well be that more of the carving was intact when he saw it. 14
Padel read a further two sloping strokes but his account of them is confused: he described them as to the left, saying the h-side (sic) was illegible, and suggesting O or L (sic). His sketch shows them to the right. It seems to me that any strokes are indeed on the left, that it is the b-side which is illegible, and that the letters are to be taken as a T.
15
Two strokes to the right of the stem. These might be the remains of angled vowel-strokes (O). Padel has the first of these perpendicular to the right of the stem but continue across it sloping
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
151
forwards, which is a peculiar arrangement. As he says, the two are surely too close to be MB. Any forward sloping projection of this first stroke is no longer visible. 16
Four strokes to the right of the stem sloping very slightly forward - S. Beyond this point the surface is entirely destroyed, the text may or may not have continued further.
And reconstruction would be so tentative that it is scarcely wonh setting our die alternatives, but the following is possible:
C
-]R(g/o)I(n/r)(ag/mg/gw)(tt/.c)(h+)( /.)qO(t/d)TOst(o/mb/g)s[-
Back right
This is the clearest of the three Brodie inscriptions, but even it is severely damaged in its middle and upper sections. The surface is more or less intact around groups 1- 4 and these are fairly clear, the top layer has, however been lost in the section that follows and groups 5 to 9 are less clear. Only traces of the bottom of the scores of 10 and 11 have survived. Groups 12 - 14 are more strongly present but thereafter very little can be discerned for certain. It appears that originally die text ran the entire length of the slab, but the top has sheered off. It is by no means certain that the text would have begun at the original ground level of die slab (at least a further 100-150mm below the current terminus), but letters may well have been lost from the beginning of die text. Starring at the ground, immediately after the fracture [see fig. for key to numbering]: 1
Four long strokes perpendicularly across the stem - E. The break in the stone crosses the ogham band at an oblique angle, but just sufficient of die surface to die left of the stem has survived between the fractured edge of die stone and the first stroke, to demonstrate that there was not originally a fifth stroke to this letter. The third stroke appears to be angled at the stem ( f | > |) to point upwards in the direction of reading. This mixing of angled and straight strokes within the one letter is unprecedented, but the spacing of the strokes does not warrant die subdivision of the group. The device may be intended to make it clear diat the letter is a vowel, not a member of the m-aicme. Though die fracture of die stone has carried off the portion of the first stroke
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
152
to the right of the stem, the reading is not in doubt. 2
Two strokes to the left of the stem, sloping backwards sharply - D.
3
As above - D. The gradient of 3 is slightly less steep than 2, but this does not affect the reading. The two pairs are widely spaced to prevent conflation as C.
4
A single long vowel-stroke perpendicularly across the stem - A.
5
Five long oblique strokes across the stem, sloping forwards - R.
6
As above - R. Letters 5 and 6 are carefully spaced.
7
Five strokes to the right of, and roughly perpendicular to, the stem - N. Padel thought the. first three slightly more sloped than the last two, but the spacing is enough to prove the reading which is not in doubt. There are undulations in the badly abraded h-surface opposite these strokes but these are much less substantial than the proper letter-strokes of the preceding groups. If the baicme strokes of 7 had continued onto the left of the stem the final two would have intersected with the first of the backward sloping strokes of 5, thus we can be sure that 7 is a member of the b- not the m-aicme.
8
Two long strokes across the stem, following on almost immediately after the final stroke of 7. There can be no doubt that these two are not perpendicular to the stem, but slope backwards, against the direction of reading. This must be to differentiate them clearly from the adjacent, forward-sloping, m-aicme letters. There is no doubt that this letter is O.
9
Five strokes to the right of, and perpendicular to, the stem - N
10
The next two groups are extremely worn. Two strokes to the right of, and perpendicular to, the stem, then, after a gap, a third stroke. The gap is sufficient for two strokes, which would bring the total to five - N. The space between the last stroke of 9 and first stroke of /0, i.e. between letters, is only marginally greater than the average space between strokes within a letter, but the division is in no real doubt. Padel complained that the sloping of strokes in these two groups was not consistent (the first are almost perpendicular, the last slope forwards. The reconstruction of
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
153
the text strongly suggests, however, that this should be taken as a second N. 11
The following section must have contained at least two letters. After a gap there appears to be two long strokes across the stem, sloping very slightly forward. If anything has been lost before these two it could be no more than a stroke or two, but the gap is not that much larger than normal inter-letter spacing. The forward slope on these strokes is so slight that they may be vowel-strokes, in which case O, or possibly U or E, but given the backward slope on the vowel 8 it may be that these strokes are pan of an m-aicme letter. Two strokes would be G, any more would be Gw or Sl but these letters are extremely rare in epigraphic oghams and we should be slow to see them here. Following this group there are four strokes to the right of the stem, again sloping very slightly forward. There is a generous gap between the last of the long strokes and die first of the short, but it is not so long that lost strokes must be posited. The four are far from certain, ponions could have been lost to die left of die stem (there are hints of such strokes but most of die surface is lost at this point). Further entire strokes immediately following it could have been lost, so the reading S or N is highly tentative. It is further possible that the four should be divided into two pairs - LL. Padel made no attempt at this section, and Macalister's claim to see a 'clear' I is dubious.
12
Following a short gap, some clarity returns. Three strokes to the left of die stem sloping backwards. Pan of die wrought iron bracket by which die stone is held upright almost entirely obscures the surface of the stone to the right of die stem, therefore it is impossible to see if the strokes continued across. The backward slope would tend to suggest T rather than U, but tliis is not certain because of the backwards slope of 8.
13
Three strokes to the left of die stem sloping backwards. The third stroke comes after the bracket and clearly does not continue to die right of the stem, confirming the reading T. The propensity for doubling ogham consonant thus supports, though does not compel, the reading of 12 as T also.
14
Five long strokes perpendicularly across die stem - 1 . The fifth is very faint and the second and
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
154
third have lost whatever was to the right of the stem. 15
The next section is severely pocked and abraded. There are hints of diagonal strokes, perhaps even forming the RR-foifid (jg$Q, but this is highly doubtful. There follows an extended section (more than 300mm) of highly degraded text in which nothing but tantalizing disjointed pans of possible letters survives. Nothing is legible.
16
Level with the top of the curl of the monster's tail there appear to be five long strokes sloping forward across the stem - R.
17
A further group of five sloping strokes - R. Letters 16 and 77 are carefully spaced.
18
The next section is badly damaged. There appear to be one or two b-strokes, perhaps B or L, but it cannot be ruled out that they are part of a larger letter, or they may even be vowels.
19
The following strokes appear to be vowel-strokes, perhaps four or five (E/I). They seem perpendicular and thus are less likely to be m-strokes, but neither this, nor the possibility of subdivision can be discounted.
20
The next section is all but gone. There does not appear to be carving to the left of the stem, so b-strokes are more plausible. Since it is not clear when 19 ends and 21 begins, 20 may have contained two, three, four, or even five strokes.
21
Level with the neck of die monster, three strokes are visible to die left of, and perpendicular to, the stem. They may have continued across die stem.
22
After an apparent gap just larger than the usual gap between strokes there are a further four or possibly five strokes similar to the preceding group. These look more like vowel-strokes - E? I?
23
It is clear that there were further letters beyond this, but they are too worn to reconstruct. There is space for another five or six letters before the top of the slab. A recessed band several centimetres wide has been removed from the top edge of the stone in recent centuries, and it is impossible to see how and where the text originally ended.
The first ten letters may be reconstructed with reasonable certainty:
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
EDDARRNONN[-
BRODIE
155
or -]EDDARRNONN[-
thereafter the reading is so uncertain that there is little merit in offering a reconstruction. The best I can manage is:
-(o/a+/g+)(ll/s+/si+)(t/u)]TI[(m+/n)[-
beyond which the carving is too badly damaged to attempt even something as tentative as this.
FORM OF SCRIPT Despite the extreme wear the Brodie inscription has suffered, it is still possible to analyze the form of script used. All three sections are carved in the same hand, a neatly consistent Type lib form of the script. In a couple of places the distal tips of strokes have been lost where the outer edge of the ogham band has flaked off, but even at their full extent the strokes are of no more than medium length. Straight vowelstrokes are long, occupying the entire width of the ogham band. There are occasional angled-vowels, and one case of what appears to be a vowel comprising a mix of straight and angled strokes (CI-
||>f).
In the more legible sections component strokes are strictly parallel and evenly-spaced, and individual letters are, in general, carefully spaced, especially when two letters from the same aicme are side-by-side. In most cases, b- and h-aicme strokes slope in the anticipated direction, the only exceptions are dubious through wear. Weathering cannot always be invoked to account for apparent inconsistencies of sloping between straight vowel and m-aicme letters. In a few instances the former appear to slope backwards, against the direction of reading, perhaps to differentiate them from forward-sloping m-consonants. Others are slightly further forward than upright and it is not always easy to distinguish them from m-strokes. Most of these problems occur, however, in sections which are damaged to some extent.
This is a substantial ogham, neatly andformallycarved, as befits a major public monument. In general terms the closest parallel to the script is found at St. Ninian's Isle, though the latter lacks angled vowels. Scoonie is also similar, though its letters are not as consistently sloped and it too lacks angled vowels. Birsay is closest, but it is such an informal and cursive text that its over-all appearance is quite different.
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
156
Golspie is broadly similar, and equally stately, but it has supplementary letters which Brodie lacks. The inscription is so damaged that we should be wary of arguing from the absence of letter forms, but it is noteworthy that, despite the length of the text, Brodie appears to lack both the x-forftd and the hammerhead A, the two most common forfeda in Scotland.
INTERPRETATION OF TEXT The appropriate transliteration of V is discussed in the Introduction where it is suggested that in inscriptions such as this, which clearly post-date the seventh century, V should be transliterated HI in Gaelic contexts and /w/ in Pictish. Angled vowels are also discussed in the Lunnasting entry, where it is suggested that they may have had orthographic rather than merely palaeographic significance. Their phonetic value, however, remains unclear. Perhaps the most important feature of the Brodie ogham inscription is its sheer length. Even if the three lines of text are to be taken separately as independent sense units, they comprise a long and possibly complex message. Unfortunately we have no models for epigraphic ogham texts which go much beyond the listing of names, patronymics, and gentilic affiliations. Despite the nearly total illegibility of large stretches of text, there does not appear to be a 'MAQQ' at any point, implying that the three sections of the Brodie inscription do not comprise names plus patronymics. We may instead have a list of half a dozen or more simple personal names, no more than loosely connected, but it is perhaps more plausible to imagine text incorporating more than proper nouns. The prospect that the Brodie inscription might include common nouns, prepositions, or even verbs, makes its current impoverished state of preservation all the more frustrating. Only one section can be made out with any certainty, this is the EDDARRNONN at the beginning of C (back right). This is the male personal name Ethernanlldarnon which is discussed, along with the saint who bore it, in the entry on Scoonie. Though the name could refer to an ordinary secular figure, the Brodie text may be evidence of the cult of St. Etheman in Moray. Though the name is apparently Brittonic, this spelling could be either Irish or Pictish. The other sequences are too short or doubtful to throw further light on the language of the text.
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
157
The section immediately following EDDARRNONN may be OLL, which, as it stands, could be either Brittonic or Gaelic. In Welsh oil 'all', 'wholly', cf. Irish idle,followsthe word which it modifies [Evans 1964:98 §107]. Though unattested in inscriptions, it is possible to formulate plausibly appropriate phrases containing the two words in this order, especially since text may have been lost from the beginning as well as the end of the line. Irish oil, 'great', is perhaps more easily integrated into a meaningful text, appearing in afigurativesense, as it sometimes does, following a personal name, eg. Aed ollfii andud n-áne 'Aed great at kindling of radiance' [Stokes & Strachan 1903,295.2, cited in DH]. The title ollamh (the highest grade of poet) seems to be ruled out by the extant remains of letters in thefollowingsection, but oil occurs widely as thefirstelement in compounds with substantives meaning 'great', 'vast'. If oil and what follows is a compound, this would explain thefollowingTT, since a double letter would not normally be expected in word-initial position. If the double T were intended to indicate lack of lenition then /ti(:)r/ could be tlr 'land', 'region'. Given the lack of certainty of the lettering at this point it is perhaps going tooforto suggest tiruairse 'relics', 'fragments', 'saintly relics', cf. túairse [DIL], though if the cross-slab records some major benefaction, or saintly dedication, 'great holy relics' might well be mentioned in the text. An alternative is that olltir is in some way connected with Irish ailithir (pile 'other' + tir 'land') 'pilgrim', 'stranger', which is sometimes spelled oilithir [DIL], though the loss of the middle vowel is a problem. Brittonic cognates of both these words exist, cf. Welsh ail + tir, but that leaves the problem of OLL for AIL. One 'Colman ailithir1 is commemorated in an ogham inscription from Maumanorig, Corcu Dhuibhne (CIIC 193).
DISCUSSION One of the most striking features of the Brodie ogham is its overall length. Taking the three sections together, the text must have been originally up to seventy or eighty characters long. This is considerably longer than the next two longest ogham inscriptions from Scotland (Bressay, under fifty, and Lunnasting, under forty), and certainly far longer than the longest Scottish texts in non-ogham scripts (Newton c.44,
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
BRODIE
158
Tarbat c.55?). This represents a substantial commitment to ogham literacy. That the script was used so prominently on a major public monuments such as this is one of the most striking examples of the high status accorded the script in post-seventh century Scotland. The precise nature of the Dyke site in the ninth century, whether small monastery, minster church, burial ground, or whatever, is not known. It is notable that the two ogham-inscribed slabs from Moray both come from sites with no other explicitly Christian early medieval sculpture (Dyke and Roseisle), and that conversely, two sites with large collections of early sculpture, Burghead and Kinneddar, both lack ogham inscriptions. This dichotomy, which is mirrored elsewhere in Scotland, may reflect a contrast in the nature of the sites involved. Crudely put, on a continuum running from small domestic habitation sites to major monasteries, the more 'ecclesiastical1 a site, the less likely one is to find ogham there. Other Moray cross-slabs from sites with no other early sculpture, Elgin and Sueno's stone, Forres, are notable for the secular themes, hunting and warfare, of their decoration. This may throw further light on possibly secular patronage at Brodie.
BIBLIOGRAPHY
OS record card - NH 5 NE 3
(Rev. M Aitken) NSA 221, 226 (stone only, no mention of ogham); Stuart 1856:9,11 pl.xxii, xxiii (no mention of ogham); Southesk 1886:14-22; Ferguson 1887:142-5; Rhys 1892:287-8,1898:367; ECMS1325; Nicholson H; Macalister 1940:199-201, pl.IIIe; Padel 1972:67-73; Shepherd & Ralston 1979:31, ill.22; Jackson 1984 ill. opp. p. 117; Foster 1996 fig.48 (Tom Gray).
PREVIOUS READINGS Ferguson
Back R (C+B) EDDARRNONNmeQIiJQeN(r/gg)RuN(urb/eel/stf etc.)URH Front
(A)
(e/st)(g/o)NN(m/a)(c)QQDIM(g/t)A(c/t)PLL(u/o)MB(a/o)(n/s)[-
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
Bhfs
Front R (A)
(b/n)ONaroTaTC(o/e)
Back L (B)
(r/f)(am/o)SANAGhhdOhotTOIa(l/g)S
BackR (C)
EDDARRNONn ehhTUMo[-
ECMS
BackR (C)
EDDARRNON[-]Q[-
Macalister
Front R (A)
VONSEOdTO
Back L (B)
MADE OEVV IPENN GARIOZ (reading down)
Back R (C)
EDDARRNON ILCIUDOVORR VITEOR
Front R A
(v/al)ON[-]ECCO[-
Back L B
R(g/am)iN£&chqOOTOS(o/l)mbs[-
Back R C
EDDARRNOnn[»](t/u)(t/u)IHggw[-
Padel
BRODIE
159
BRODIE - Ogham-inscribed cross-slab (reverse face) [Foster 1996 fig.48 (Tom Gray)]
'Front*
'Back'
BRODIE - Ogham-inscribed cross-slab showing position of inscriptions [ECMS fig. 1361
3 2 1
BRODIE - Front right - schematic representation of ogham inscription (key to numbering)
16 15 14 13 12 11 10 9 8
mwm
6 5 4 3 2 1
BRODIE - Back left - Detail of ogham (key to numbering) [After Foster 1996 fig.48]
17 16
14
BRODIE - Back right - Detail of ogham (key to numbering) [After Foster 1996 fig.48]
160
B UCKQUOY
[In modified form the following has been accepted for publication in PSAS 125 (forthcoming November 1996) under the title The ogham-inscribed spindle whorl from Buckquoy: evidence for the Irish language in pre-Viking Orkney ?\]
DISCOVERY The spindle-whorl was discovered by Dr Anna Ritchie in 1970 during her rescue excavation at the Point of Buckquoy, Birsay, on Mainland Orkney (NGR HY 243 282). It was described and illustrated in her excavation report [Ritchie 1977:181-2,197 fig.8 no.84, 199 pi. 13a, item 84 in the catalogue offinds]with a detailed discussion of the inscription by Kenneth Jackson [Jackson 1977] and of the geology of the stone by G. H. Collins [Collins 1977]. The inscription has been further discussed in two unpublished MLitt theses [Padel 1972:73-5; Holder 1990:66-70] and by Dr Ritchie herself [1983:62,65]. Along with all the other finds from the Buckquoy excavation, the ogham inscribed whorl is now in the Tankerness House Museum in Kirkwall, Orkney (ref.no. 1976.56).
THE SITE The threatened site, known locally as Sinclair's Brae, was a long, low mound (20m long, maximum height 0.5m), truncated at one end by coastal erosion [Ritchie 1977:174]. Ritchie estimated that, by 1970, at least half the original site had been lost over the encroaching cliff [ibid. 174]. Ten weeks of digging revealed a series of farmsteads, dating from the seventh century to the tenth, built and re-built one on top of the other [175]. The Buckquoy excavation proved seminal, for it provided the first identified example of a distinctively Pictish house-type, of cellular form [182-3; see fig.], which subsequent discoveries have shown to have been widespread [Morris 1991:72], In the absence of other clear indicators, the remains were interpreted on the basis of the contrasting and distinctive house-types. Ritchie identified two major phases (MI) of Pictish occupation, followed, after a period of disuse, by three phases (III-V) of Viking
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
161
settlement. The Viking occupation appears to have been short-lived because the site was ruinous by the time it was used for an isolated burial, datable on coin evidence to the third quarter of the tenth century [1977:190-1].
The inscribed spindle-whorl was discovered immediately outside the south-west entrance to the main hall of house 4 [ibid. 181]. This structure, which has been described as having a 'distinctly anthropomorphic air' [Ritchie 1983:56], is perhaps the most sophisticated example yet uncovered of the Pictish 'figure-ofeight' house type. It differs from the older cellular type in having its cells arranged in a linear plan, not set in a circle round the central area of the house. Not enough is known about Pictish domestic structures to date this type closely [Ritchie 1977:182-3]. A similar, though simpler structure at Coileagan an Udal, Nonh Uist, was dated by its excavator between the seventh and ninth centuries [I. Crawford 1974:9]. Figure-of-eight-shaped houses made of wattle have been excavated more recently at Deer Park Farms, Glenarm, Co. Antrim [Lynn 1989], which calls into question the assumption that this type of dwelling is necessarily diagnostic of Pictish inhabitants. Most of the artefacts found in the pre-Norse layers at Buckquoy were of simple types common in various parts of north-west Europe in this period, though the painted pebble was recognized as being diagnostically Pictish [Ritchie 1972]. The ogham was also listed as a characteristically 'Pictish' item, without reference to the emphatically Irish background to the script [McManus 1991]. The possible Irish connotations of the Buckquoy inscription are discussed more fully below.
The settlement at Buckquoy was not itself of high status [Ritchie 1983:54] but its position at the nonh side of the strategically imponant Birsay Bay, very close to the tidal islet of the Brough, give it an importance beyond its size. The name derives from the ON bygg-kvt 'bere quoy', i.e. barley enclosure, and is a reminder that, traditionally, the soils of Birsay were considered the most fertile of mainland Orkney [Ritchie 1977:174]. The animal bones recovered in the excavation indicate a mixed farm with an emphasis on the pastoral, especially cattle [ibid. 191]. Ritchie has put forward the persuasive idea that Buckquoy
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUO Y
162
was 'the home farm for the inhabitants of the Brough' [1985:198]. The Bay of Birsay is one of only three sheltered bays on the west coast of the mainland of Orkney and is ideal for fishing [Ritchie 1977:174]. More importantly, perhaps, it is, as Ritchie has pointed out, particularly well-suited for boats setting out to the Hebrides or Ireland [1983:47].
Dating The discovery of an ogham inscription in the course of an archaeological excavation conducted to modern scientific standards raised the exciting possibility of the first absolute dating of an ogham inscription. Sadly, however, very little dating evidence had survived for any, bar the final, stage of activity at Buckquoy, and in the official report it was the ogham which was used to date the context, not vice-versa [Ritchie 1977:192].
The inscribed whorl was recovered from the second of the Pictish phases of
occupation [181]. Unfortunately there was nothing to permit the close dating of either phase. Only twenty-seven artefacts and some pottery sherds were found in the Pictish levels. These artefacts were, in the main, simple domestic items such as bone pins, a comb, and a spoon, none of which are closely datable [179].
The lack, until recently, of properly excavated ogham inscriptions has made it impossible to construct an accurately datable typology. On the basis of certain assumptions about the relationship between the forms of the script used in Ireland and in Scotland, and on the basis of the an historical dating of the slabs on which some inscriptions appear, it has been customary to assign most so-called 'Pictish' oghams to the eighth or ninth centuries, with simpler ones perhaps slightly earlier [Jackson 1955:139]. Since the script of Buckquoy is not typologically the most simple, Jackson's dating of the inscription to the eighth century was in keeping with general opinion at the time. Ritchie used his date to assign phase II to the eighthcentury and phase I to the seventh. While the other finds are not incompatible with this dating, they could be older. In the interim since Jackson gave his opinion, however, radio-carbon analysis of the contexts
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
163
of ogham stones excavated at Pool, Sanday and the Brough of Birsay has stretched the chronology of Orkney ogham back to the sixth- and seventh-centuries, and thereby thrown open the question of dating all the Scottish oghams, including the whorl [Pool - Hunter 1990:185; Birsay - Prof. Christopher Morris pers. com.]. It would be rash, however, to place too much emphasis on these new radiocarbon dates in advance of final publication of the excavations concerned.
A terminus ante quern for the abandonment of the farmstead, and therefore the deposition of die ogham, is provided by the conventional date of the beginning of Norse settlement in Orkney, c. 800. It has been assumed that sites in and around Birsay would have been settled early on, since Birsay was an important centre in the Pictish period [Ritchie 1983:47]. If one accepts this, and allows 'a brief interval, perhaps half a century* [Ritchie 1985:194] for the ruins to develop, die conclusion is that the buildings fell out of use in the mid-eighth century. In recent years, however, authors have come to admit that there is 'as yet no firm local, historical or archaeological evidence* for the date of the landnám, and that c.800 'is not a comfortable assumption' [Bigelow 1992:10; see also Morris 1985:210-3], While raiding bases might indeed have been established at the beginning of the ninth century, permanent settlement may not have got underway for at least another generation [B. Crawford 1987:40-2], The ninth century Latin Life of Fintan paints a picture of an Orkney still Pictish in the 840s [Thomson 1986, Lówe 1986]. There is no independent dating for the Viking phases at Buckquoy, instead these are retrospectively dated from die burial (phase VI) in the ruins of the last farm, which can be securely dated to the mid- to late tenth-century [Ritchie 1977:190-1]. Thus, while Ritchie's relative chronology for the site is inherently plausible, both upper and lower limits for the Pictish occupation are fluid. Jackson's dating of the ogham may well be correct, but cannot be relied on without independent corroboration. The form of the script used is compatible with any date from the seventh to the eleventh. We fall back, then, on a chronology based on the interpretation of the history of die site. This can provide a date no more refined than 'seventh-, eighth, or early nindi-century'.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
164
DESCRIPTION OF OBJECT A chalk spindle-whorl incised on one face with ogham arranged round a circular stem [see fig.]. Material:
'A fine-grained cream coloured sandy limestone, with quartz grains up to 0.5mm in diameter' [Collins 1977:222]
Dimensions:
36mm diameter, 10mm thick
Condition:
Intact and in good condition, inscription well-preserved.
A spindle-whorl is a heavy ring, in this case a perforated stone disc, attached to the pin of the spindle to give it the necessary weight and momentum for spinning thread by hand. A total of six spindle-whorls were recoveredfromBuckquoy, all of very similar size and shape, though madefroma variety of materials [Ritchie 1977 Nos. 82-86]. None of the others are marked or decorated in any way. The two examples from phase II (Nos. 83 and 84) were both made of chalk 'presumably chosen for its creamy colour and tractable nature' [ibid. 181]. Geological analysis of the three limestone whorls showed them to be 'made from closely similar rock type, varying only in detail' [Collins 1977:222]. Collins concluded that they 'could well have been madefromchalk pebbles, obtained from local glacial deposits' [ibid. 223]. Several other limestone whorls are known from Orkney, including one from the Broch of Lingrow [NMS GE11; MacGregor 1974:92] which is made from very similar chalk to that of No. 84 [Ritchie 1977:181]. The likelihood that the whorl was made from locally obtained stone is important because it implies the ogham was carved in Orkney rather than that the whorl was, say, imported from Ireland ready-carved. Ritchie was surprised that so few spindle-whorls were recovered from Buckquoy and concluded that any sheep at the site were kept more for their meat and hides, a view supported by the evidence of sheep bones recovered. Nonetheless, the presence of the ogham inscription on No.84 would seem, she feels, 'to imply that the whorl was special to its owner' [ibid. 182].
CARVING TECHNIQUE The ogham inscription was lightly, though clearly, incised with a fine sharp blade. Such a text would be easily and quickly produced on as workable a medium as chalk and can have taken no more than a few
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
165
moments to carve. The inscription is complete and well preserved. Jackson remarked that t(a)lmost every stroke is quite clearly made out when the whorl is held to a strong light at an angle and examined with a magnifying glass* [Jackson 1977:221], The carving is most readily seen on the photograph of the whorl taken under magnification by the Institute of Geological Sciences [see fig. (first published in Ritchie's excavation report, 1977 pi. 13); for further written descriptions see Jackson 1977 and Padel 1972].
A close examination of the published photograph reveals a zig-zag line traced very lightly across the surface between the perforation and the stem. It underlies the first two strokes of 12 and re-crosses under the stem just before letter 3 [see fig. for key to numbering]. This feature appears not to have been noted before, but evidently is prior to the ogham. The strokes are less substantial than those of the ogham and there is no confusion between the two. The sweep of the stem-line away from the perforation may have been to avoid this zig-zag, but in any case the stem is fairly haphazardly placed.
The stem-line, which is c.120 mm long, was incised only once and not re-cut on top of the subsequent letter-strokes. This is most clearly seen in the case of letters 5 and 11 (A and M). The individual letter strokes are c. 3-4mm in length, with 5mm for the hammer-head A and 12mm for the M, and were carved with one knife-score each. There appears to have been no re-cutting or augmentation of strokes (though see discussion of 7 and 72).
The spacing of letter 9 (N) requires some explanation. The second and third strokes are full-length, parallel, and spaced and sloped as one might expect. The first, however, is at an eccentric angle, meeting the second in a point at the stem. Even more strangely the fourth and fifth are squeezed into the inadequate space before 10, cramped to such an extent that the final stroke intersects with the following stroke well below the stem. This indicates that 10, or at least its first stroke, must have been carved before 9, or at least its last two strokes. This is puzzling since it would imply that the letters were not carved in the sequence in which they were to be read. If the carver was copying a circular model he or she might start
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
g UCKQUO Y
166
at any point, especially if he or she was not fluent in ogham, but surely it would have been apparent by the last stroke of 8 that space was running out and all the letters of 9 would need to be tightly spaced. That this was not done implies an alternative explanation.
The parallel second and third strokes of 9 appear firmer then the other three. They are accurately sloped and, if one ignores the first, fbunh, and fifth strokes, are nicely spaced relative to the letters on either flank. Could it be that, having completed the text, the carver realized he or she had carved two strokes where five were needed and squeezed in the first, fourth, and fifth strokes as best they could ? Perhaps closer examination of the depth of individual strokes, and therefore the pressure with which they were cut, might be instructive. Miscounting strokes in this way is the kind of slip that even an experienced oghamist might make and need not imply that the carver was following an exemplar he or she did not understand.
Several features combine to give the Buckquoy ogham a cursive appearance. These are, principally; the stem is not equidistant from the perforation all way round, the parallelism of component letter-strokes is maintained only erratically, there is disparity in the length of strokes both within and between groups, the letters are unevenly spaced (note especially the cramping of 9 and 1211). These are the kind of features often attributed to ignorant copying by an illiterate carver, yet the perceived shortcomings of the Buckquoy ogham are not blunders due to miscomprehension. Since the legibility of ogham depends entirely on layout and spacing, one can't help feeling that someone not fully comfortable with the script would have taken more care to lay it out properly. While the use of examplars is highly likely in grand monumental ogham epigraphy, such as at Branasbutt, Whiteness, or Dupplin, the texts of which are carved with special, labour-intensive carving techniques, it seems scarcely appropriate for a casual and cursive text like Buckquoy.
No special equipment or carving ability would have been required to scratch such an
inscription, anyone who could write could have carved it. While it can never be proven, I feel the features mentioned above are most naturally explained in terms of a text carved straight onto the whorl by the person who composed it.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
167
DESCRIPTION OF OGHAM INSCRIPTION Provided the curve is in keeping with the scale of the lettering there is no particular difficulty in reading a circular ogham, that is, as long as one knows where to start and in which direction to go. Indeed, a circular stem seems a very natural arrangement of a Type II ogham (i.e. one with draw-in stem-line) on a disc: much more obvious than, say, the peculiar configuration on the Ennis bead (CIIC 53). Since ogham always reads from left to right (whether horizontally or vertically), the point at issue is: does one stand, as it were, at the centre and look out, thus reading clockwise, or stand on the perimeter looking in, and read anti-clockwise. If one were keen to preserve the convention of reading vertically up the left, then one would adopt the former stance. This appears to be the direction of the Logie Elphinstone ogham, as indicated by the slope of the letter-strokes. Both Jackson and Padel preferred to read the Buckquoy ogham in this direction too, though the slope of the Buckquoy letters unequivocally indicates that is should be read owz-clockwise. Padel attempted to establish a pattern for the direction of reading circular oghams [1972:13-15], but, since the wheel oghams in the Book of Ballymote are quite different in character, the only true parallel is the circular ogham from Logie Elphinstone, which reads clockwise. If anything, the Buckquoy ogham should set the precedent for the much more inscrutable Logie ogham, but there is no need to set the authority of one against the other, since the letters of each slope in opposite directions and leave no room for doubt.
Jackson recognized that the M (11) was, as would be expected, diagonal to the stem-line, but, because he was convinced that 'die other strokes are evidently all intended to be at right-angles to it', he was not sufficiently sensitive to the pitch of the letters ~ '(m)ost are in fact more or less atright-anglesto die point where they reach the line'. The few which he conceded were 'more sloping' could be explained, in his opinion, by the distorting effect of a circular base-line [Jackson 1977:221], A close examination, however, reveals that all of the b- and h-aicme consonants are oblique. Compare, for instance, letters 4 and 5 (DA). There can be no doubt that the slope is deliberate - across any individual letter-group die stem is so short as to be, in effect, straight - but this sloping is exactly what one would expect.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
168
The sloping of b- and h-aicme consonants is a widespread feature of Type II inscriptions in Scotland. Although the gradient may vary, the incline is always in the same direction relative to the direction of reading, i.e. with the proximal end further 'forward' than the distal. Placed next to one another a correctly orientated B and H will form an arrow-head which points in the direction of reading. The pitch of maicme consonants, which slope left-to-right (up) across the stem, is not diagnostic since their outline is maintained even if the inscription is inverted, and read back-to-front. B- and h-aicme consonants, however, would, in such circumstances, point the wrong way. Even if the slope of consonants were not sufficient to indicate the direction of reading the Buckquoy text, further, unequivocal evidence is provided by die hammer-head A. Tlúsforfid appears on five other Scottish ogham inscriptions always with the cross-stroke on the right/lower distal end. There can be no doubt that Buckquoy is to be read anti-clockwise with the b-surface nearest the perimeter.
In Irish dessel, ? sun-wise\ 'right-hand-wise', means by extension 'lucky' 'favourable', 'propitious' [DH s.v.], and a similar semantic association is reflected in other Celtic languages. Given this well-attested preference for motion in a sunwise direction it might at first glance appear surprising that the Buckquoy text reads anti-clockwise, but it is all a question of perspective: though die text runs anti-clockwise, to read it one must turn die whorl clockwise. Thus as die spinner spins her yarn, and die spindle turns sunwise, the text passes legibly, if at speed, before her eyes. The presence of die pin would in any case make it exceedingly awkward to carve and read the text if it ran in the opposite direction, if, that is, one assumes the inscription was carved while the whorl was still in use for spinning.
The only remaining question is where to start. Circular texts in any script are not a common feature of Insular epigraphy. Where they occur their layout is usually a result of die shape of die objects on which they are carved, for instance, rings, coins, seals and stone fonts. The most common device for indicating the starting point of a text is the small square cross (e.g. Aldborough sun-dial [Okasha 1971:47]). Occasionally spacing is used to indicate word-division (e.g. die font from Partrishow, Cardiganshire (CIIC
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
169
988)), but quite often the text is continuous and the inscriber relies on the skill of the reader in discerning where to start (e.g. the Attleborough ring [Okasha 1971:49-50]). Since circular texts are usually short, this is rarely a problem.
There are marked differences in the spacing between the various letters of the Buckquoy ogham, and it might seem logical to start after one of the larger gaps, 516 and 112, i.e. at C or N. In both case, however, the resulting text is unintelligible, and it seems more likely that the variation is merely the result of the casual way in which the text was carved.
In most cases it is obvious where an ogham texts begins, but in some Type II Irish oghams, whether manuscript (St. Gall Priscian) or epigraphic (e.g. Ballyspellan brooch (CIIC 27), Killaloe slab (54), Tullycommon bone (52), Colman Bocht slab, Clonmacnoise (749)), a non-phonetic character is used to indicate the starting point and direction of reading. Both Padel and Jackson interpreted Buckquoy's letter / ( * ) as an instance of this 'feather-symbol' or 'feather-mark'.
Yet a preliminary survey of Irish
epigraphic and manuscript oghams showed that in every case the 'feather-mark', perhaps more aptly labelled 'arrow-mark', is shaped > — or -^—, never x — or -*— [Brash 1879:322-3]. There are two Scottish examples of inscriptions beginning with x-(Lunnasting and Cunningsburgh 3), but in both cases it seems preferable to take the character as the first foifid and accord it its usual phonetic value (Id). There are no Scottish examples of a > — or - > — shaped directional indicator. It appears, therefore, that 7 is to be taken as /e/. This interpretation is based on its unambiguous shape, but has the added advantage of removing an otherwise unfeasible cluster of consonants.
The stem of the Buckquoy ogham is not a seamless ring, like that of Logie Elphinstone, but rather a looped line with two distinct ends. These overlap slightly, just above 1, and it would seem natural to take this break in the stem as the start of the text. This leaves the problem of the relationship of 1 to the rest of the text, because it is out of alignment with the other characters, and lies outside (below) rather than on
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
170
die stem. Strictly speaking, the first letter after the break in the stem is 2, but a close examination shows that 1 is in fact attached to the stem. The left-hand end of the stem curves down slightly and meets the upper left arm of the forfid just above its intersection. It is as if 1 was indeed sitting on the tip of the stem, but that the whole thing had been bent down 90°. There is ample space for 1 to sit on the stem between the break and 2 and no obvious reason why it should be suspended from rather than intersected by the stem.
The precise arrangement of strokes in this portion of the text is not easy to rationalize, however one chooses to interpret it. It is clearly confused to some degree - the final stroke of 12 is hard-up against the x-forfid and partially, but not unequivocally, differentiated from the first two strokes of 12, Why the large gap before 2 ? Why the intersection of the two ends of the stem and the placing of 7 on a different alignment ? Dr Oliver Padel suggests to me that the curling-out of the stem is a deliberate device to show the stan of the inscription [pers.com.]. While this is without parallel, certainly it is a possibility.
The palaeographical conventions of ogham have to be deduced solely from observation of the extant inscriptions. Since there are relatively few post-seventh century examples and these are heterogeneous, it is sometimes difficult, in analyzing variations in individual inscriptions, to differentiate the deliberate and telling from the incidental and insignificant. This is particularly true of informal inscriptions such as Buckquoy, where the difficulty is further compounded by the virtually unique circular arrangement of the text. Generally, unambiguous mistakes are rare, but it would be foolish, either to see errors everywhere, or to deny that they ever happen - a corrected mistake is the obvious explanation of the configuration of Buckquoy 9. If one were prepared to accept the possibility of another mistake, or at least a change of plan, a possible explanation is suggested by the observation that the two ends of the stem overlap to form a little cross. Perhaps, having quickly cut the stem, the ogham-carver was struck by this x-shaped intersection and decided to press it into service as an x-forfid, but, having cut die rest of the letters, concluded that
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
171
this device was not sufficiently unambiguous and added a separate x (/), putting it to the side to avoid it being interpreted as a second E.
Balancing the demands of an unforced interpretation of the carving against plausible readings of the text is the epigrapher's perennial problem. We might feel intuitively that the beginning and end of the text should coincide with the break in the stem, but given the obvious casualness with which the inscription was carved, the poor spacing of the letters and the fact that at least one mistake was made and corrected, how much confidence can be placed in such a hunch ? If an intelligible reading is provided by starting with the stroke before / to what extent should this be allowed to influence our interpretation? For the sake of convention only and without prejudice to the final reading, I take the x-forfid as the first letter and read anti-clockwise, as follows [see fig. for key to numbering]: I
A small cross below the line, consisting of two short lines intersecting more or less at their midpoint, the angle between the upper two arms being more than 90°, joined to the tip of the stem as described above. This character, the x-forfid, appears on a total of seven Scottish oghams including Buckquoy, though in initial position only on Lunnasting and Cunningsburgh 3. In both cases the letter is followed by TTE, and, even though neither text has been satisfactorily interpreted, given the apparent avoidance of double letters in word-initial position, it seems preferable to take the symbol as having its normal phonetic (vocalic) value /e/. Above theforfid, the stem curves round sharply, crosses its other end, then commences its circuit. There is a generous gap before the next letter. Jackson described it as sufficient for five strokes, although I doubt it could have taken more than two, properly spaced. Given the comfortable spacing of the next few letters, it need not be considered remarkable and is certainly no bigger than the gap after 5.
2
Five strokes to therightof the stem, sloping forward - N. The first stroke is somewhat doubtful, being faint and at a slightly greater angle than the rest (making due allowance for the curve of the stem).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
3
BUCKQUOY
172
Two strokes to the left of the stem, sloping very slightly forward - D. These two are roughly parallel for most of their length but converge at their distal ends. According to Jackson, they do not join. The first stroke appears to over-shoot the stem very slightly. There is a line, like a bind-stroke, which continues well beyond the end of the first stroke. Jackson dismisses it as 'a fortuitous scratch', but it is possible that it is indeed a bind-stroke, similar to one found on the Burrian stone which has a single curved line do duty for the bind-stroke and the final letter-stroke of the group.
4
As above - D. These two are slightly closer together than the previous pair, and at a sharper angle to the stem.
5
A single long stroke almost perpendicular across the stem with a horizontal line across its lower distal end - a hammer-head A. This letter is clear and definite. Padel thought the cross-stroke unusually long. There follows a generous gap, the largest in the inscription.
6
Four strokes to the left of the stem, sloping forward with constant gradient - C. Jackson describes the second stroke as crossing the stem slightly.
7
Three strokes to the left of the stem, sloping forward with increasing gradient - T. Padel saw a very short possible score just after these but decided it was probably not part of the letter. It should be noted that there are a number of nicks and scores in the surface of the stone which are clearly just casual damage.
This first half of the text takes up two thirds of the stem, thus, although the first seven characters are very comfortably spaced, the next six are rather cramped. 8
A single stroke across the stem - A. This stroke is of very similar length to those of 10 and substantially shorter that 77, so there is no doubt that it is a vowel. It occupies the middle twothirds of the ogham band and appears to slope very slightly backwards.
9
Five strokes to the right of the stem, sloping forward at markedly different angles - N. The second and third strokes almost meet at the stem, and the final stroke intersects the first stroke of 10 and thus fails to reach the stem. The form of this letter has been discussed in detail above,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
173
there is no justification in taking it as 2+3. 10
Five strokes fairly perpendicular across the stem - 1 . Though the individual strokes are a little straggly, they are all more or less parallel, of similar length, and spaced as widely as 2, 6, and 7.
11
A single long stroke obliquely across the stem, slopingforwards- M. This is the longest stroke of all, about double the length of the preceding vowel-strokes, and very firmly sloped. There follows a generous gap, perhaps left to avoid the end of the underlying zig-zag.
12
Three strokes to the right of the stem, sloping forwards. The first two are exactly parallel and very tightly spaced, the third follows at a slightly more gentle gradient after a larger gap. There is some doubt over whether these should be taken as a single group of three strokes, in which case - V, or as two separate letters, 2+1 - LB. Elsewhere in the text the component strokes of individual letters are generally evenly spaced, but in contrast there is considerable variation in the size of the gaps between letters. The interval between the second and third strokes of 12 is comparable to that between 415, 718, and 8/9. There is no gap at all between 9110. The ogham text does not appear to have been carefully laid out, if the carver was running out of space this may have been the maximum spacing possible. There is certainly a larger gap between the second and third strokes than between the third stroke and the intersection. Furthermore, the second stroke is far longer than the other two. This may be nothing more than a slip of the blade, but, if intentional, might be intended to differentiate two closely packed though separate letters of the same aicme. On balance I would prefer to take 12 as two separate letters [cf. Holder 1990], though Jackson and Padel took it as one.
Which gives the following reading: ENDDACTANIM(v/lb)
In essentials this is in agreement with Holder's more tentative reading - E(s/n)DDACTA(n/lv)IM(v/lb) - and the possible variant passed over by Jackson and Padel - (e/ )(s/n/ )DDACTANIMV - dieir preferred reading was (e/ )TMIQAVSALL(c/q).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
174
FORM OF SCRIPT The Buckquoy whorl is inscribed in a simple Type II ogham script with the incorporation of the two forfeda most common on Scottish oghams - x (Id) and the hammer-head A J_ • The hammer-head A occurs on six other Scottish ogham stones: Birsay 1, Burrian, Lunnasting, Latheron, Golspie, and Formaston. The x-forfid, which isfoundon numerous Irish pillars inscribed with Type I ogham, is used on Burrian (three times), Newton (once), Golspie (twice), Cunningsburgh 3, and Lunnasting (once each at the beginning before TT), and at Formaston (twice). All the consonants of the Buckquoy text are sloped. The vowels, with the exception of the wojorfeda, consist of perpendicular strokes across the stem, taking up, perhaps the middle two-thirds of the ogham band, though the variation in length of individual strokes makes it difficult to be accurate on this last point. Within letters, the strokes are fairly parallel and evenly spaced. Substantial gaps are left between all letters, not just successive ones of die same aicme. If the suggested reading is correct, there is no attempt at word-division.
In general terms the script is very similar to that on the cross-slab at Latheron, which has the hammer-head A and consonants which slope, thought not as consistently as Buckquoy. The other ogham-inscribed crossslab from Sutherland, Golspie, is also similar in script to Buckquoy, but has angled vowels, and letters which are not so generously spaced. The Buckquoy script is simpler than those used on the three Birsay stones. Birsay 1 has the hammer-head A and sloping consonants, but, like Golspie, has angled vowels. Birsay 3 is written with bound letters. The Burrian ogham provides an example of the final stroke of a letter group being a continuation of the bind stroke, and it is possible that Buckquoy *s 3 is an attempt at such a letter, but equally it may be a mere slip of the blade.
The only exceptional aspect of the Buckquoy ogham is that it is carved on a circular stem-line. This, however, does not affect the aspects of the script mentioned above. Changing the shape of the stem is one of the ways to create a cryptic ogham mentioned in the Ogham tract in the Book of Ballymote [Macalister 1937:48), although, interestingly enough, a circular stem is not given as an example. While the only odier
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
B UCKQUOY
175
instance of a circular ogham, Logie Elphinstone, may indeed be cryptic, there is nothing enigmatic about Buckquoy. The use of a circular stem is an obvious solution to the practical problem of fitting the text to the object and is, in a sense, comparable to the 'up-top-down', or boustrophedon, arrangement on the oldest Irish pillars. There is no need to assume influence between Logie Elphinstone and Buckquoy, in either direction. The circular stem is a simple innovation which may have arisen independently in different areas. The fact that the two are read in opposite directions, and are very different in length and in context, is sufficient to suggest no direct relationship between the two.
INTERPRETATION OF TEXT For the reasons given above, eTMIQAVSALL(c/q), the favoured transliteration of Padel and Jackson, is based on an incorrect reading of the text and must be rejected. There is slight doubt over letter 12 (probably LB, but could possibly be interpreted as V), but the rest, ENDDACTANIM-, is secure. The only other doubt is over where to start. Though Jackson did in fact give ENDDACTANIMV as one of his eight alternative readings, he rejected it as 'wholly unintelligible and cannot be Celtic', so strong was his conviction about the nature of Pictish. In conclusion he stated himself to be 'content to write off this inscription as unintelligible, like all the other 'Pictish' inscriptions' [Jackson 1977:222]. It is simply not true to say that all Pictish inscriptions are unintelligible. Although a number do continue to resist interpretation, e.g. Inchyra and Logie Elphinstone, following Padel [1972] I hope to show that many, e.g. Latheron, Scoonie, and Ackergill, contain recognisable Celtic personal-names. I have argued against Jackson's hypothesis of a second, non-Indo-European Pictish language [Forsyth, forthcoming 1996c] and prefer to view the not inconsiderable difficulty of many Pictish inscriptions as an inevitable result of trying to interpret often damaged inscriptions in an otherwise only minimally-attested P-Celtic language. The proportion of doubtful or disputed Anglo-Saxon runic inscriptions, as given by Page, is a forceful reminder that epigraphic opacity is not an exclusively Pictish phenomenon [1973:14-51.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUO Y
176
If one is prepared to begin with the stroke before 7, then die reading BENDDACTANIML suggests itself. This I would segment BENDDACT ANIM L, which I take to be Old Irish bendact anim I , 'a blessing on die soul of L \ According to diis new interpretation, BENDDACT is Old Irish bendacht, 'a blessing', from Latin benedictum [DEL and Vendryes 1959 s.v. bennachn Thurneysen 1946:450 §727] and ANIM, Old Irish animy 'soul' [Vendryes 1959 s.v. anim; D1L s.v. ainim(m); Thurneysen 1946:214 §333]. As it stands anim could be nominative, accusative, or dative singular, as attested in die Old Irish Glosses, but in the current context is most likely to be dative or accusative (see discussion below). This word is not to be confused with Old Irish ainm 'name', which occurs, in its older form ANM, on a number of classic ogham pillars in southern Ireland [McManus 1991:118 §6.27], and continued to be used in a technical sense in die Irish phrase ainm n-oguim 'funerary inscription in ogham' [ibid. 154 §8.8].
The spelling anim is relatively straightforward, diough as accusative or dative it alternates in die Glosses widi die forms anmuinlanmain [Thurneysen loccit.).
BENDDACT for bendacht is perhaps more
interesting. The gemination of consonants is a much discussed feature of ogham orthography [McManus 1991:124-6 §6.30; Harvey 1987a], and is particularly common in Scottish ogham. To die extent that it can be determined, diere is an apparent tendency for non-initial D to be doubled in Scottish ogham texts (exceptions being Blackwaterfoot 2, Latheron, Lochgoilhead) though whether this had phonetic significance is unclear.
If it were thought diat die Buckquoy text perpetuated aspects of die traditional ogham
ordiography, DD could be taken as indicating (non-lenited) /d/, on analogy widi CC and TT representing (non-lenited) /k/ and /t/ [McManus 1991:125]. Similarly, single C for /x/ could be an example of die same geminate/non-lenited--simplex/lenited distinction. However, as Dr McManus points out to me, diere are numerous parallels for such spellings in Old Irish manuscript orthography, and it is equally possible that the Buckquoy ogham is merely a cipher for die manuscript spelling. In Old Irish die spelling ct 'not infrequently represented' cht [Thurneysen 1946:21 §28], and bendact is attested in die Glosses [Wurzburg 19bl5, see Stokes and Strachan 1901-3] and on one roman alphabet inscription (CIIC 868).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
177
If, for whatever reason, the reading L were rejected, it would be possible to take 12 (LB) as a single letter consisting of three strokes, V i.e. /w/ or /f/, and still read it as Irish - not BENDDACT ANIM L, but VENDDACT ANIM or, rather, ANIM VENDDACT fthe soul of Vendact' [I owe this suggestion to Dr Paul Russell, Radley College and Jesus College, Oxford]. This would require taking Vendact as an otherwise unattested compound of the popular personal name element vendu- 'white, bright' [McManus 1991:103 §6.4] with the suffix -oct- (EOIr. acht) which was commonly used to form personal names from nouns and adjectives [McManus 1991 §6.13], or an archaic form of Findach* The third letter of the baicme may be transcribed as V or F depending on the period [McManus 1991:122; Sims-Williams 1993:140-3], but the spelling ven- is problematic. One would have expected the first syllable to show vowel affection, a change which occurred in Irish by the sixth century, certainly before syncope and apocope [McManus 1991:87, 93]. Apart from problems of absolute chronology, in terms of the relative chronology of sound changes in early Irish, vowel affection occurred before apocope (loss of final syllables), yet Buckquoy is a post-apocope text (*Vendu-oaos). Thirdly, we would expect the name to appear in the genitive case, yet it displays no genitive ending (-achta). These linguistic objections render preferable the original interpretation Bendacht anim L.
DISCUSSION The formula bendacht for anim M, 'a blessing on the soul of N. \ is well attested in the Irish epigraphic record, mainly on recumbent cross-slabs (e.g. bendacht for anmain N. (CIIC 551, 933,935(; bennachtfor anmain N. (CIIC 916, 917(). Other related forms are benfdachjt Die for anfmainj N. (CIIC 529) and bendacht arN. (CIIC 586,958). Of particular interest is the examplefromLemanaghan, Co. Offaly (CIIC 868) bendactfor anfmaijn ailbenig - with c for /x/, as appears mutatis mutandis in ogham at Buckquoy. All of these are written in the roman alphabet, but there is a single, very late, example of the use of the formula in an ogham inscription -- the eleventh century, bi-lingual, runic-Norse/ogham-Irish slab from Killaloe (CIIC 54) which has bendacht [ar] M, with the first E spelled with the x-forfid: BxNDACHT.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
178
In all these examples anim is in the accusative, governed by the preposition ar or for, the latter occasionally abbreviated (as on CIIC 529, 916, and 917), to/[Thurneysen 1946:162 §251]. The omission of the preposition from the Buckquoy text therefore demands an explanation. Epigraphers are usually quick to identify grammatical and other 'errors' in the inscriptions they study [see Macalister's attempt to 'correct' Pictísh grammar, 1940]. The Buckquoy oghamist was certainly casual, it could be that he or she was careless or ignorant as well With funher study, however, it often turns out to be the epigrapher who is ignorant, and 'bad' Irish should be invoked in this case only once all other possibilities have been exhausted. In feet there is an alternative explanation.
In his discussion of Archaic Irish, Greene confirmed that 'the elimination of unneccessary particles is one of the outstanding features of the archaic style1, which he says 'accounts for the consistent use of the independent dative where Old Irish would require a preposition1 [1977:19]. Thurneysen makes brief mention of the prepositionless dative in his Grammar of Old Irish, characterizing this nominal construction, which is attested in certain Irish legal texts and paralleled in Gaulish inscriptions, as 'archaic and extremely rare' [Thurneysen 1946:162 §251]. If Buckquoy f s anim were in the dative case, then we might have here an example of the 'independent dative'. The extreme rarity of this construction in Old Irish warrants caution, as Dr McManus has reminded me, yet archaic syntax would not be out of keeping with a seventh or eighth-century date for the Buckquoy whorl. However, as relics of an earlier period of the language, 'archaic' features continued to be used in Old Irish as a stylistic device [Breatnach 1984:458-9]. This need not, of course, rule out an early date for our text, especially since a heightened literary register is unlikely in a casual epigraphic context such as this. Nonetheless, an independent dative, if indeed that is what it is, need not necessarily indicate an early date, and thus the text cannot be dated on the basis of syntax alone.
A possible objection to the reading BENDDACT ANIM L is that the use of an initial letter, in this case L, to represent a personal name is unprecedented in a Goidelic context at this period. In the ogham
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
179
tradition, as in the runic, each character was identified by a meaningful name on the acrostic principal. The letter-name for ogham L was luis, though there is some doubt whether this is luiselloise 'flame, blaze* or his 'plant, herb* [McManus 1991:36]. Later glossators identify luis as a tree, either the rowan or elm [McManus 1988:150]. Rune-writers exploited the letter-names to great effect in writing riddles in which the runic character was made to stand for its name [Page 1973:203-11], but there do not appear to be any examples of such a conceit in the early ogham corpus. Luis is not attested as a personal name, so 'L' would have to be interpreted as a straightforward abbreviation of some other personal name.
The concept of abbreviation of divine names and epithets was well known from the use of the nomina sacra but does not seem to have been extended to profane personal names. To establish this beyond doubt, however, is not easy since there are few close comparisons. Irish personal names are too varied for a comprehensive system of abbreviations such as was used in Roman epigraphy. For this reason the contexts in which it would be possible or appropriate to abbreviate a Goidelic personal name are limited. Unless there is some cryptic intent, a personal name is likely to be abbreviated only if the identity of the individual is obvious and widely known. Perhaps the most usual circumstance is the marking of ownership on a personal belonging. Most of the extant, non-monumental, insular inscriptions are on objects of great value, usually highly-accomplished pieces of secular or ecclesiastical metalwork such as brooches, weapons, and shrines. One would not expect the names of the owners, patrons, makers or honorands to be abbreviated on such formal, in some cases public, items inscribed with posterity in mind. It is instead on instmmenta domestica, transient and of little economic value, that one might expect to find the initials of an owner or maker. But of course, such objects are rarely preserved in the archaeological record. Though there are no clear examples from Celtic-speaking areas, the corpus of Anglo-Saxon runic inscriptions includes a handful of examples of objects inscribed with a single letter, or a pair of letters, most readily interpreted as owner's marks (the wooden York spoon, the Sleaford brooch, the Willoughby-on-the-Wolds bowl, mentioned by Page [1973:172]).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
180
Reasons why the alleged personal name was abbreviated in die Buckquoy text when the rest was written out might be the obvious lack of space (if die individual was well known to all the inhabitants of die farmstead, to give it in full might have been considered superfluous), or perhaps die Irish-speaker who carved it was unsure how to commit an unfamiliar native name to writing. Of course, an abbreviated message inscribed on a gift provides a shared understanding between giver and recipient thereby enhancing die intimacy of die bond between diem, and this may be sufficient explanation of itself. An appropriate analogy might by a locket engraved widi a sweetheart's intials.
On die semantic level, one might object diat 'a blessing on die soul of N.' is an essentially commemorative phrase and as such scarcely appropriate for a spindle-whorl. Certainly all the examples of bendacht or If or cited thus far occur on funerary monuments. The formula, however, was not restricted to gravestones: Macalister read bendacht ar Artgal on die Cross of Kells [CIIC 586]; in metalwork, the mid-eleventh century Stowe Missal Shrine [CIIC 932J has bendacht De ar cech anmain as a hairilliuth 'God's blessing on every soul according to its deserts'. These are commemorative or dedicatory without being funerary, as are die occurances of die formula in die scribal colophons of manuscripts. An example of die latter widi Scottish provenance is die colophon to die Gospel extracts in die Book of Deer (Cambridge UL MS.I.i.6.32). Written in die same hand as die main text it is to be dated to die ninth century. It asks of die reader a bendacht fforj anmain in truagain ro-d scribai 'his blessing on die soul of die poor wretch who has written it' [Jackson 1972:8; Stuart 1869:lx, 89].
The examples are all public texts on high status items written widi posterity in mind and thus, still radier far from the informal domestic object from Buckquoy.
Perhaps a closer parallel is provided by die
inscribed pebble found in 1822 in a grave at Temple Brecan, Na Seacht dTeampaill, Inis Mór, Aran Islands (CIIC 532) [Higgins 1987.1:23, 26-7, 137-9; 1987.2:268-9, fig.l, plate 54]. The black limestone pebble, 75mm in diameter, was lost for many years and known only from drawings. It was dien widely interpreted as a stone lamp, and thus appeared to provide a cognate example of a domestic implement inscribed widi
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
181
a dedicatory text. A few years ago, however, it was re-discovered in the National Museum of Ireland, enabling the most recent writer to rule out this explanation [Higgins 1987.1:26]. Instead it belongs to a class of 'decorated pebbles' of uncertain, probably religious, use. The circular inscription in Insular minuscule reads +or(oit) arbran n-ailither, 'a prayer for Bran the pilgrim' [Macalister 1949:6,202-3 ill.]. This formula, Or(oit) ar/do, is by far the most common of Irish roman alphabet inscriptions and is found widely, though not exclusively, on funerary monuments. According to Higgins the Temple Brecan stone 'has all the signs of wear from use and handling to suggest that it had a previous function' [1987.1:23], yet the fact that it was discovered in a grave clouds the issue somewhat; it is unclear whether the inscription had been present while the item was in use, or was carved expressly for deposition in the grave. Whatever the precise interpretation, the existence of another informal, portable, personal belonging inscribed with a formula that is more familiar from grave slabs makes the Buckquoy whorl seem less exceptional.
Parallels A number of inscribed spindle-whorls have been recovered from Insular and Continental excavations. Such an intimate and personal item might be considered particularly suitable for inscribing, though the surprising frequency with which inscribed whorls are recovered may be due, perhaps, to their relative durability, coupled with how easy they were to lose, rather than because they were particularly favoured for inscriptions. Historically and cross-culturally, spinning is an activity most closely associated with women, so opening up the possibility of interest, or even skill, in literacy on the pan of women. The spindle-whorl texts vary widely in length, complexity, and meaning. Most simple are those which carry merely a personal name, presumably that of the owner. An example is the seventh or eighth century Anglo-Saxon jet spindle-whorl from Whitby Abbey inscribed with three runes, Wer [Peers & Radford 1943:74], which has been interpreted as a (male) personal name [Page 1973:171*2]. Next is the writer/maker formula for instance, another nine inscribed spindle-whorl, this time from the Northern Isles (exact provenance unknown), which reads 'Gautr wrote the runes' [NMS BE 360; Liest0l 1984:232], Personal belongings,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
182
especially tools and weapons, are sometimes inscribed with talismanic or protective inscriptions, whether pagan or Christian. Scottish examples would be runic ruthorks, such as the seal's tooth from Birsay [Liest0l 1984:232; Curie 1982:59-60 III 37], or the St Ninian's Isle chape [Small et al. 1973]. The Buckquoy whorl may tall into this category, as might the pseudo-ogham-inscribed whorl from Burrian [MacGregor 1974:91 fig. 18; see Appendix - Dubia]. A fourth category, though attested much earlier, shows that such inscriptions may have a ludic rather than strictly practical or talismanic function. These are the corpus of Gaulish-, Gallo-Latin-, and Latin-inscribed whorls from eastern Gaul [Whatmough 1949, Lambert 1994:122-5]. Meid has interpreted these texts as subtly, or more directly, erotic messages, from implied male speakers, addressed to young women. As such he characterizes them as reflecting 'spinning room amusements' [Meid 1992:52-54, ill.]. According to Lambert such objects, bearing 'des souhaits ou des compliments amoureaux\ were given 'en cadeau par des galants dont les intentions sont claires* [1994:23]. The fact that whorls were made by men and given to women, especially sweethearts, in other cultures too, is indicated by a Norse example in Rekjavik Museum, Thorá migfrá Hruna Thora owns me, from Hruni' [referred to on NMS label for spindle-whorl BE 360]. The Christian sentiment of the Buckquoy whorl is a long way from the double-entendres of the Gaulish carvings but it is not inconceivable that 'L' was a man (an Irishman ?) who inscribed the Orkney spinster's whorl in the hope of future remembrance. Page gives as an alternative interpretation of the wer of the Whitby spindle-whorl the northern form of West Saxon wcer 'token of friendship' [1973:171].
All of these comparanda are examples of inscriptions carved while the whorl was still in use for spinning. In more recent times, however, former spindle-whorls were used in the Scottish Gáidhlteachd as charms [pers.comm. Dr John Maclnnes of the School of Scottish Studies]. It is possible that die Buckquoy text relates to a time when the whorl had ceased to be perform its primary function and had taken on talismanic significance. There are no explicit early medieval references to whorl-charms, but Adomnán provides seventh century evidence of the use of pebbles as amulets. In VC he refers first to a small, unspecified benedictio, ('object that has been blessed'), which was housed in an inscribed box, and, once dipped in
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
183
water, was used to cure a broken hip [VC ii.5 p. 103 n.134]; then to a piece of rock salt blessed by the saint which was hung on the wall above the bed of an invalid [ii.7 p. 105]. Most relevant to Buckquoy is the lapis candidusy 'white stone', which was taken from a Scottish river, blessed and used to work cures among the heathen Picts. Adomnán has Columba say, Signate ... hunc candidum lapidem, per quern domimts in hoc gentili populo mulms egrotonim perficiet sanitates, 'Mark this white stone. Through it the Lord will work many cures of the sick among this heathen people' [VC ii.33]. Though signate doubtless means mark with the sign of the cross, is not the Buckquoy spindle-whorl a white stone, marked with a Christian message ? Though there is considerable doubt about the interpretation of the ogham on the famous Ennis bead (CIIC 53), it should be noted that it was believed to be of assistance to women in labour, and efficacious in the treatment of eye complaints [Brash 1879:321; see Meaney 1981 for discussion of Anglo-Saxon curing stones].
Historical Context As a complete, legible but. according to the two authorities, unintelligible ogham, Buckquoy might be considered one of the stronger planks in the argument for a non-Indo-European Pictish language. If it could, however, be proved to be comprehensible, and the proportion of uninterpreted inscriptions thus reduced, the case for non-Indo-European Pictish would be weakened accordingly. Of course, if the ogham is Old Irish, this implies nothing about the nature of the Pictish language, merely that there were Irish speakers in Orkney in the eighth century. Thus, if accepted, the interpretation of the Buckquoy ogham text as a Christian phrase written in Old Irish is an important breakthrough, not only for the linguistic information it provides, but also because of what it indicates about the spread of Irish Christianity to the North.
When discussing external influences on pre-Viking Orkney, commentators have, quite rightly, tended to stress the eastern connections with mainland Pictland and Northumbria [see most recently, Lamb 1993]. The importance of the western seaboard as a conduit for influences from Ireland should not, however, be
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
184
overlooked. There is documentary evidence for voyages by Irish peregrini to Orkney and beyond as early as the sixth century.
Both Brendan and Cormac Ua Lfatháin, who, according to Adomnán, visited
Columba on Hinba, came from areas well stocked with ogham stones [VC i.6, ii.42, iii. 17], a number of which have explicitly Christian associations. Of those marked with the cross, two refer to ecclesiastics Arraglen [CIIC 145] QRIMITIR 'priest', Maumanorig [CIIC 193] AILITHIR, 'pilgrim' (though the reading is problematic). Continued familiarity with the ogham script in the Irish church is suggested by the presence of ogham marginalia in manuscripts of the ninth century and later written by Irish scribes, not least by the ogham signature in the Stowe Missal, RIA MS.D.II.3 f.l lr, of one Sonid who describes himself as 'peregrimts1 [Warner 1915: xlii and plate viii].
Peter Harbison has amassed archaeological and ait historical evidence for a long tradition of pilgrimage along a western maritime arc stretching from the far south-west of Ireland to Shetland [1991:192-4, 221]. A spindle-whorl would be an unlikely piece of baggage for a monk on pilgrimage, unless, that is, it was already functioning as a relic or charm. The geological evidence, however, indicates that the Buckquoy ogham was not imported from Ireland, but instead carved locally, the implication being that there were people at Birsay in the eighth-century who understood the Irish language, either Irish settlers or bi-lingual Picts. The Orkneys were regarded by contemporaries as Pictish territory [Dumville 1976], but it is not inconceivable that there might have been an Irish colony there, as in so many other areas of along the western seaboard of Britain. The similarity of house-type at Deer Park Farms, die Udal, and Buckquoy, is just one feature worthy of further investigation in this regard.
The Christian sentiment of the Buckquoy text throws interesting light on the current debate over the relative strength of the Irish and the Nonhumbrian input in die evangelization of the Northern Isles [Morris 1990:9]. Continuing links between the Irish church and Orkney are demonstrated by the reference in die ninth-century life of St. Fintan of Rheinau [Thomson 1986, Lówe 1986] to an Orcadian bishop, presumably a Pict, who had studied in Ireland in his youth and could speak Irish to die fugitive Fintan. That he was
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
185
the only person in the area who could, implies that knowledge of Irish was not widespread in Orkney in the 840s. It is impossible to evaluate the Buckquoy ogham without reference to the excavated remains on the Brough of Birsay, just a few hundred metres away. Although Buckquoy was a farmstead, not a high status site, its proximity to the Brough means it cannot be disassociated from it. Might the Buckquoy farmers have been monastic tenants of the Brough ? Perhaps, although recent interpreters have been turning away from the traditional identification of the pre-Viking structures on the Brough as a monastery [Hunter 1986:171, Morris 1990:15, 1993:287]. The ogham text implies no more than that the owner was Christian, certainly it need not imply that he or she was a monk or nun, nor that Buckquoy was an ecclesiastical site. The oghams most closely comparable to Buckquoy, those from Gurness, only ten miles away, from Pool, and from Bac Mhic Coonain, North Uist appear to have no ecclesiastical connotations. They share with the oghams from the Brough itself an informality which suggests that ogham literacy was not highly restricted, or reserved for solemn, formal purposes. It would be wrong, therefore, to assign an ecclesiastic interpretation to the Brough on the basis of the presence of ogham.
CONCLUSION In the above discussion I have attempted to explain the evidence as best I can, while frankly acknowledging the difficulties in both reading and interpretation. It is a frustrating reality of early medieval Insular epigraphy that vernacular inscriptions which do not conform to a firmly established textual formula rarely yield an utterly unambiguous reading [e.g. Williams 1949, Clancy 1993]. Even the intensely studied runic inscriptions of Anglo-Saxon England, written in a language vastly better attested than Pictish, present numerous difficulties and are much disputed [Page 1973]. In the circumstances it is perhaps unrealistic to expect such an informal ogham inscription as Buckquoy, from a little understood period in the development of the script, to be completely unequivocal. The most I have been able to show is that mine is a possible interpretation. If I have been unable to demonstrate that it must be so, then I await with interest a more convincing explanation.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BUCKQUOY
186
BIBLIOGRAPHY
OS Record card - HY 22 NE
Padel 1972:73-5; A. Ritchie 1977:181-2, 197 fig.8 no.84, 199 pi. 13a; K. H. Jackson 1977; Collins 1977; A. Ritchie 1987:62, 65; Holder 1990:1990:66-70.
PREVIOUS READINGS
Jackson Padel Holder
(e/ )TMIQAVSALL(c/q) or (e/ )(s/n/ )DDACTANIMV as above E(s/n)DDACTA(n/lv)IM(v/lb)
BUCKQUOY - 'Figure-of-eight' house in which inscribed whorl was found [Ritchie 1977:177 fig.3J
BUCKQUOY - Ogham-inscribed spindle-whorl [Crown Copyright: Institute of Geological Sciences, NERC]
I8p5
it,
10
12
BUCKQUOY ~ Ogham inscription (key to numbering)
187
BURRIAN
DISCOVERY The ogham-inscribed slab was discovered in 1870 during excavation of the Broch of Burrian on the small island of North Ronaldsay, Orkney (NGR HY 7627 5138). The excavator and proprietor of the island, Dr William Traill of Woodwick, was assisted by the antiquary Sir Henry Dryden and the excavation, which lasted from 1870-71, was comparatively rigorous for the period. Traill identified three phases of use, the primary broch occupation, a secondary occupation inside and outside the broch, and thirdly, early Christian activity, evidenced by finds only. It was in this third and final phase that the ogham slab was discovered 'towards the south side of the broch, where the wall was so low that, though the slab lay not much above the floor of the tower, it was also notforfromthe surface* [Traill 1890:346]. As early as 1871 Sir Henry Dryden had solicited Brash's views on the ogham, as is apparent from Brash's letter reproduced in 77?£ Ogam Inscribed Monuments of the Gaedhil [1879:362-3]. This is the first mention of the ogham in print and Brash's is the first illustration [see also pan of a letter from the late Brash published in Traill 1890:350-1]. Though Traill had published a report on his excavation [1890], die bulk of his finds remained unpublished until 1974 when MacGregor re-published die site, paying particular attention to die finds, including the inscribed slab. A sevendi- or eighth-century date was suggested for the early Christian phase at Burrian and theformof the incised cross has been assigned to the late eighth century on stylistic grounds [Thomas 1971:187]. In 1871, along widi odierfindsfromthe site, the ogham-inscribed slab was donated to the National Antiquities Museum where it is currently on display (NMS Cat. No. GB1).
SITE The four square miles of North Ronaldsay comprise die most northerly of the Orkiley tsles and 'surely one of the most bleak' [folacGregor 1974:63]. No-where on the island is more than 15m above sea-level. The Broch of Burrian is at the tip of a promontory in the very south-east of the island, overlooking a rocky shore and separated from the arable land by four earthen ramparts. The promontory is just to the east of
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRI A N
188
the broad sandy 'South Bayf where the present day harbour lies [Hedges 1987.III: 105]. Some of the site has been lost to marine erosion and only the broch tower was excavated in earnest, so knowledge of the site is partial. Traill describes the 'one to two feet of rubbish' he found above the original broch floor, above this was 'an upper paved space, and various partition walls, built upon the debris that overlay and concealed the original floor* [1890:341]. These two pavements formed, in MacGregor's words, a clear stratigraphical reference point, dividing the architectural remains into primary and secondary phases. According to MacGregor, however, the separation is not as clear-cut as is often assumed. He points to discrepancies in plans and lists and wains that 'a great deal of caution is called for when using the stratigraphy in relation to the finds' [69-70].
A number of finds are diagnostically Pictish, most famously the ox phalanges inscribed with symbols (NMS GB 227), and the pebble inscribed with a pentangle and other designs (NMS GB 9). But there are other more humdrum items. Two artefacts, the cross-slab and an iron hand-bell (NMS GB 306), point to ecclesiastical activity at the site.
Radford claimed the site as a monastery or Christian hermitage
[ 1962:170], a view more recently supported by work on hand-bells of Irish type in Scotland [Bourke 1983]. Bourke has concluded that in Ireland the distribution and associations of iron bells, in possible contrast to those of bronze bells, suggests a connection with monastic churches. Iron bells of this kind were produced in Ireland from about AD 700-900 and their presence in Scotland 'can reasonably be attributed to the influence of the Irish church' [465]. At only 60mm tall, the Burrian bell is the smallest of all the extant early medieval hand-bells from the Celtic-speaking regions. Bourke entertains the possibility that it may be a cow-bell, but points to other clearly ecclesiastical bells which are only slightly larger [464]. Two other hand-bells of Irish type survive from Orkney, but both of these, from archaeological excavations on the Brough of Birsay and at Saevar Howe, were discovered in Norse contexts.
As MacGregor points out there are other examples of the foundation of monasteries on formerly defensive sites, with the ramparts then serving as a symbolic barrier between the clerics and die world [103].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRIAN
189
Marwick ascribed the early Christian remains at Burrian to a Ninianic mission [1952:1], but his identification of the name of the island, Rinansey or Rinarsay, with the popular form of the saints name, Rinan, has since (sic) been refuted [Taylor 1931]. The indisputably Irish background of the bell and, of course, of the ogham, renders any Ninianic, or for that matter, Northumbrian, connection unlikely. The subsequent history of the site is unknown, but impressive standing remains must have survived into the Norse period to have attracted the name 'Burrian' (< ON borg-in 'the fortress') [Marwick 1952].
DESCRIPTION OF OBJECT A small, smooth, rectangular slab, incised on one face with the outline of a cross and, parallel to the left margin, an ogham inscription. Stone:
Clay slate
Dimensions:
0.7m x 0.39-0.29m x 0.05m
Condition:
The slab is complete but its surface is severely laminated and much has been lost, including some of the carving. The ogham is, however, largely intact and clear.
The slab itself is a natural flag only roughly shaped. What survives of the carved surface is, however, very smooth. To such an extent, in fact, that Macalister suggested it had been deliberately smoothed to receive the carving. Unfortunately the stone is highly laminated and much of the friable surface has flaked off, especially in the lower and right-hand portions. According to Traill this damage was caused by the roots of plants which, on discovery, had penetrated the surface of the stone [1890:346].
From the excavation report there is no indication of how the stone lay when it was discovered, whether vertical or horizontal, or whether the excavators had an opinion as to whether it was in a primary or secondary position. From Traill's discussion [346], it appears he considered the slab an upright gravemarker. Whilst the stone bears a superficial resemblance to a recumbent grave-cover, the fact that the carving is restricted to the upper two-thirds may suggest that the stone originally stood upright in the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRIAN
190
ground, perhaps at the head of a grave. A cursory glance at CIIC II suggests that the carving on Irish recumbent grave-slabs tends to be more centrally placed and to cover a wider surface area. If it did stand upright, the slab would have been less than two feet tall (0.6m). Leaving aside the lost surface area, the extant carving is remarkably unweathered, which may imply that the slab was not exposed to the elements for long. Of course, the monument need not have been funerary, and the suggested interpretation of the text is non-memorial. The cross may instead have been devotional. If the position inside the broch is primary (which may account for the lack of weathering) and the site is to be interpreted as ecclesiastical, the cross may have been an item of church furniture, perhaps propped upright against the altar. In which case the comparison with the Flotta altar front gains added significance [see below].
DESCRIPTION OF CARVING Both cross and ogham are very lightly and delicately incised. The cross is a simple though pleasing blend of neatly executed straight lines and smooth curves, measured and regular, as if drawn with compass and ruler in hand. There is no trace of preliminary guide-lines or points, most likely the outline was drawn on with charcoal or traced round a template. Several writers have noted the contrastingly scrappy appearance of the ogham and commented unfavourably on its wavering stem-line and irregular lettering. The contrast between the two is bound to raise doubts over their contemporaneity, especially since the cross is slightly morefirmlyand deeply incised. But it would be wrong to make too much off this, the technique is basically the same for both, and the fact that the cross is placed a little to the right, as if to allow for the ogham, seems to prove that both are part of the same over-all design. If the suggested interpretation of the text is accepted the inscription refers to the cross which, though not incontrovertibly so, further supports the view that the two were carved at the same time.
DESCRIPTION OF DECORATION In addition to the ogham the slab is incised with the outline of an equal-armed cross (220mm broad), with circular joints and square terminals, on a narrow shaft which disappears into a patch of wear (total extant
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRIAN
191
length c.370mm). Several writers have noted the very similar cross (though lacking the shaft) on the altar frontal from Flotta, one of the most southerly Orkney islands [Thomas 1971:186-8]. The cross form is not particularly widespread, though it does occur on the huge cross-slab at Rosemarkie ROS, again minus the shaft. Perhaps a closer parallel is found at Monymusk ABD or less precisely at Ulbster CAI, both of which share the distinctive feature of the narrow shaft which gives the Burrian example the air of a processional or free-standing altar cross. All four of the comparanda mentioned are decorated with interlace in some way, but the Burrian cross is plain. Of course, no trace of pigment survives, but if, as seems likely, the practice of painting sculpture was common, the smooth fine-grained stone of this slab would make it ideal for illumination. Thomas suggested a date in the late eighth or ninth century for the Flotta altar, based on a comparison of its interlace with that of a ninth-century cross in Cumberland [188] and on the assumption that the Burrian ogham is 'unlikely to be any older than the eighth century' [187]. To use Thomas's date for the Flotta altar to confirm the date of Burrian is thus a circular argument, based ultimately on the unfounded assumption that a developed form of ogham such as this can be no earlier than the eighth century. The probably seventh century radio-carbon dating of the bind-ogham-inscribed Birsay 3 shows that such an early date cannot be ruled out for Burrian. The Monymusk comparison may be as relevant as the Flotta, and it is doubtful if a date as late as the late eighth or ninth century could be accepted for the simple Class n Monymusk slab. No clear date for Burrian can be established on stylistic grounds alone.
The two other shafted crosses mentioned, Monymusk and Ulbster, stand on small rectangular bases, but the base of the Burrian cross, if it ever had one, is now lost. The disjointed patches of original surface which survive in the area beyond the current end of the shaft display some lightly incised lines. To the right are five or more oblique parallel lines some with mirror images below, the whole looking something like a feather; to the left are more nebulous curves. From the outset this has been referred to in print as a fish [Brash 1879:362], an idea perpetuated by no less an authority than Romilly Allen [ECMS 24]. While the carving to the right could be seen as a fish's tail (though unlike any in Pictish an), the design
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRI AN
192
to the left is hard to rationalize as a fish-head. The original interpretation rested on no more carving than survives today, and there seems slender justification for it. The Burrian 'fish' faces left, yet with only a few exceptions (e.g. Edderton SUT which points up to the left at an angle of 45°) most Class I fish face to the right, their fins indicated, if at all, by a few slightly curving lines. There are only two Class II examples which face to the left, Golspie SUT and St. Vigeans 1 FOR, but neither has fins represented in this feather-like way.
Padel suggested a correlation between the fish symbol and ogham inscriptions and this idea has been repeated by others. In fact, of the nine ogham stones occurring with symbols, only four, Ackergill, Inchyra, Golspie and Latheron have fish. While this is higher than might be anticipated, given that the fish is not one of the more frequently occurring symbols (twelve examples on Class I, less than 10%), there is doubtful merit in performing statistical analysis on such a small sample. The marks are certainly artificial, but they are much less substantial than the cross and less substantial indeed than the ogham, so that even if one were to concede that they were intended to represent a fish, they look more like subsequent graffiti than a proper 'symbol'. There is certainly insufficient justification for classifying Burrian as a Class II slab. Macalister rejected the 'fish' and instead interpreted the carving as the lower ends of the shaft turned out and decorated with feathering. There are a variety of unusual cross bases in both Pictish and Irish an, though none of the form he describes. The NMS photograph reproduced here shows up a number of scratches and scores, some of which look more deliberate than accidental. On balance I think the 'feathering' is as likely to relate to this than to the original design.
OGHAM INSCRIPTION The ogham inscription runs vertically up the face of the slab a few centimetres in from the left margin, level with the cross. The extant length of the text is 345mm, the individual letters ranging in size from 15mm to 3mm. Thus, with 26 letters in the space of 35cm, Burrian is carved on a very fine scale. Ferguson described it accurately as 'the most minute lapidary Ogham hitherto found' [136]. The letters,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRI AN
193
most of which are bound, are very clearly spaced one from another though there is no explicit word division. Surface damage has rendered the beginning of die ogham indistinct. It is possible there were furdier letters below what is now visible, but no traces remain. The rest of the inscription is fairly clear diough there is slight doubt over a few groups, especially 213 and 10. As it stands, die lettering finishes level with die top of the cross. Above there is a pitted gap before the fracture line of die stone. There are no further markings apparent and the text probably ends here. Reading from die bottom up [see fig. for key to numbering]: 1
Five strokes perpendicularly across the stem -1. The left bind-stroke is visible but not die right since the surface is lost beyond die distal end of die strokes. Any letters which may have preceded have not survived.
2
The surface between here and 4 has suffered disruption, resulting in confusion and disagreement between the various audiorities. Padel and Macalister agree that first diere are two strokes to the left of the stem, bound, with die second protruding to the right of the stem. But one chose to interpret die bottle as half-full, the other as if it were half empty. Padel diought the right surface 'not too worn' and, while conceding a possible O, preferred to interpret the lines as a mistaken attempt at a D. On die other hand, Macalister saw it as an imperfectly preserved O. I diought I saw a diird stroke before the others, with the left distal end lost in a spall and continuing across the stem a similar distance to that of the third stroke. I concur with Padel that there is no obvious wear to the right of die stem at this point, but the first and third strokes certainly protrude onto that side. I am reluctant continually to postulate slips and errors so, for want of an obvious reading, proffer die following alternatives - T, D, O, U.
3
After a gap diere are two short strokes to the right of die stem, sloping forwards. Macalister interpreted these as L without further comment. The lack of a bind-stroke renders this unlikely but in any case it is clear that the second stroke, at least, continues across the stem. There is a nick at the point where die right distal end of die first stroke would lie, but no evidence of anydiing between diat and the stem. Padel interpreted these as two separate letters, BM, but was
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRI AN
194
unhappy that they were so cramped. Neither mentions what appears to be a third stroke, sloping in the opposite direction, most obvious to the right of the stem, but continuing across to intersect with the second stroke just to the left of the stem. It is not clear if this stroke ever continued further. Given the ample spacing of the rest of the inscription I think it unlikely that there is more than a single letter here. That the first stroke is not bound to the second throws it into some doubt, since all other letters are bound. The third stroke is doubtful, but if accepted could be intended for the first forfid > <. That the two opposing diagonals do not intersect exactly on the stem would be explained as just a careless slip caused by carving the letters first and then adding the stem to join them up [see below for interpretation of forfeda]. One might speculate that the reason this letter is so indistinct is because it was a botched job deliberately effaced by the carver, but really this is clutching at straws. 4
Six strokes perpendicularly across the stem. The fourth stroke is very worn, to the extent that it might not be a stroke at all, but to reject it would leave an otherwise inexplicable gap. Macalister describes it as 'a short stroke interpolated between third and fourth strokes', Padel does not refer to it. It is most plausibly explained as a simple slip in counting out the strokes. The right bind stroke is visible, but the left is largely worn away. The left portions of the third and fourth strokes, and the right portions of the fourth and fifth are very worn. The first few strokes of this group are perpendicular to the stem, the final few less neatly placed, but there is a clear distinction in gradient between this letter and the next two which are obviously oblique. In any case, three consecutive Rs would be unlikely. There is no doubt that the letter is a vowel, hence probably I. From here the inscription is clearer.
5
Five strokes across the stem, sloping forward, bound at both ends - R. Each is slightly longer on the right side than the previous one with the effect that though the left bind-stroke is parallel to the stem, the right bind-stroke slopes away from it. The short stroke between this and 6, mentioned by Macalister and Padel, is not bound to the rest of the group, though the right bindstroke over-shoots the fifth stroke slightly. Padel suggests it is a false start for the next letter.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRIAN
195
Given the clear spacing elsewhere, this is the most likely interpretation. 6
As above - R. This letter is bigger and its strokes increase more markedly than 5's. Padel thought he saw a possible 'small stroke' to the left which he interpreted as H 'if a letter at air. Macalister saw the same to the right - B. The area in question is chipped and I could see no trace of the alleged nicks to either side. Unless the surface was already damaged when the letters were cut there would be a surprisingly large gap before the next letter.
7
A single stroke perpendicularly across the stem with a bind stroke on the right distal end - a hammer-head A (cf./i).
8
Five short strokes to the right of the stem, sloping forward, bound - N. The individual strokes are of equal length thus the bind stroke is parallel with the stem [pace Padel].
9
As above - N. The letter bows out slightly with the middle strokes longer than the first and last. There follows a faint and dubious short line parallel with the stem as if a right bind-stroke for a two-stroke letter, but there is no trace of the relevant strokes and to accept it as a letter would make this section unfeasibly cramped. The line is perhaps best dismissed as a casual scratch (there are plenty of these elsewhere across the surface) and its position relative to the stem as merely coincidental.
10
Next is a very worn patch out of which emerge the distal ends of three strokes to the right and three to the left. These slope in the direction of reading and if produced would meet at the stem to form three right-feeing arrows > > > ; each slightly larger than the previous one. The lack of bind-strokes is surprising, but nonetheless this group is probably best interpreted as an angled vowel - U. As Padel notes, the wear could be obscuring the further two or three strokes for which there is room, but there is no trace of anything further. The worn area is not panicularly wide and it would be thought that the distal ends of any lost strokes would have survived. If there are no letters missing, the apparent gap is hard to explain unless that area was already damaged when the ogham was carved.
11
Five or six strokes across the stem, bound - R. The lower tips of both bind-strokes are clearly
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRIAN
196
visible before the first stroke (if five), the extra stroke, if present, would run along the fracture line of the laminated surface. It is curious that the first R of both pairs, 516 and 11112, may consist of six rather than five strokes. Each stroke begins progressively nearer the stem on the left and ends further away to the right with the result that the left bind-stroke slopes towards and the right bind-stroke away from the stem. 12
As above - R. The final stroke is angled at the stem with the point facing to the right. Slightly smaller and more compact than the previous letter.
13
A single stroke perpendicularly across the stem with a bind stroke on the right distal end - a hammer-head A.
14
Four short strokes to the left of the stem, sloping forward, bound - C. Each stroke is slightly longer than its predecessor with the effect that the bind-stroke slopes slightly leftwards, away from the stem.
15
Three short strokes to the left of the stem, sloping forward, bound - T. As with the previous letter the strokes get progressively longer.
16
One long stroke perpendicularly across the stem intersected by a second lying at an angle of less than 45° to the stem. The perpendicular stroke is deeper than the oblique one. Tliis has been interpreted as thefivstforfid > < and assigned a value K by Padel. It is, however, quite unlike 21 which is formed, not of straight intersecting lines, but of opposed angled lines which do not quite meet. The intersection does not occur exactly on the stem but slightly below it. If the slighter oblique stroke could be shown to over-lap the perpendicular one it might be possible to explain the carving as an attempt to correct a mis-carved M. If the initial stroke was carved perpendicularly rather than obliquely the second stroke may have been added to rectify the situation. The fact that this surprisingly long stroke runs so far to the right that it intersects with the first stroke of / 7 might suggest it was added after / 7 had been carved.
17
Four strokes across the stem, bound - E. The left bind stroke starts well before the first stroke (the way it trails off suggests it was cut with a stroke contrary to the direction of reading and
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRI A N
197
overshot its mark). 18
Tliree strokes to the right of the stem, sloping forward, bound - V. They are well spaced and get progressively longer.
19
As above, though more tightly spaced and of equal length - V.
20
Four strokes to the left of the stem, sloping forward, bound - C. Each stroke of this tiny letter is slightly longer than its predecessor.
21
Two long angled strokes, opposing, not meeting > <. Macalister says that there is no stem in the space between these two, and goes on to draw certain conclusions about the level of literacy of the oghamist and the way die text was carved. In tact die surface is so worn that, even though there is indeed nothing visible between the two strokes, no conclusion can be drawn about the original presence or otherwise of die stem. This same character appears on die Formaston and Lunnasting stones but it is unclear whether it is to be considered a variant of die first forfid or a separate character in its own right.
22
Cross-hatching - two superimposed sets of five parallel oblique strokes across the stem in opposite directions. This forfid is known from die manuscript tradition and is found in stone at Bressay. It has the value RR [Sims-Williams 1992:72J. It is unclear if there is any difference in the sound represented by this character as opposed to the more usual pair of Rs, e.g. 516,11112. In carving technique the strokes which runs left-to-right-down are lighter than those in the opposite direction, it appears therefore that the heavier right-to-left-down have been superimposed, in odier words were carved second.
23
Two long strokes perpendicularly across the stem, bound - O. The first stroke overshoots the left bind-stroke. At this point the stem begins to veer to the left but the letters maintain die correct position relative to it.
24
Four short strokes to die left of die stem, sloping forward, bound - C.
25
As above - C. The fourth stroke is bent about half-way along and points in die direction of reading.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
26
BURRIAN
198
Four short strokes to the right of the stem, sloping forward, bound - S. Each stroke is longer than the one before.
Which gives a reading of: I(t/d/o/u/?)( /e/b)IR( /a)RANN(u/)(./ )RRACT(k/e/m/?)EVVC> <EROCCS
FORM OF SCRIPT Macalister took the untidy execution as symptomatic of the 'unintelligent copying of a prepared model' by an inexperienced and possibly ignorant writer [1940:209]. He erred, however, in assuming that the oghamist's aim was necessarily to produce a neat, calligraphic and monumental text, like, for instance, Lunnasting. The Burrian script is of a quite different register. Of all the Scottish oghams this is the one most resembling ordinary 'hand-writing'. As such, it is a perfectly acceptable cursive text, which would not look out of place if it appeared in ink on the margin of a vellum manuscript. The slips and irregularities, such as over-shooting bind-strokes and variable stroke and letter size, are of the kind made by someone who knew what they were doing and understood the workings of the script. Rather than indicate ignorance of the script, I think the lax execution implies the reverse, familiarity with ogham and confidence in its use. Despite its somewhat careless execution, the Burrian text is clear and legible except where damaged. This is due to a number of features including the consistent use of bind-strokes, the varied but usually generous spacing between letters, and the general proportions of the script.
The Burrian oghamist has been criticized for not forming a proper intersecting X for the úisi forfid (27), but we are perhaps wrong to assume that this was the carver's aim. An identical character, consisting of two angled non-intersecting strokes, occurs at Formaston and possibly at Lunnasting, in the former case the two halves are even further apart than at Burrian. The characters -X-and > <- may be two quite separate forfeda. Nor is it clear that the úvstforfid is what was intended at 16. The third 'mistake' is the superfluous sixth stroke of the R at 5 and 11. While in both cases the extra stroke is doubtful, it is curious and perhaps not accidental that in both cases the first R of a pair should have six strokes.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRIAN
199
There is insufficient evidence to provide a contextual date using archaeological methods. A re-assessment of the comparanda for the form of the cross may permit art-historical dating and a full interpretation of the text may allow linguistic dating. The third potential dating factor is the typology of the script. Burrian exhibits one of the most developed forms of the script extant in Scotland. All its letters are bound, as appropriate, it contains angled vowels, hammer-head As, and other forfeda, including the cross-hatched double R. While several of these features are found on the Bressay slab which dates probably from the ninth century, the first two appear on Birsay 3 which has been archaeologically dated to perhaps the midseventh century, and the third occurs on Buckquoy which is probably earlier than the mid-eighth. So, while a date in the mid-to-late eighth century is a possibility, an earlier date cannot be ruled out.
It has been conventionally assumed that all Celtic ecclesiastical establishments in the Northern Isles were abandoned in the face of pagan Norse invasion. But, as Wainwright has pointed out Celtic Christian an survived the Scandinavian settlement [1962:115], a point more recently reinforced for Shetland by Stevenson [1981]. Place-names of the Papa indicate Norse recognition of Christian hermitages or monasteries [MacDonald 1977], and, perhaps most striking of all, the ninth century Vita Findani, written on the Continent by an Irish monk, relates the adventures of the Leinstennan Findan who, in escaping from Viking slavers in Orkney in the 840s, came to the seat of a bishop. Though not himself an Irishman (presumably therefore a native Pict), this bishop 'had been instructed in the study of letters in Ireland and was quite skilled in the knowledge of this language* [Omand 1986:287; Lowe 1986], This important evidence for Irish educated clergy in Orkney a generation or more after the Norse settlement means there is no de facto reason that the Burrian slab must necessarily date before 800.
INTERPRETATION OF TEXT The first problem encountered in attempting to interpret the Burrian text is the value to be attached to the various forfeda:
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRI AN
200
22 'Cross-hatched RR': There is no obvious reason why RR should have been written here with theforfid rather than in full (cf. 5-6,11-12\ since there is apparently no phonetic difference and space is not at a premium. What appears to be the same word is written at Bressay with two Rs. The use of the forfid may be just an elegant variation, what phonetic significance it might have is hard to imagine. Sims-Williams suggests this character may reflect a desire to differentiate the geminate or long liquid /rr/ from /r/, which is plausible, though his sole examples are here and Bressay [1992:72].
7, 13 'Hammer-head A': See entry on Formaston where it is tentatively suggested that this character may be the equivalent of manuscript ai.
10 'Angled vowel (U)': There is no indication whether or not angled vowels have a value any different from their straight counterparts. The angling may be a purely visual device aimed at further differentiating vowels from consonants and thus enhancing legibility. See entry on Lunnasting for the possibility that vowel-stroke may be angled to indicate vowel length.
21 (possibly 16 and 3): The difficulty lies in judging whether or not this character with opposing but not touching angles (>
The second problem is word division. On the basis of the possible identification of URRACT and CERROCCS, and aided by the principle that geminate letters are not usually found in word-initial position, the following word division is suggested:
I(..)IRRANN URRACT KEVV CERROCCS
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRI A N
201
From the outset the final word has been equated with 'cross1 [Brash]. While this most obvious interpretation is probably the correct one, the unusual spelling has never been satisfactorily explained. Of course the ultimate source of the word in all the Celtic vernaculars is Latin crux. In Latin, the grapheme x represents the sound /ks/. In Brittonic this became in intervocalic position /xs/, with subsequent vocalization of the /%/ to short /i/, i.e. L. crux > W. crwys, OC. crois, MB. croes, croas [Jackson 1953:535 §125]. In Irish c/ragave cross [Thurneysen 1946:575]. The Burrian spelling appears to indicate a stage when the /ks/ of the Latin was preserved, with the double C perhaps indicating a non-lenited sound, /k/ as opposed to l%l. The E in the first syllable looks like an epenthetic vowel, which as Padel points out, might arise in a borrowing if the target language didn't have initial /kr/, but this is puzzling since initial /kr/ poses no problems in Celtic. An alternative explanation is that, though 21 is the úrstforfld, it has, in this case, die consonantal value K = l%l. This would give die reading /kxroks/ which is no less inexplicable, but if the initial C were, in fact, the final letter of the previous word (the unintelligible KEVV), then the final word would read /xroks/ (i.e. /kroks/ with lenited initial). If the basic form is /kroks/, then we have a perfectly regular *CRROCCS /kroks/ from cntx /kruks/. This leaves us to explain the third word and why it would cause lenition, or rather, since we appear to be dealing with a text in a Brittonic language, spirant mutation (which entails the same initial mutation, /k/ > /%/, but for entirely different phonological reasons) [Jackson 1953:634-8]. In British, proclitics ending in a consonants gave rise, in close speech groups, to an external sandhi known as 'gemination1. In Neo-Brittonic the result of such gemination was that in the second word initial p-, t-, c- became /f/, /6/, /%/ (spelled in f, th, £h manuscripts). By analogy, spirant mutation spread to other contexts, and conversely was lost in some phonologically appropriate contexts. In Middle Welsh spirant mutation occurs only in a few specific grammatical contexts [Evans 1964:21 §24], but as Jackson comments 'there is considerable disagreement between the three languages [WCB] and it is obvious that there was much levelling out by analogy and other causes working in different ways' [1953:63].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRIAN
202
The Pictish word for cross may be preserved in the place-name Applecross, Apur-Crossan [ATig 722] since the first element is clearly Brittonic, i.e. Pictish. It is quite conceivable however that Adomnán wrote the name as it was known to Irish-speakers, who were able to substitute the Irish form of the very similar second element, but not the less familiar first. The word 'cross' seems to appear in another Scottish ogham, that of Bressay, though in this case the spelling, CROSSCC is very close to the Scottish Gaelic form crosg. It may be relevant to note that one of the Scottish runic inscriptions inscribed on an ultimately Celtic-inspired cross-slab at Kilbar, Barra, has a metathesized spelling of 'cross' - kors [Barnes SC 8].
The other word which may be readily identifiable is the second, URRACT. Unfonunately the initial letter of this word is damaged and doubtful. If it is indeed to be taken as UU or U then we may have here a most interesting form. John Koch suggests to me that URRACT may represent the Pictish cognate of the Old Welsh 3sg.preterite *guract 'he/she made', cf. gwreith 'I made' [Williams 1938:44.1102 (Canu Aneirin - Peis Dinogat)]. In MW the form was gwnaeth since r had become n here by analogy, though it was preserved in Cornish and Middle Breton cf. C. gwmf MB. groqff 'I do' [Evans 1989:130-2; Thurneysen 1946:111].
If the above interpretation of URRACT and CERROCCS is accepted, then the Burrian text appears to represent a sentence - 'blank he/she made blank cross'. In which case the first 'blank' is probably the name of the person who made the cross. Given that the text appears to be written in Brittonic, we might expect a Brittonic name. In addition to Birran, the Book of Llandaff, for instance, includes a number of compound names with second element bran 'raven' (Conbran/ Cinbran cf. Irish Branchu, Dibran/Dofran/Dubron, Gabran, Loitbran, Morbran, Davies 1979]. Uhlich discussed Irish compound names incorporating the element bran [1993:184-6], and there are many more, both male and female, which end in -ran(n) (e.g. Morann (m), Eórann (f) and Cochrann (f), also potentially relevant are the male Odrán, Ailerán, Sárán, Ciarán, Barrán, Láarári) [Ó
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRIAN
203
Corráin & Maguire 1990:sw.]. The Burrian text is too corrupt for certainty, but the Irish male Idbrann [CGH LL 335 b 37] is a possible contender.
This leaves the third word (.)EVV or (.)EVVC. The doubt over the value of the initial letter is a serious problem. It is usually interpreted as the x-forfid and given the consonantal value K = /g/, but its resemblance to thisforfid is not overwhelming. Ferguson carried out an inpossibly far-fetched inversion to arrive at THETTS which he translated as 'this', which may be the right interpretation for the wrong reasons [1887:136], Comparison has been made with the HCCVVEVV of Lunnasting, which is clearly delimited by word-dividing dots, interestingly it is also the third of four words in a text. The correspondence, which is far from exact, may be purely coincidental since the personal names appear to be in different positions and Lunnasting is not a cross-slab. To accept it would rule out the word division (,)EVVC KRROCCS suggested above. Substituting the alternative interpretation of 16 produces MEW or MEVVC, neither of which is immediately suggestive of anything.
Notwithstanding the uncertainty over the third word, the above interpretation yields a sentence of SVO structure which is at variance with the expect Celtic word order VSO. The syntax of the verbal sentence in Welsh is an involved topic, complicated by the so-called 'abnormal order* of MW prose and the great variety of word order in early poetry (including SVO) [Evans 1989:179-81]. The value of the evidence of the Burrian sentence is compromised, however, by its epigraphic context. The general prominence given to personal names in inscriptions may be sufficient to account for its fronting here.
DISCUSSION At a total of twenty-six characters Burrian is a medium length text by Scottish ogham standards, its significance lies in its constituting the only extant sentence in Pictish. Not only is URRACT the only example of a Pictish verb, but in being so close to its Old Welsh cognate it provides important further evidence for the fundamentally Brittonic nature of Pictish. The text demonstrates the use of ogham to write
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRI A N
204
a language other than Irish, not merely non-Irish personal names, but non-onomastic parts of speech. From this we can infer the adoption of ogham by the natives of Orkney, not merely its use within a circle of Irish immigrants. In being other than a commemorative inscription the Burrian text expands the range of known purposes of epigraphic texts in Scotland, in this case to record the name of the person who 'made' the cross. We should perhaps bear it in mind when we come to interpret otherwise unqualified personal names on cross-slabs before we rush to assume that they are commemorative.
As discussed above, the informality of the physical appearance of the inscription may be taken as evidence for easy familiarity with ogham on the part of the carver. It is more difficult to evaluate the implications this might have for the audience of the text. The Burrian ogham is carved on such a tiny scale that it can only be read close-up, unlike, say, the much more obviously monumental Brandsbutt ogham, which could be read at a distance of several metres or even yards. If the Burrian cross was an item of church furniture it is possible that individuals knelt before it for meditation and prayer. The oghamist may have recorded his name in the hope that he would be remembered in their prayers, though he makes no explicit request. The formula of the craftsman's request for prayers is common enough in the roman alphabet inscriptions of Ireland, a famous example being the or do Maél-ísu Mac Bratdan U Echan dorigni in gres-sa, 'a prayer for MaeMosa mac Bratdan O Echan, who made this handiwork', on one arm of the early twelfth century Cross of Cong [CIIC 552].
BIBLIOGRAPHY
OS Record card - HY 75 SE 3
Brash 1879:362 pLXLK; Southesk 1884:200; Ferguson 1887:135-7; Traill 1890:346, pl.46 (excellent); NMS 1892:231-3; Rh?s 1892:292-4; Nicholson 1896:27-8, App.15-7,72-5; Rh^s 1898:372-3; ECMS24
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
BURRIAN
205
fig.20; Macalister 1940:209-10, fig.6; Padel 1972:75-9; MacGregor 1976:96-7, fig.21; Sims-Williams 1992:48, 72.
Site:
RCAHMS 1946:45-7 No. 192, 51 No.201
PREVIOUS READINGS
Ferguson
-RRA(n/g)(n/g) U(u)RRACT THETTS CCRROCCS
Brash
IALELRARBANN(u/ng)£ieRRACT(ea/p)EFFX(mm/aa)RROCCS [Traill 1890:350-1]
Southesk
NAALLVORARANNIfliRRACTMHEFFCMESOCCS
Rhys
-VORRANNV(u/e)RRACTPEVVCERROCCS
ECMS
UORRANN UURRACT PEW CERROCCS
Macalister
IOLIRRBANN UTRRACT KEVV CEROCCS
Padel
IdbmlRRhANNuRRACTKEVVCERROCCS
<^
9 NORTH RONALDSAY ^t&T^
ORKNEY ISLANDS
o Os
»o 15
ami. 30 km6ROCH o( BURRiAN
BURRIAN - Location map [MacGregor 1974:64 fig.l]
ta^^^
ll4l ll,llll,tlll||l |,, l
*
*
* **í**l*»#|l|
/
<#
^
10
0
liflíiliíl
#
50
1 1 1 1
»00 FEET
1 1 1 1
BURRIAN - Broch site [RCAHMS 1946 fig.88]
Cí>!C
BURRIAN - Ogham-inscribed slab [© Trustees of the National Museums of Scotland, Neg. P9554]
BURRIAN - Ogham-inscribed cross-slab [MacGregor 1974:97 fig.21
v„_ l
— j
BURRIAN - Detail of cross
26 25 24 23 22
21 20 19 18 17 16 15 14 13
TF
B
f\
12 11 10 9 8 7 6
Í
4 3 2
1 BURRIAN - Ogham inscription (key to numbering)
206
CUNNINGSBURGH
1-3
DISCOVERY Three ogham-inscribed fragments were recovered from the vicinity of the churchyard of the ancient, longdisused, church of Cunningsburgh on the east coast of the southern mainland of Shetland (NGR 4324 2792). All three are now in the NMS. The first (NMS IB 114) was found in the mid-1870s by Robert Cogle [Goudie 1876:20]. The second (NMS IB 115) wasfound,probably in 1874 or 1875 'close by the burying-ground' by the Rev. George Clark, the Free Church Minister of Cunningsburgh, and presented to the NMS in April 1883 [Goudie 1883:306]. The third (NMS IB 182) was uneardied in the spring of 1903 in digging a grave in the churchyard [Goudie 1904:54].
There is dispute over the existence of a possible fourth ogham-inscribed fragment from Cunningsburgh. Padel [1972:88] dismisses thefourthCunningsburgh ogham as a ghost which arose from a misreading by the RCAHMS of Goudie 1883:306 [=1904:53], and which was subsequently picked up by Wainwright and reflected in his distribution map [1962 fig.27]. In fact, a close reading of Goudie suggests that four fragments were indeed recovered. In the passage in question Goudie refers to afragment,consisting of only two letters, already in the Museum, this must be Cunningsburgh 1 (IB 114). He goes on to mention two fragments later found by Rev. George Clark 'at the same place and about the same time' as one another. The larger of the two (IB 115 = Cunningsburgh 2) had been given to Goudie 'some years since, shortly after its discovery'. The smaller was still in Clark's possession at the time of writing. The latter cannot be Cunningsburgh 1, as Padel thought, because by then it was already in the NMS. Nor can it be Cunningsburgh 3, which was not discovered until 1903. Corroboration is provided by Brash who mentions Cunningsburgh 1 and 'two other inscribed fragments' in The Inscribed Monuments of the Gaedhil, published posthumously in 1879. He states that Gilbert Goudie informed him these had been discovered 'quite recently' [366], which, since Brash died on 18 January 1876 [Brash 1879:xv], must mean 1874 or 1875. Neither Goudie's nor Brash's account is in the least muddled, and the conclusion must be that there
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
Q U N N I N G S B U R G H (GENERAL) 207
were indeed four ogham-inscribed fragments recovered from Cunningsburgh, that the fourth was not donated to the NMS, and is now lost.
SITE The ancient graveyard at Mail, Cunningsburgh is still in use, though nothing other than the eight or nine fragments of sculptured stone survives to indicate its period of foundation. No trace of an earlier enclosure remains, though the small rectangular platform on the north side may be the site of die chapel [MacDonald & Laing 1968:131], The only other antiquity is a seventeenth century grave-stone [RCAHMS 1946:13-4]. The area, Konningsburg 'King's burgh', was of great economic importance in the Viking period [see map]. Not only does it lie at the head of sheltered, south-feeing bay, with a rich hinterland of good agricultural land, but nearby, to either side of die Catspund Bum, lay die largest deposits of economically exploitable steatite in Shetland [P. R. Ritchie 1984:67]. Archaeological remains indicate that the scale of production was so large that the soapstone must have been intended for export, either within Shetland (it is known to have been taken to Jarlshof) or further afield. The coastline immediately adjacent to the deposits is rocky but access is provided via the sandy beach at Mail, one kilometre to the north-east [P. R. Ritchie 1984:70]. The graveyard lies at the eastern edge of the beach. In the mid-nineteenth century Mr Clark's predecessor as Free Church minister, Rev. Frederick Soutar, had noticed in the 'exposed face of the sand-bank facing the sea' 'fragments of charred bones, of molluscs, and other remains, which seemed to indicate refuse-heaps and human internments' [Goudie 1904:62], These could, of course, relate to any period. Mr Clark himself had unearthed 'a quantity of charred grain, and a small deep cup or vessel of steatite so thin and so carefully scooped out as rather to resemble a piece of pottery [63].
Iron Age activity in the vicinity is indicated by the ruins of a broch at Aithsetter, immediately opposite the graveyard, from which a gaming piece was recovered [RCAHMS 1946:45 No. 1187], and a little further away at Mousa. In 1992 a slab incised with what purported to be a Picdsh figure was recognized in the graveyard, however, the nature of the carving and the circumstances in which it was found are such that
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH (GENERAL) 208
it would be wise to be cautious about taking the slab at face value. It has been accepted as genuine by some authorities [Turner 1994], but, even though the formal eccentricity of the Cunningsburgh oghams would provide a possible context for the eccentric design of the slab, I would prefer to reserve judgement. The natural resources of the area and its demonstrable importance in the Iron Age and the Viking period, mean it was probably also important in the intervening Pictish period, even if the direct evidence for Pictish activity at the site is limited [Turner 1994:319]. Since ogham continued in use in Celto-Norse circles into the eleventh century at least, the presence of ogham alone is insufficient to indicate Pictish activity, though one of the fragments appears to incorporate a Celtic name, which makes the inscriptions easier to place in the pre-Norse period. Turner mentions current 'rumours in Cunningsburgh of buildings and hearths having been located while digging graves [in the kirkyard] in the past' [1994:319]. She notes that over the years 'a number of other stones have been dug out of the graveyard ... most of which have been thrown onto the adjacent storm beach*. According to the local grave-digger 'many of these were "scratchet stanes"', but Turner's brief search of the beach produced nothing [315].
RUNIC INSCRIPTIONS The importance of the area in the Norse period is further underlined by the discovery of three of Shetland's five runic inscriptions in and around the graveyard at Mail (the others are from Papil and Eshaness). The three (Barnes 1992 SH 1-3) are now in the NMS (IB 103) [see fig.]:
SH 1
A fragment of sandstone, found in 1872 by Robert Cogle, three feet below the bed of a stream at Aith's Voe, Cunningsburgh. The first runic inscription to be found in Shetland. Approximately 280 x 200 x 50mm. Water-worn, part of surface missing. Inscribed on narrow edge -krimr+
SH 2
A fragment of sandstone, found by Cogle in 1875, 'a very short distance from [SH 1] ... close to the old burying-ground' [Goudie 1904:62], inscribed on narrow edge -
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH (GENERAL) 209
(kt)+(tk). '[B]eyond reconstruction' [Liestol 1984:227], this text may be a palindrome. Barnes has suggested that SH 1 and SH 2 could be part of the same inscription [1992:37], both are split such that only the lower or middle part of the letters are intact. The two could date from 'any period after Norsemen reached Shetland until almost die end of the Middle Ages' [Barnes 1992:37].
SH 3
A slab of sandstone, approximately 1.30 x 0.15 0.15m, found in the burial ground in 1877 by Rev. George Clark after a deliberate search, built into the outside of the western wall of the churchyard, the inscribed face was exposed and covered in moss [Turner 1994:319]. Inscribed on narrow edge. Only the last pan of the text is preserved -eftir fodur sinn Porbjorn 'after his/her father Porbjom'. Barnes thinks this may be an example of the mainly eleventh century formula 'NN raised/placed/laid this stone/cross/etc. in memory of NN'[1992:36]. Liestol considered this 'the best of the older inscriptions in Orkney and Shetland' [1984:228].
BIBLIOGRAPHY Runes:
Stephens 1875; Goudie 1879:144-53; Anderson 1881:227; ECMS 19; RCAHMS I 47, III 14 No. 1136 (5-7), figs. 470-71; Shetelig 1954:158-60; Liestol 1984:227-8; Barnes 1992:36, 37, 40.
Site Only:
RCAHMS 1946.111:13-4 No. 1136; MacDonald & Laing 131; P. R. Ritchie 1984:67-70.
OGHAM-INSCRIBED FRAGMENTS - CUNNINGSBURGH 1
DESCRIPTION OF OBJECT An irregular rectangular fragment of greyish sandstone, carved on both faces» including pans of three bound ogham letters [see fig.].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 1
Stone:
Grey sandstone
Dimensions:
240mm x 200mm max. width x 45mm thick
Condition:
Very poor, fragmentary, heavily laminated in thick layers.
210
The fragment has been severely trimmed down, though the top edge may be intact. The bi-facial decoration suggests the original was free-standing, perhaps a cross-slab, though its original form and size are impossible to determine. The large scale of the letters suggests the original may have been substantial when complete. If the top edge is indeed primary then the text would have read vertically up the left-hand edge, the most common position. If the free-standing cross-slab hypothesis is abandoned, a horizontal arrangement, on the analogy of Abernethy, is also possible.
DESCRIPTION OF CARVING 'Front': The loss of considerable areas of the surface does not enhance the appearance of the decorated face of the fragment, but from the extant remains it is clear that, even when newly carved, the design was crudely executed. The relief bands waver in direction and are of varying thickness, the carved areas are wide and roughly pocked. The decoration, which consists of a false-relief pattern of opposed double curves enclosed in a rectangular border [ECMS fig. 1 la], may represent the remains of one arm of a cross-shaft. The outer border is not quite parallel with the edge of the stone, but the symmetry of the rest of the design is so rough that this is probably not significant.
'Back': The ogham on the reverse face is sharper, more regularly spaced and, though very little survives, appears slightly more accomplished than the curvilinear decoration on the front. The two, of course, are not necessarily contemporary. The individual ogham strokes have been pocked and smoothed, the stem-line cut again on top of the cross-strokes. Though much of the surface has been lost to lamination, it does not appear that there was any other decoration on this fragment, nor that the lettering was arranged on a separate band as at Whiteness, Brodie, and Abernethy.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
Q UNNINGSBURGH 1
211
DESCRIPTION OF OGHAM The surviving ponion of stem-line is 225mm long and clearly continued beyond its current limits. The surface is severely laminated and large chunks of it are missing. The individual strokes are evenly carved and the letters spaciously arranged. The maximum length of the long strokes is 65mm. The angle of the third letter indicates the text is to be read in the following direction [see fig. for key to numbering]: 1
Five long strokes perpendicularly across the stem, bound at both ends -1. The upper ponion of the first stroke is lost at the break, but is not in doubt. After the fifth stroke, a thick chunk has been lost immediately above the stem. The right edge of the missing ponion is straight, perpendicular to the stem, and its outer limit is a straight line the same distance from the stem as the bind stroke of /. This suggests that the surface has fractured along the line of a lost stroke, a theory at least partially strengthened by the slightly large gap between 1 and 2 (in comparison with the gap between 2 and 3\ though obviously, from such a shon section it is difficult to judge average spacing. The distance between the midpoint of the last stroke of / and thefirststroke of 2 is 50mm, the equivalent distance between 2 and 3 is 30mm, with the tips of the strokes even closer together. At most two bound strokes have been lost, and these must have belonged to the h-aicme because the surface below the stem is intact (possibly D or H). The outer edge of the recess curves in at the point where the first of a pair of strokes might be. There is no trace of these hypothetical strokes on the exposed surface and it is possible the shape of the missing layer is merely coincidental. Elsewhere, however, the laminations havefollowedthe line of bind- or letter-strokes.
2
Five long oblique strokes across the stem sloping forwards, bound at both ends - R.
3
Two long strokes across the stem angled at the mid-point to theright( Z > ) , bound at both ends. The lower bind-stroke clearly continues past the second stroke indicating that the letter originally contained at least three strokes - U, E or L
This yields a reading:
-]I( /d/h)R(u/e/i)[-
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 1
212
FORM OF SCRIPT The slight gap between the I and R is unlikely to mark word-division because where indicated, this is shown with pairs of dots on either side of the stem, not by spacing of letters. Angled vowels occur on Birsay 1, Burrian, Cunningsburgh 3, Formaston, Lunnasting, and possibly on Golspie. It is not clear whether the contrast between 'straight* and 'angled* vowels is a purely stylistic variation in script, or whether it was intended to differentiate between vowel sounds. Since both commonly appear in the same inscription, often along with further supplementary vowel letters, it may well be the latter (see Lunnasting for discussion of angled vowels).
INTERPRETATION Since so little survives, and from the middle of a possibly extensive text, it is difficult to extract linguistic information from this inscription. There is nothing about the sequence which requires or precludes interpretation as Celtic.
BIBLIOGRAPHY Goudie 1878:20; Brash 1879:366; Southesk 1884:206; Rh?s 1892:294-5; 1898:376; ECMS 17fig.lla,b; Goudie 1904:53; Nicholson 1896:30,57,85 N; Macalister 1940:215, pi.vie; RCAHMS 1946 No. 1136(1); Padel 1972:79-81.
PREVIOUS READINGS Southesk
IR-
ECMS
(e/i)R-
Macalister
IRu
Padel
IRu+
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 2
213
OGHAM-INSCRIBED FRAGMENTS - CUNNINGSBURGH 2
DESCRIPTION OF OBJECT A triangular fragment of grey sandstone, carved with two incomplete lines of ogham, across the face and along the arris. Stone:
Sandstone
Dimensions:
300mm max. height x 190mm max. width x 65mm thick
Condition:
Severely trimmed. Ogham on face well-preserved, but arris badly damaged.
The Cunningsburgh 2 ogham is more substantial than a piece of graffiti, but the original form of the monument is unclear. Since three of its sides, at least, must have been visible, it might have lain horizontally as a cover, or stood venically, either free-standing or as an end-slab in a composite construction. If the slab had stood upright then the two oghams would have read vertically upwards, as one would expect, but the upper surface could not then have been horizontal.
DESCRIPTION OF CARVING The only carving on the surviving portion consists of two lines of ogham letters, one written on the arris, the other across the broad face with a drawn-in stem line. On the surface, near thefracturedinside edge, there is the odd score which might be the remains of a third row of letters, but could as easily be casual damage. The short narrow face is evidently intact since the lettering continues over the edge and down the sloping side. Both sections are written in the same hand and doubtless form a single text. The slope of the consonants indicates that both are to be read in the same direction, but whether they are to be taken separately or sequentially, and if the latter, in which order, is hard to determine. It is impossible to estimate how much has been lost beyond the fracture. Reading from the fracture to the intact edge, the inscription is as follows [see fig. for key to numbering]:
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 2
214
Angle: 1
Two short strokes to the left of the stem. A third stroke would have occurred at the point of fracture - D, or any subsequent letter of the h-aicme (T, C or Q).
2
One short vowel stroke across the arris (A), with perhaps another two (U). Rh?s, Allen, and Macalister all read E, but four notches would be very cramped. Padel thought two (O), but was prepared to go up to three (U).
3
Three strokes to the right of the stem, sloping markedly forward, clear - V. The upper surface of the b-side abutting the arris is lost from this point to the first stroke of 6. Sufficient remains opposite 4 and 5 to indicate that no strokes have been lost there, but the missing portion is larger between 3 and 4. On the h- (i.e. narrow) side there follows a gap sufficient for a few strokes, possibly vowels. Padel thought he saw 'part of a single h-score'. Allen suggested O.
4
Two strokes to the left of the stem, sloping backwards - D. The distal tips of this pair of strokes are joined by a curved stroke, giving a 'hair-pin' impression.
5
Two short strokes to the left of, and perpendicular to, the stem - D. Slightly shorter than 4, though similarly joined at their distal tips. No doubt the difference in gradient is intended to distinguish between two successive letters of the same aicme.
6
Five long oblique strokes across the stem, sloping forwards. The portion of the first stroke to the right of the stem is lost but not in doubt - R.
7
Four short strokes to the right of the stem. The last two are badly spalled and only their tips survive - S. There may have been a fifth stroke - N. Thereafter most of the arris has been lost, but there is room for another two or three strokes. There are hints of a pair of short vowel notches, followed by a possible single vowel notch (OA, or U), but these are doubtful.
This gives a tentative reading: -]d+(o/u/e)V( /o)DDR(s/n)( /oa/u)[-
Face:
A stem-line 250mm long is carved across the broad face of the stone, parallel to the ogham-
inscribed arris. The stem-line continues over the far edge and down the narrow end face. The line is more
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 2
215
or less straight, but wavers slightly. The carving is generally clear and well preserved, the only doubt comes where the lettering crosses the short arris. There is an extensive spall to the right of the stem taking away the dps of the last few strokes of 1 and continuing to the edge of the first stroke of 4y but it is clear from the configuration of the extant strokes that no carving has been lost. The inscription is as follows [see fig. for key to numbering]: 1
Four strokes perpendicularly across the stem - E.
2
A single oblique stroke to the left of the stem, sloping backwards - H. The section nearest the stem is lost. The slope proves that this is an h-aicme stroke (cf. 3 and 5), not the remains of an M (cf. 8).
3
Three oblique strokes to the left of the stem, sloping backwards slightly - T.
4
Four vowel strokes - E. Macalister has the central pair bound at the right-hand edge. Padel mentions the possibility, but I could see no sign of it.
5
Four oblique strokes to the left of the stem, sloping backwards, roughly parallel with 2 rather than J-C
6
Two vowel-strokes across the stem, apparently bound with a curved 'hair-pin' stroke at the left edge only - 0.
7
Five oblique strokes to the right of the stem, sloping forwards - N.
8
A single long oblique stroke across the stem, sloping forwards - M.
9
Two vowel strokes across the stem, apparently bound with a curved 'hair-pin' stroke at the left edge only - 0.
10
Four or possibly five long oblique strokes across the stem, sloping forwards - S' or R. A stroke is visible to the right of the stem, just before the short arris, but to the left of the stem it appears to run along the arris itself, however the arris is damaged at this point. Macalister thought this final groove was less substantial than the other four and therefore doubtful, but most other authorities have accepted it. Padel mentions the possibility that this letter was 'bound or partially bound', but I could see no sign of it.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
11
CUNNINGSBURGH 2
216
Padel saw four strokes to the right of the stem on the end of the stone - S. The slight doubt over the status of the possible fifth stroke of /0, raises the possibility that the final letter could have five strokes (N).
Which gives a reading: -EHTECONMORs
FORM OF SCRIPT The form of the script used is distinctive, consisting of long, thin strokes, quite different from the thick, formal, square-ish lettering of Cunningsburgh 1. The marked distinction in length between shon vowels and long consonants is noteworthy. The component strokes of letters are evenly, though closely, spaced, but the letters are very tightly packed. In a few instances, e.g. face 3/4 or (5/7, adjacent letters are closer than component strokes. The slope of m-aicme consonants is not great, but all b- and h-aicme consonants are sloped also. It is unusual to have vowels doubly distinguished from consonants, by both length and gradient.
The inscription contains noforfeda as such, but there are four instances of an otherwise unattested feature a single curved bind-stroke on the left-hand ends of a pair of strokes, giving a 'hair-pin' appearance. On the face this appears twice for a vowel (6 and 9 = 0 ) , on the arris it twice appears with consonantal force (3 and 4 = D). None of the other letters in the inscription are bound. Since the device is unprecedented there is no way of knowing whether this is merely a scribal variant of an ordinary 0 and D, or whether it is meant to convey some linguistic nuance. The transliteration of the third member of the b-aicme is discussed in the Introduction [see also Sims-Williams 1993:162-70] where it is suggested that it should be transliterated as /f/ in a Gaelic context and /w/ in a Pictish (Brittonic) context. The transliteration of H is discussed in the same entry [see also Sims-Williams loc. cit.], it may indicate a spirant. In its general proportions, the script of Cunningsburgh is comparable to that of Pool, though the resemblance ends there.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 2
217
INTERPRETATION OF TEXT The arris portion is too doubtful and fragmentary to permit detailed analysis. The face, however, is more promising. In the Irish annals Latin mors, 'death', is the word which appears to be used for the violent death of churchmen [Anderson 1922:248 n.4]. Though the ogham alphabet is used to write a Latin leonine hexameter in the margin of the Bodleian Library manuscript of the Annals of Innisfallen, fol.40 v [Macalister 1910], Latin here seems rather far-fetched and a vernacular interpretation, either Gaelic or Brittonic, is to be preferred. The final -s causes some problems. It is clear and at the end of the inscription, and cannot be explained away as a segment of a following word now lost (unless the two lines are intended to be read sequentially, the face first, and the division of the two lines has not been guided by word boundaries). As Thomas Clancy has pointed out to me, the recurrence of S in what appears to be syntactically genitive position on a number of Scottish oghams (Altyre, Brandsbutt, Bressay, Gurness, Inchyra, Lunnasting) is striking and problematic, and certainly hard to explain as Celtic* it looks more Germanic. While this might be plausible at sites within a Norse ambit, including Bressay, Lunnasting, and Cunningsburgh, its recurrence in the Pictish 'heartland' makes some other explanation seem more likely. This apparently non-Celtic termination [Padel 1972:85] is all the more puzzling since the preceding letters appear to be the securely Celtic name Conmor, as identified by Rhys [Goudie 1904:54]. Welsh Conmor, attested, for instance, in the Book of Llandaff [Davies 1979], has been derived from Brit. *Cunoburros [Williams 1937:cxvii, Jackson 1953:485 (explaining the confusion between m and b)] t Cuno- being the well-known Celtic male personal name element 'hound' [Jackson 1953:413 §63, Uhlich 1993:209-20], and -burros 'puffed, bloated, big, proud', cf. W. bwr, MIr. borr, Gaulishbor- [Evans 1967:154-6]. The name is not attested in Gaelic. To interpret Cunningsburgh 2 in Irish terms would require taking
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 2
218
Rh?s pointed out the record in the Annals of Tigernach for 686 of the death of one Tolair aithicain, Tolar(g) (mac) Aithkain, (Ehttecon for Aithkan ?). There is doubt over this name, however. Stokes favours the Tigernach form [1890:392], but AU hasAcithaen (= Accidan?). The intensifying prefix a(i)th is common enough in Irish personal names [Uhlich 1993:168].
It appears in the gentilic name
ATHECETAIMIN on an ogham pillar from Corrower, Co. Mayo (CIIC 7). The noun aithechán, a diminutive of aithech 'rent-payer, vassal, peasant', occurs meaning 'little churl', but also 'giant, monster' [DH]. 'Great giant' is an intriguing translation in the light of the figure on the recently-discovered Pictish incised slab, which has been interpreted as a giant [see fig]. All of these interpretations leave the final -s unexplained.
That the segment, EHTEC, calls to mind the ETTEC(UHETTS) of Lunnasting and the ETTEC(A-) Cunningsburgh 3, both of which occur at the beginning of the first word in the inscription, suggests an alternative explanation may be required. If John Koch's suggestion that ETTECUHETTS is ette-cuhett-s 'this is as far as' is accepted (cf. MW. kyhyt 'of the same length, as long', cihit-on, cohiton used in Old Welsh and Old Breton boundary clauses as 'limits, extents, what is as far as', see Lunnasting entry), Cunningsburgh 3's ETTECONMORS may be a parallel ette
BIBLIOGRAPHY Goudie 1878:20; Brash 1879:366; Goudie 1883; Southesk 1884.206; Rh?s 1892:294-5; 1898:374; ECMS 16-7 fig.10; Goudie 1904:53-4; Nicholson m - 21-23, 35-36; Macalister 1940:215-7, plvia: RCAHMS 1946 no. 1136(2); Padel 1972:81-85.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 2
219
PREVIOUS READINGS Southesk
-ROMOOSEFBe+-
Rhys
Angle: DEVODDRE
Face: EHTECONMOR
Rhys
Angle: D+(u/e)V(o/u)DDRS
Face: EHTECONMOR (r/f)(s/v)a+
ECMS
Angle: DeVoDDRe
Face: EHTECONMOR
Nicholson
Angle: ERLLOTEL
Face: EHTE CON MOR
Macalister
Angle: DEV...ODRS
Face: EHTECONMOZS
Padel
Angle: -d+(o/u)V...(d/o)DRs+
Face: -EHTECONMO(r/z)(s/n)
OGHAM-INSCRIBED FRAGMENTS - CUNNINGSBURGH 3
DESCRIPTION OF OBJECT A thin parallelogram-shaped fragment of sandstone, incised with three lines of ogham and otherwise undecorated [see fig.]. Stone:
Micaceous sandstone
Dimensions:
440mm tall x 260mm wide x 50mm thick
Condition:
Severely trimmed, much of intermediate surface lost to lamination.
As with Cunningsburgh 2, so much of this slab has been lost that it is impossible to determine its original form or orientation. Much of the surface has been lost to lamination, but it does not appear to have borne any decoration other than the ogham letters. It seems rather slight to have been part of a monument, unless a lid, end-slab, or some other element in a composite construction. For the purposes of discussion it will be described as if the oghams are running vertically, with the longer two to the left.
DESCRIPTION OF CARVING Three sections of incised ogham letters survive. The three lines are straight and roughly parallel except for the lower end of the middle line, which curves towards the centre. The strokes, which vary in length
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 3
220
from 7-16mm, are shon and widely spaced, giving the letters a squat, stumpy appearance. The individual scores are deep incisions with U-secrion and rounded ends. A pair of dots indicating word-division is preserved in the middle section. I follow Padel in taking each line separately, from the wider end to the narrower:
A
Extant portion of stem-line - 225mm long. The slope of the consonant strokes indicates that the
text should be read as follows [see fig. for key to numbering]: 1
The stem begins at the intersection of the strokes of a small diagonal cross. See discussion below for the possible value of this letter.
2
Three shon strokes to the left of the stem, sloping slightly backwards - T.
3
As above - T. The strokes are unevenly spaced within the letter, but 2 and 3 are sufficiently spaced to make it clear that this is a single group of three rather than 2 + 1.
4
Four strokes across the stem, angled at the mid-point to point forward ( - > ^ - » ) - E.
5
Four shon strokes to the left of the stem, sloping backwards - C.
6
A single long oblique stroke across the stem, sloping very slightly forwards. There follows the right-hand ponion of a parallel stroke. Though rather widely spaced, these are probably a pair of strokes, either a vowel, or a member of the b-aicme. The two could be either a complete letter 0 or G, or the first two strokes of a subsequent letter of the a- or -m-aicmi. Alternatively, they might be two singletons (AA/MM), or a singleton and the first of a group of strokes from a subsequent letter (Ao + /Mg + ), but for linguistic reasons this seems less likely.
Which gives the following reading:
B
xTTEC(o + /g + )[-
Extant ponion of stem-line - 230mm long. The slope of the strokes indicates that the text should
be read as follows [see fig. for key to numbering]: 1
The stem-line meets the fractured edge of the stone in a smooth arc. After a shon gap there is
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 3
221
what may be the faint shadow of a single stroke across the stem, perhaps slightly angled at the mid-point ->-. Padel may be correct in taking this as a natural mark, which would leave a large space before the first letter, left empty perhaps because of the difficulty of writing ogham on the curve of the stem. If the mark is deliberate it would be an angled vowel, either complete (A), or partial (the break occurs where the next stroke would have been). 2
Three strokes to the right of the stem, sloping slightly forwards. The outer two are roughly parallel, though the stem at this point is coming out of the curve, but the middle stroke is more acutely angled. The three lines of lettering on the stone are so fragmentary that it is difficult to judge the precision of the spacing. If Padel is correct and the preceding section is empty, then the three are clearly meant to be taken togedier - V. If however 7 is a letter then 2 may be intended as 1+2, with the sharper slope of the second stroke meant to differentiate the two groups. The slope of the middle stroke may not be significant resulting merely from the problems of the curved stem.
3
A pair of dots, one on either side of the stem, the left-hand one farther back than the other, as if orientated to the curve of the stem. The pair are further apart relative to the stem than the dotpairs on Bressay or Lunnasting. Padel suggests that they are an afterthought, added to clarify the text; the greater separation, however, might be designed merely to make them stand out in contrast to the short letter strokes, thus mirroring in reverse the close dots between the long strokes of Bressay and Lunnasting.
4
Two strokes to the left of the stem, sloping very slightly backward - D.
5
A single short stroke perpendicularly across the stem - A.
6
Three strokes to the left of the stem - T. Only the proximal tips of these strokes are preserved. There follows a generous space, perhaps intended merely to differentiate between two successive letters of the h-aicme.
7
Three strokes to die left of the stem, sloping backward - T. The mark resembling a bind-stroke at the distal tips of the strokes is the result of natural wear.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND 8
CUNNINGSBURGH 3
222
Three strokes to the right of the stem, sloping forward - V. These might have crossed the stem U, but the relevant portion is missing.
9
A single stroke to the right of die stem abuts the broken edge of the stone. The portion to the left of the stem is lost - any letter of the b- or m-aicmi.
Given the predilection for doubled
consonants, it may be that this letter was a second V. Which gives the following reading: -]a+(v/bl) : DATT(v/u)(v/b+/m+)[-
C
Extant portion of stem-line - 175mm long. The slope of letters 2 and 3 indicate that this line is
to be read in the opposite direction to that of A and B [see fig. for key to numbering]: 1
The shadow of a single long stroke, perpendicularly across the stem, is discernible on the surface of the stone - a letter of the a-aicme. The surface at this point has been lost through lamination, it is possible that the stroke was slightly sloped, in which case it might belong to a letter of the m-aicme.
2
Three short oblique strokes to the right of the stem (following the direction of reading), sloping slightly forwards - V. The surface to the left of the stem has been lost, so it is possible, though perhaps unlikely given what follows, that the strokes continued across (- Gw).
3
Three short oblique strokes to right of the stem, sloping slightly forwards. The surface to the left of the stem is intact so nothing has been lost - V.
4
Five oblique strokes across the stem, sloping slightly forwards. The left-hand portion of the first two strokes has laminated off, but the letter is not in doubt - R. Any subsequent letters are lost at the fracture.
Which gives the following reading:
-](a +/m+)VVR[-
FORM OF SCRIPT Over and above the uniquely meandering stem, the most notable feature of the script of Cunningsburgh 3
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 3
223
is the distinctively stumpy appearance of the short and widely-spaced strokes. The letters too, are widely spaced, with up to a stroke's wonh between them. Letters of the b- and h-aicme letters slope consistently in the expected direction. There is one straight vowel (B 5), or perhaps two (A 6). The former occupies only the middle two quarters of the ogham band, the latter, it's full width, but B 5 is in an area of damage and A 6 may in fact be an m-consonant. Like Lunnasting the inscription begins with the x-foifid (it may even be the same word), the most common of all the supplementary letters. Its possible phonetic value in Irish contexts is discussed by Sims-Williams, who concludes that, after an initial period when it signified /g/ [see discussion of Newton], it was used for /eá/ later / s : / , and subsequently as the equivalent of the manuscript digraph ea and diphthong éo [1992:62, see also discussion of Formaston and Buckquoy]. The other supplementary letter in Cunningsburgh 3 is the angled vowel (A 4) which is such a common feature of Shetland oghams (Bressay, Cunningsburgh 2, Lunnasting) but is seen elsewhere in Scotland at Birsay, Burrian, and Formaston [see Formaston for a discussion of angled vowels]. Another feature shared with Bressay and Lunnasting is the pair of dots to show word-division. While this feature indicates a late date, it need not necessarily imply one in the Norse period [see Introduction for discussion]. The transliteration of H and V is problematic, the former is probably a spirant, and the latter HI in a Gaelic context and /w/ in a Brittonic [see Introduction].
There are no other inscriptions with letters quite so squat and widely-spaced as those of Cunningsburgh 3, though both Lunnasting and Burrian are tending in this direction (unlike Cunningsburgh 3, both have bound letters). No other inscription has exactly Cunningsburgh 3 f s combination of letter-types, though it would be dangerous to make an argument based on the non-appearance of certain forjeda in a fragmentary text. Nonetheless this inscription clearly belongs in a group with Bressay, Lunnasting (to which it is quite close), Burrian and Formaston, in other words, the later Scottish oghams.
INTERPRETATION Originally, the three sections of ogham text may have been entirely separate, but the curving stem of the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
CUNNINGSBURGH 3
224
middle section seems to indicate that B and C were joined. Macalister thought the three were pan of one continuous line of ogham [see fig.], though Padel was correct in pointing out that the contrary slope of section B mitigates against his reconstruction. The three segments can, in fact, be interpreted as one continuous line but, in the form of a spiral [see fig.]. This would entail reading them in die sequence A, C, B, with perhaps at least five letters missing between C and B, and an unknown number (depending on the original length of the slab), lost at the other end. If Cunningsburgh dates to the Norse period, such a layout might have been inspired by the common practice of arranging runic texts in spirals, though the circular ogham from Buckquoy shows that already in the eighth century there was flexibility over the alignment of the stem. Since the stem terminates at the intersection of the initial > < , this must be the beginning of the text. The slope of the letters indicates it cannot be the end.
The extant segments are: A: xTTEC(o + /g + )[- C: -](a+/m+)VVR[- B: -]a + (v/bl): DATT(v/u)(v/b+/m+)[Because of the breaks in the stone, there is considerable doubt over a number of individual letters, and, as discussed above, there is uncertainty in the transliteration of certain characters. Since word-division is indicated at one point, we are not at libeny to introduce it elsewhere in the extant sequences. The following are two possible interpretations: Etteco-f ]-qffr-[ ]-af: datturEtteco-f ]-aw(w)r-f ]-aw : dattwfw)The opening sequence may be compared with EHHTECON- of Cunningsburgh 2 and ETTECU- of Lunnasting which would imply a Brittonic interpretation (see discussion above). The remaining fragments of text are too shon to give much clue as to meaning. The final segment calls to mind Bressay's : DATTRR: which has been compared with ON dottir 'daughter1, but insufficient remains at Cunningsburgh 3 on which to base such a reconstruction. Similarly with -ajfr (if indeed that is the correct transcription), which, if at a word boundary, could be a Norse male personal name in the nominative. Cunningsburgh 3's word-final (a+)V should be compared with the end of the third word in the Lunnasting text -EVV.
CATALOGUE OF THE OGHAM INSCRIITIONS OF SCOTLAND
CUNNINGSBURGH 3
225
BIBLIOGRAPHY Goudie 1904.54-5; Macalister 1940.216 pLVlb; RCAHMS 1946 No. 1136(4); Padel 1972:85-8.
PREVIOUS READINGS Macalister
A ETTEC[o+- B -a+]TVVALT-
C -RTTA-
Padel
A ETTEC[a+- B - V : DATT(u/v)[a+/b/m]
C -RTT-
or
-VVR-
THE CUNNINGSBURGH OGHAMS - DISCUSSION The three ogham fragments from Cunningsburgh are markedly unlike one another in several respects. Cunningsburgh 1 may have come from a cross-slab, but fragments 2 and 3 are quite unlike any known monument, except perhaps Lunnasting, which is also carved with only an inscription. The comparison with Lunnasting is all the more important given the similarity in the texts of the three. All contain the sequence Ettec-: Lunnasting and Cunningsburgh 3 have it in text initial position (Cunningsburgh 2 is fragmentary so we cannot tell if any tiling preceded it). Lunnasting and Cunningsburgh 2 both have -s at the end of this word, and Cunningsburgh 3 has word-final ]-(vowel)V in a position in the inscription analogous to Luiinasting's #HCCVVEVV#. Though there is barely enough to go on, it is at least possible that all three have similar texts meaning something like 'this is as far as (the territory of) NN\
The most striking differences are in the forms of script used. Not only are 2 and 3 quite unlike 1 and each other, but both are sui generis in comparison with the corpus as a whole. Fragments 1 and 3 incorporate developed, and therefore probably late, features, including bind-strokes, word-division, and supplementary letters (also short, widely-spaced letters).
Fragment 2, in contrast, has three features which taken
separately might indicate an earlier date, viz. the use of the arris, the close-spacing of long thin letterstrokes, the disparity in length between vowels and consonants. But seen another way, the use of arris and face together, and the unique 'hair-pin' characters, give the inscription an experimental air. The use of the arris at Golspie shows this practice was not restricted to the earliest inscriptions. The feet that all
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
C U N N I N G S B U R G H 1-3
226
Cunningsburgh 2's consonants slope seems at variance with the comparatively short length of the vowelstrokes. In the light of the eccentricity of fragments 1 and 3 one wonders if some of the supposedly 'early* features are in fact merely mannerisms, and the inscription is as late as the other two. It is worth stressing, however, the difficulty of trying to translate script typology into absolute chronology, and the possibility remains that the three Cunningsburgh oghams span a considerable period of time.
With its bind-strokes, Cunningsburgh 1 has letters of average proportions, similar to those of Whiteness, Abemethy or Inchyra. In contrast the long thin strokes of Cunningsburgh 2 are like Pool (and some of the oghams of Co. Cork), and the widely spaced strokes Cunningsburgh 3 are more squat than any others in the corpus. This diversity of ogham hands and styles at Cunningsburgh may reflect inscriptions from different periods, or, given the eccentricity of layout and other features, merely an experimental attitude to the script. It is surely significant that, as at Birsay, the same site has produced both ogham and runic inscriptions. At least one of the Birsay oghams is demonstrably pre-Norse, so a late rune-inspired enthusiasm for ogham is not the only explanation for the Cunningsburgh collection. If the language of the Cunningsburgh texts is correctly interpreted as Pictish this makes a Norse period explanation less likely, but it cannot be ruled out. The 'ethnic-cleansing' model of Norse settlement in Atlantic Scotland can no longer be upheld and the Pictish language may have persisted for several generations. Stevenson has argued cogently for the continuing presence of indigenous Christians in Norse Shetland, and for two-way cultural borrowing [1981].
HU4I26
8
gravsy*rtf
•
Runic i t o n *
o
Oroch I t i t * o f )
findipot
MU472S
CUNNINGSBURGH - Location Map (Turner 1994:316 Illus.l]
CUNNINGSBURGH 1 - Ogham-inscribed fragment [© Trustees of the National Museums of Scotland, Reg. IB 114]
CUNNINGSBURGH 1 -
Ogham-inscribed fragment
Front
Back
Schematic representation of the ogham (key to numbering)
ÉFWSE
CUNNINGSBURGH 2- Ogham-inscribed fragment [° Trustees of the National Museums of Scotland, Reg. IB 115]
(face)
(arris)
4 5
CUNNINGSBURGH 2
Schematic representation of the ogham (key to numbering)
CUNNINGSBURGH 3 - Ogham-inscribed fragment
CUNNINGSBURGH 3 Macalister's reconstruction
>
Alternative reconstruction
-—*N
I
\ I
-. —
\
— i -
1
^
CUNNINGSBURGH 3
Schematic representation of the ogham (key to numbering)
B
3
r i
4
5
6
7
8
9
CUNNINGSBURGH - Runic inscriptions.
j^^aesaEn^
Cunningsburgh 1 [ECMS fig. 14]
Cunningsburgh 3 [ECMS fig. 15]
The Ogham Inscriptions of Scotland: An Edited Corpus
A thesis presented by
Katherine Stuart Forsyth
VOLUME TWO
The Department of Celtic Languages and Literatures
Harvard University Cambridge, Massachusetts
April 1996
227
DUNADD
DISCOVERY The presence of rock carvings at the great citadel of Dunadd in the parish of Kilmichael Glassary, Argyll [NGR NR 837 935] has been known since at least the nineteenth century, but only in the 1950s was the ogham inscription recognised. Writing in 1879, F. W. L Thomas commented that he thought the footprint 'has probably been sheltered by the turf till recently' [1879:31]. He records a local tradition that the mark was the result of a leap made by Ossian from Rudal Hill to Dunadd, from which a footprint was left at each. Thomas goes on to say that Ossian fell onto his knee at Dunadd (presumably explaining the rockbasin) and 'put out his hands to prevent him felling' [36]. I wonder if this is a reference to the finger-like impressions of the inscription. Even if the strokes of the inscription were known locally, they were not recognised as an inscription by the scholarly community until attention was drawn to them in 1953 by Stewart Cruden of the Ancient Monuments Inspectorate. Kenneth Jackson examined it on behalf of the RCAHMS and published his account in 1965. To protect the carvings from vandals and the elements the entire carved surface, except for the rock basin, was covered over with a fibreglass facsimile in 1978. As pan of this operation the Department of the Environment lifted the turf in order to examine the surrounding area but no further carvings were discovered [M. Campbell 1976:6].
SITE The fortress of Dunadd, Dim Ad, 'fort of the river Add' [Watson 1926:45], has been extensively excavated, most recently in the early 1980s, and has produced one of the largest assemblages of artefacts from any early medieval site in Scotland. These finds are now in the NMS. The nucleated fort is situated on an isolated rocky hillock, 54m OD, above the flood plain of the meandering Add, 6km north-nonh-west of Lochgilphead. It commands the boggy land of the surrounding Móine Mhór (Crinan Moss), and controls ponage across the Kintyre peninsula from the head of Loch Crinan to Loch Fyne [Alcock 1988:28]. Dunadd is twice mentioned in the Irish annals (Dun Att)y a siege in 683 and its capture by the Picts in 736,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUNADD
228
which confirms, if confirmation were necessary, its prime importance in the history of Dál Riada [Bannerman 1974:15-6; see RCAHMS 1992:159 for historical note]. Campbell and Lane describe Dunadd as 'one of the primary royal sites of Scottish Dalriada' [1993:52]. The heavily defended site may, in feet, have been the inauguration place of the Dál Riadic kings [F. Thomas 1879].
Construction of the fort may have been begun in the Iron Age, but the main focus of activity at Dunadd was in the early medieval period, primarily the fifth to ninth centuries. The site is comprehensively described in Volume VI of the Argyll Inventory [RCAHMS 1992] and concisely, with colour photographs in Ritchie & Breeze 1991. A detailed discussion of the site is outwith the scope of this work and the reader is directed to the extensive bibliography on Dunadd. Before turning to the rock-carvings and ogham inscriptions I merely allude to the political, military and ritual significance of Dunadd, its economic role in the collection and distribution of tribute, its involvement in long-distance trade, its fostering of craftproduction of the highest calibre and its role as a centre of cultural exchange [E. Campbell 1987; Thomas 1990; Campbell & Lane 1993].
FURTHER INDICATIONS OF THE USE OF LITERACY AT DUNADD The rock carving is the only evidence for knowledge of ogham at Dunadd, but two otherfindspoint to the use of roman letter literacy there. Most unusual was the finding of a small piece of sulphide of arsenic. This Mediterranean mineral, also known as orpiment, was used in manuscript painting to produce a brilliant yellow colour and appears in the Book of Durrow and the Book of Kells, both of which may have been produced on Iona. The extant piece may have been passing through the trading node of Dunadd on its way to the great scriptorium on Iona. The small pebble found at Dunadd inscribed with the letters INOMINE is described below. It may imply the presence at the site of literate clerics. Campbell and Lane have posited a primitive chancery at Dunadd [1993.62] doubtless staffed by monks from Iona [for connections between Dunadd and Iona see E Campbell 1987].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUNADD
229
DUNADD (Roman alphabet inscription) DISCOVERY: This inscribed slate was discovered during the 1904 excavation inside fort D, near the roadway to fort A [Christison 1905,311; see fig.). It was presented to the National Museums in that year by the landowner Col. Malcolm of Poltalloch and is now on display in the Queen Street Museum (Cat No. GP 219).
DESCRIPTION OF OBJECT: A small pebble incised across its smooth face with one line of letters, c.35mm tall [see fig.|. Material:
Pyritous slate, probably Easdale Slate
Dimensions:
Circular, diameter c.4cm
Condition:
complete, uncarved except for lettering
DESCRIPTION OF CARVING: The inscription is complete and reads in clear and confidently formed minuscule letters: INOMINE which has been taken for the latin in nomine 'in the name (of the father...)'. The language and formula imply an ecclesiastical context. Okasha considered the text unlikely to be primary, 'it is more likely to consist of practice letters' [1985.65]. Alternatively the pebble may have had a talismanic significance. To Okasha the archaeological and historical context of the object suggested a seventh to eighth century date, but there is no other dating evidence [65]. Regardless of the precise significance of the inscription it implies a high degree of calligraphic skill at Dunadd, the letters are neatly seriffed and have a strongly 'book-hand' aspect. The informal aspect of the inscription is heightened by the apparent haplography of the N. Given that it was probably the Irish (Gaelic) language which was spoken at Dunadd, might it be that the carver has become confused with the Irish preposition / ?
BIBLIOGRAPHY: Christison et al 1905:307, 311, 321 and fig.; J. Curie 1932:386; Jackson 1960:39, 40-1; Campbell & Sandeman 1964:117, Appendix 34; Jackson 1973:168, 171; Alcock 1981:166-68; Higgitt 1982:308; Okasha 1985:64-65, pl.vii; RCAHMS 1980:156 (repr. inRCAHMS 1992).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUNADD
230
ROCK-CARVINGS Just below the summit, at the point marked 'a' on the plan [see fig.] is a flattish terrace of bare rock riven with fissures and bearing a number of carvings. These consist of a rock-cut basin, below the entrance through the NE wall of the summit fort, and 2m to the north-east spread over a distance of about 3m, a shod foot-print, an incised boar, another foot-print and an ogham inscription [RCAHMS 1992:157]. The first footprint is probably to be associated with modern graffiti of a pipe-smoking head and the words 'King Fergus* [M. Campbell 1976]. The rock basin and footprint were known by the time of the first published description in 1878. The boar came to light during the 1904 excavation but only in the 1950s was the ogham recognised [as discussed above]. The carvings are conventionally dated between the mid-seventh century and, at the latest, the end of the ninth, most probably within the period late-seventh to early-eighth century.
DESCRIPTION OF OGHAM CARVING The Dunadd ogham comprises two lines of text arranged horizontally across the rock-face, one above the other, and roughly parallel with two natural horizontal fissures in the rock [see fig.]. The inscription is largely contained in an area bound by two oblique fissures, though the upper line continues for a few strokes beyond the right-hand fissure. The horizontal arrangement of ogham text is unusual, but not without precedent. The ogham recently discovered on a low-lying rock outcrop in Knockbrack townland, near Castleisland, Kerry, is arranged in a single horizontal line along a natural edge [Fionnbarr Moore, pers. com.]. There is no stem-line carved at Dunadd, but instead the lips of the horizontal fissures serves as arrises. Both are curved rather than crisp, but, while the upper is weathered and indistinct, the lower is fairly well defined. The upper strokes of each line are carved on the horizontal surface of the stone and thus have been more exposed to weathering. The lower strokes, however, which reach down into the fissures, have been shielded to some extent and are better preserved. Whilst drawings and photographs give the impression of a two-dimensional carving, in fact, at close-quarters the inscription is little different from the typical pillar ogham. The strokes are too weathered to determine whether they were cut or
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUNADD
231
pocked. The lower strokes of both lines continue to another ridge inside the fissure, this distance is shorter in the case of the upper line, hence the strokes are not as long as those below. Though little detail survives in the upper line, there is no reason to believe the two were not carved by the same hand.
There are three published readings of the Dunadd ogham inscriptions, all made before the rock-carving was covered with the facsimile. Little weight can be attached to that of Campbell, which seems a long way from the extant carving, and I have not reproduced it. The first estimation of the text was given by Jackson, who described it as 'gibberish as it stands'. He commented disparagingly that 'it might be possible to make it look rather less nonsensical by juggling with the hypothetical stem line in such a way as to produce some sorely needed vowels'. Even widi this treatment, however, to Jackson's eyes the Dunadd ogham would be 'only a little more nonsensical than some of the other Ogams of Scotland, which, though they may have a rather more generous allowance of vowels, are still very odd indeed'. He concluded, 'the whole inscription, in its layout, the form of its letters, and its unintelligibility, is typical of the enigmatic Ogams of Pictland, and not of the perfectly interpretable inscriptions of the Goidelic world' [1965:302],
Jackson's preconceptions about 'unintelligible Picrish' and 'perfectly interpretable Goidelic' ogham are further discussed in the introduction. His conviction that the Dunadd ogham was 'Pictish' and his expectation that it would thus be unintelligible lead to a very unsympathetic reading. An almost obtuse reluctance to read vowels resulted in the following (reading left to right): Upper: HCSD[.]t[..]v[.](N/LV)H[.]T Lower: L[....]VQR(R/I)HMDNHQ Padel was more open-minded, though his debt to his supervisor is clear. To him Dunadd was 'uniquely quite unpronounceable'. He felt that unless some vowels could be 'worked in', it too would have to be discarded as gibberish [89]. His reading differs from Jackson's principally in allowing a number of vowels and, though still difficult and, by his own admission, 'highly conjectural', is a clear improvement: Upper: AESD[-]T[-]V[-](N/LV)A [-]T[V Lower: L [....]VIRR(H/A)MDNA
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUNADD
232
I visited Dunadd in August 1995 by which time the facsimile had been in place for 18 years. Though it is beginning to show its age at certain particularly exposed spots, fibre-glass does not of course 'wear* like stone and thus there is no reason to believe that the carved surface, which is still completely intact, is significantly less clear than when it was first cast in 1978. My reading of the inscription is based on this person inspection and on careful study of the pre-1978 photographs in the NMR. I was informed by the RCHAMS illustrator that the Inventory drawings were made under Prof. Jackson's close supervision, thus, they are not to be taken as an independant witness to the reading.
There can be no doubt that both lines read in the normal direction from left-to-right, as is confirmed by the direction of the slope of individual letters.
Upper Line:
The upper ogham, and especially its upper (horizontal) surface, seems to me very badly
worn and ill-defined. There is little I can add to what Padel has written. Reading from left to right [see fig. for key to numbering]: 1
Five strokes on (I) or above (Q) the hypothetical stem. There is a clear space between the first and second strokes, so it is possible that the group is in feet - 1 + 4 , i.e. A E or H C.
2
Four strokes below the stem - S.
3
Two strokes on (0) or above (D) the stem. Though it is virtually impossible to ascertain the relative position of the stem at this point, the strokes of this group are apparently much higher than those of /. If there is such a distinction it implies that 1 should be taken as a-dicme and 3 as h-aicme.
4
On the original photographs in the NMR there is the slight trace of what may be a stroke on the stem at this point. It is not included on the published RCAHMS diagram, nor mentioned by previous writers. Without seeing the actual rock face it is impossible to be sure. The stroke may be a singleton, in which case - A, or pan of die next group.
5
Three possible strokes tentatively drawn on the RCAHMS diagram. They appear slightly lower
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUN ADD
233
than i , so may be on die stem and therefore vowels, but they are higher than 1. Their full length is not preserved so it is difficult to ascertain their original position. If there are only diree - T or U. If taken with 4 - E (or C), or since there would be room for an intervening stroke, more likely - 1 (or Q). There follows a worn patch with sufficient room for one or two letters. 6
Three faint strokes below the stem - V, or, since the letter may not be complete, any successive letter of the b-aicme. There follows a shon gap which could have contained a one- or two-stroke vowel - A or 0 . Since die next letter belongs to die b-aicme also, the gap may be deliberate, intended to differentiate clearly between two successive letters of the same aictne.
7
Five strokes below the stem - N. These could be grouped 2 + 3 (L V).
8
A single stroke, on or perhaps above the stem - A (or H). Padel thought this stroke probably a vowel.
9
In the gap between 8 and the break (a deep, oblique, natural fissure in the rock surface) there is what appears on the photograph to be a possible further two strokes above or on the line, close by the break. The inscription resumes immediately after the fissure with another three strokes. Since there is ample room for these before the break, the implication must be that something has been lost in the vicinity of P.
10
Three strokes on or above the stem, with possible traces of another on the photograph, - T or C.
11
The traces of a single stroke on (A) or above (H) the stem are just about visible on die photograph. Padel thought these to be the possible remains of a V, Jackson dismissed diem as 'faint marks ... not letters but appear to be accidental scratches '. This may be true, but they appear to be regularly spaced which implies they are indeed letters.
Which gives the following tentative reading: (ae/i/q)S(d/o)(u/at/i/q/e/c)[--]V(/a/o)(n/lv)A(/d/o)[crack](t/ch/v)
Lower Line:
I found the lower line to be better preserved and far more legible. At first glance, my
reading may appear wildly divergent from that of Padel, in fact we are in substantial agreement over the
CATALOGUE OF THE OGHAM INSCRHTIONS OF SCOTLAND
DUN ADD
234
majority of letters: 1
Previous authorities claim to have seen two faint strokes below the stem (L or any successive letter of the b-aicme) to the right of the main body of lettering. I examined the surface of the facsimile with care, but could find no trace of them. For a short distance following this the surface of the rock is uneven. If there ever was any carving in this area, and there is room for about three or four letters, it has been completely obliterated. The roughness may, however, be ancient, and on balance, I prefer to think the inscription begins with the next letter.
2
Three strokes below the stem - V. Since this letter does not begin immediately at the edge of the worn section, but after a short gap, it appears that a complete letter is preserved. The RCAHMS drawing gives the impression that these strokes slope backwards, and this is misleading. In fact, they are more or less vertical. Any slope as the stroke curves down into the fissure, is slight.
3
Five short strokes on the stem - I.
These are substantially shorter than the surrounding
consonants. If there is any doubt that these are vowel-strokes it is completely dispelled by the extent of 7 above the stem. The distance between the third and fourth strokes is very slightly greater than between the others but is insufficient to justify the group being read as two letters (UO). Letters 4 and 5, successive letters of the same aicme are explicitly separated, in a way that these are not. 4
Five oblique strokes below the stem, sloping forward - N. Padel and Jackson considered that these strokes and those of the next letter continued across the stem, and read them as R. While there are possible traces of ribbing in the appropriate position these are far from clear, which is strange given that the adjacent carving on the upper surface (pans of J, 7) is fairly well preserved. In attempting to adjudicate between the two possibilities, the gradient of the strokes is not particularly helpful. The first stroke is at a sharper angle than the others, but all are sloped. This might be thought to differentiate this letter from the two unambiguously b-aicme letters, 2 and 9, however, 4 and 5 are not as steep as 7 (M). Letters 4 and 5 are clearly a pair but the latter is less sloped, scarcely more so than the b-aicme 9. Padel is doubtless correct to suggest that the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUN ADD
235
enhanced slope of 4 is intended to differentiate the two successive letters of the same aicme, perhaps then it should not be relied oa There is a clear gap, at least one stroke's worth, before the next letter 5
A second group of five oblique strokes, more upright than the previous group - N. The portion of the fifth stroke nearest the stem has been lost, but its distal end is clear. Again there are inconclusive marks on the upper surface, but there is definitely no trace of any upper strokes in the area closest to the stem. What appears to be the distal tip of a sixth stroke is probably just a bump in the surface. In this same area Padel saw a short single stroke on (A), or above (H), the stem. Because it was so short Padel thought it might be an off-centre vowel certainly, H would produce a very awkward consonant cluster. I could find no trace of carving in this position, only a large, ill-defined, natural indentation.
6
A single long oblique stroke across the stem, markedly more sloped than any others in the line M.
7
I could find no trace of the two h-aicme strokes claimed for this section by the RCAHMS and Padel. If there is anything, it is a single vowel stroke (A), certainly there is space for it but again the rock has an ill-defined natural indentation.
8
Five short strokes below the stem - N
9
A short single stroke, on the stem - A.
10
Five short strokes, on (I) or above (Q) die stem. Padel was of die opinion that this was a vowel, and certainly their extant length is much less than the earlier consonants of die b-aicme. They do, however, appear to be longer than 9, and the upper portion of 6 is shorter than the lower. These strokes are much more closely-spaced than previous ones (cf. 3) and the impression is that the carver was trying to squeeze them in before the oblique fissure, this may have contributed to their lack of length. This general compression means that the greater distance between the penultimate and final strokes is probably intentional, the last one has only just been fitted in before the crack. While there is still some room for doubt, my prefered reading is CH.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUNADD
236
Which gives: FI(nn/rr)MaNA(ch/q)
FORM OF SCRIPT The Dunadd ogham is written in a simple and fairly unremarkable form of the script. There are neither drawn-in stem-line, nor forfeda. The b- and h-aicme consonants are vertical or slightly sloped in the expected direction, the m-aicme steeply sloped. The vowels are short lines occupying no more than the middle quarter of the ogham band. Though the carving is badly weathered, the component strokes are generally even and consistent. Individual letters are clearly spaced with extra distance placed between successive letters of the same aicme. If the lower line is two words as suggested, then the two words have been separated by additional space. Overall, the script of the Dunadd ogham may be compared to that of Bac Mhic Connain, or more closely to that of Gurness, with which it is virtually identical. If the proposed leading is accepted then it precludes a date earlier than the eighth century. Dunadd thus provides an important indication that simple forms of the script continued in use beyond the seventh century abandonment of the individual inscribed memorial monument-type.
INTERPRETATION There is no indication whether the two lines are to be read together as one continuous text, and if so, in which order. Each line may be an independent sense unit. We would not expect the Dunadd ogham to conform to the 'X MAQQI Y' of the individual inscibed memorial monuments. Unfortunately, since the context is unique, we have no idea what kind of text to expect. It has been assumed that the inscriptions is in some way 'official' and is connected with the other carving. While this is possible, it need not be the case. The carving is neither too 'monumental' nor too 'casual' to preclude either interpretation. Notwithstanding the various claims that this is somehow a 'Pictish' inscription [see below] it seems natural to expect it to be written in the local language of the period, Gaelic. This is my working assumption.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUNADD
237
Upper With no hint at the format and only snatches of the text, our inability to derive anything meaningful from the surviving letters is disappointing, but not remarkable. AES may well be the Olr noun áes, 'folk, people, those who'. The word is almost invariably followed by a qualifying clause, adjective, or noun, in the genitive, either to denote a calling, condition or occupation, as in aisfinan 'the righteous', ais lose 'the blind', aes dam 'poets, artists' [DE], or, when followed by the genitive form of a place-name, to denote inhabitants [Hogan 1910:14]. Insufficient remains to make sense of the remaining sections of text.
Lower
I take FINNMANACH to be two words, Finn Manach. This I interpret as Old Irish, with Finn the personal name, and Manach an adjectival epithet refering either to status or ethnic origin. Finn or Fionn, meaning 'fair, bright, white, light-hued', ( < CC *vindos), was the name of a number of legendary figures, including, of course, the culture hero Finn mac Cumaill. Three legendary Finns were associated with Emain Macha and the name may in origin have been that of a divine figure. It was borne also by historical figures, appearing, for instance in the royal genealogies of Leinster and among the Dál Cais, and continued in use into the later middle ages. The name, though usually, is not exclusively male, but its use as a female name is 'relatively uncommon' [Ó Corráin & Maguire 1981:100]. No historical Finnscanbe linked with Dál Ríada but, interestingly, the name appears in the 'prehistoric' section of the Cenél nGabhráin genealogy. The Genelach Rig nAlban down to Máel-Coluim preserved in Rawlinson B.502 records a Find Féicce m. Achir m. Echdach, eight generations before Ere and twelve before Áedán m. Gabráin [CGH162 d 13].
The onhography of the name is not without dating significance. According to Thurneysen, the assimilation of nd to nn in Irish begins in the archaic period for enclitic words, but first becomes common only in the Milan glosses (late eighth, beginning of ninth century). By the middle Irish period nn is the norm
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUN ADD
238
[1946:93 §15 lc]. Thus Dunadd's FINN cannot be earlier than the eighth century, and is probably no older than the end of that century or even the beginning of the next.
Find was a popular first element in compound names, both male and female [see CGH index], but in this case it appears, rather, that Manach is a sobriquet. There are two possible interpretations. Firstly, manach, ( < Lat. monachus), appears in Irish meaning 'monk', both in the strict sense, and also, by extension, as a legal term for a tenant of church-lands not neccessarily living under a religious rule [DE\. Another word, monach ( < mon 'a feat, esp. of sleight or cunning, a trick'), which is often or usually spelled manach, refers to those who are 'able to perform feats or tricks, dextrous, skilled'. DLL notes its use as a sobriquet meaning, perhaps 'the trickster', and in the plural to refer to trick-riders [s.v. monach]. Thus manach may be an epithet refering to Finn's status, as a monk or monastic tenant, or, refer to his skillful accomplishments.
Alternatively, manach might refer to his place of origin. There were two communities of Monaig or Manaig, one in the west of Co. Down (Ui Echach Ulad) and the other near Lough Erne (Ui Chremthain). The latter, the Fir Manach, gave their name to their district, Fermanagh. According to the genealogies the two were branches of the Ui Bairrche, the tradition being that they had come originally from south Leinster [O'Rahilly 1946:30-33]. It was MacNeill [1919:58] who first made the link with Ptolemy's Manapii, located in Co. Wicklow near Arklow [O'Rahilly 30]. If Padel is correct in reading 3 and 4 as RR rather than NN, then we may, in feet, have the ethnic name Fir Manach, with the doubling of the R reflecting later ogham conventions rather than anything phonological.
The reading Finn Manach, interpreted as 'Finn the monk/monastic tenant', immediately calls to mind the small slab from Clonmacnoise carved +Colman in roman letters, with bocht added in retrograde ogham [CIIC 749]. The bocht refers, not in a general sense to Colman's impoverished or pitiful state, but to his
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUN ADD
239
specific status as a monastic bocht, a pauper or aesedc attached to the community [Ó Cuiv 1986]. The Clonmacnoise slab, however, is commemorative, probably funerary, and it is hard to imagine that the Dunadd ogham had such a purpose. The carving is conceived on a fairly small scale, though still rather grand for graffiti, but it need not be 'official \ especially if it dates to the period after the height of the site's greatness.
DISCUSSION Perhaps the main hindrance to our understanding of the Dunadd ogham is the notion that it is in some way 'Pictish', which despite the dubious logic of the argument, has proved inexplicably tenacious. The idea seems to have originated in Cecil Curie's interpretation of the Dunadd boar as 'the work of a raiding party of victorious Picts* [1940.67], Her suggestion was refined further to link the carving with the capture of Dunadd by Oengus, son of Fergus, (Onuist son of Urguist) king of Picts, in AU 736 [Curie & Henry 1943.261n]. This hypothesis and its extension to include the ogham has gained widespread acceptance and is oft repeated [e.g. Thomas 1964:40; Radford 1953:238-9; Jackson 1965:302 'the only explanation that seems likely*]. Though there have been dissenters [Alcock 1981:167; Ritchie & Ritchie 1981:154], the theory has never been seriously challenged. This unfortunate state of affairs has lead to the cart being put before the horse, and the ogham deployed as an entirely baseless, terminus ante quern for occupation of the site [e.g. Jackson 1965:302].
The traditional argument is as follows: The ogham and boar are 'Pictish'. Their appearance at the very heart of the capital of their enemies must be the result of Picts capturing the fortress. Such provocative graffiti would surely have been intolerable to the Scots who, if able, would have obliterated all trace of it. Since they did not, it follows that the site must have remained unoccupied after the Pictish raid. The latest recorded Pictish raid on Dunadd occurred in AU 736, therefore this must be the date of the carving of inscription and boar, and consequently of the final abandonment of the site. There are, of course, numerous problems with this argument, not least the necessary assumption that there were no subsequent
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUNADD
240
raids on Dunadd which escaped record. For Padel the mid-eighth century was already uncomfortably late for an ogham of this form. To his mind, such a date would make Dunadd not much earlier than the 'scholastic* oghams of Pictland, 'an incongruity that cannot be solved on the present evidence* [Padel 1972:92]. In fact Alan Lane's excavations in the early 1980s have indicated occupation of Dunadd well into the ninth century, which puts paid to the received wisdom on the supposed unacceptability of a flagrantly 'Pictish* ogham and boar. Clearly the inhabitants of Dunadd were not at all bothered by this presence in their midst.
Attempts have been made to get round the apparent difficulty. Following a suggestion by John Higgitt, MacLean proposed as a context for the carvings Pictish mercenaries in the army of the Dalriadan kings [1985:117-8]. There is no direct evidence of such a phenomenon, though it is not inherently improbable and MacLean adduces literary evidence of Pictish mercenaries in British and Irish armies. On stylistic grounds, MacLean dates the boar c. 700-736, his upper limit clearly dependant on the annal reference, though, of course, if the boar is indeed a 'memorial to a Pict who fell in the service of Dál Riada' [118], then its date need not at all be tied to a Pictish capture of the site.
Debate continues over the date and cultural affiliations of die so-called Pictish animal symbols [see recently, Hicks 1993; E Alcock 1991:113-17]. Nonetheless, the similarity between the Dunadd boar and the prize Pictish specimens from Knocknagael and Dores in Inverness-shire is striking. Though doubts have been expressed over its pedigree, MacLean has argued that the differences between the Dunadd boar and the others are a function of its later date. Campbell and Lane have maintain that Dunadd was a major centre of artistic activity, possibly even one of the crucibles in which occurred the fusion of Celtic and Germanic art into the 'Insular' style [1993]. The Pictish heritage of Insular an is well known [see eg. Henderson 1982] and in this context the appearance of a 'Pictish' style boar is perhaps no more surprising than that of moulds for St. Ninian's Isle-style brooches [RCAHMS 1992:155]. That the boar and the ogham are in some way connected is open for debate. Though not at any great
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUN ADD
241
distance from one another they are, in fact, at opposite ends of the rock outcrop, and the carving immediately adjacent to the ogham is the more prominent of the two shod foot-prints. Yet no argument has ever been advanced linking the ogham with the royal inauguration ritual usually given as the context of the foot-print. One could read a unity of design into the various carvings, but there is nothing about their relative positions or the technique with which they were carved to either prove or disprove their contemporaneity. The modern graffiti and legend 'King Fergus* underlines the danger of simplistically presuming a connection between ogham and boar [M. Campbell 1976]. In the context of Mrs. Curie's suggestion about the Pictish raiding party, Jackson was content to declare the ogham inscription 'Pictish'. Its illegibility being apparently diagnostic. Padel, on the other hand, identified an 'Irish character' to the Dunadd ogham, which made some sense to him, given that 'a 'Pictish1 inscription in the west is strange anyway' [92]. The less than obvious conclusion he drew from his observation was that, if the text were carved by invaders, 'they might have had to find a west coast, i.e. Irish oghamist to do it for them', though he thought' a difference i n date seems just as likely' [92]. In my opinion there can be absolutely no grounds for describing the Dunadd ogham 'Pictish'. It is physically rooted in Dál Ríada, it is written in a form of the ogham script common throughout all the areas in which the alphabet was used. Even its unusual location on a rocky outcrop now has its parallel in Ireland. According to my interpretation, the text is in Gaelic.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUNADD
242
BIBLIOGRAPHY
Jackson 1965; Padel 1972:89-92; RCAHMS 1988:158-9 No.248 (repr. inRCAHMS 1992).
Site only:
F. Thomas 1879; Christison & Anderson 1905:292-322; Craw 1930:111-27, 111 pl.xv b; M. Campbell 1964; Lane 1980, 1981, 1984; RCAHMS 1988:149-59 no.248; Alcock 1988; Ritchie & Breeze 1991:2,18,19-22 (ills.).; Campbell & Lane 1993.
Rock Carving:
Curie & Henry 1943:261n; Radford 1953:238; Thomas 1964:40; M. Campbell 1976; MacLean 1985:79, 113-8; RCAHMS 1988:157-9 No.248 (repr. in RCAHMS 1992).
:?;?.
,1
pas
DUNADD - Plan of fort [CampbeU & Lane 1993.53] Rock carvings, including ogham inscription^ location a
DUNADD - Inscribed pebble (enlarged) [Okasha 1985 pi. VIII]
^nowmé DUNADD - Detail of Inscription
7/-t(^,.Áfp.vl >u
i * 41
DUNADD - Incised rock surface with detail of boar [RCAHMS 1988:158]
DUNADD - Incised boar [RCAHMS 1988:157]
DUN ADD - Incised rock surface with detail of ogham inscription
[RCAHMS 1988:159]
/
X.
I H |, •
ilU
/
/
-IJli
^/////iiiiir\n,Xj_ [RCAHMS 1988:158]
A..
.al'i..
;^a
m
mm
6S&
'Upper'
7" iHLuu. 1 2 'Lower'
_ 1
3
5
ff f ff
mr "'///// ///// / 2
3
4
5
6
f^ri 7
8
9
DUNADD - Ogham inscription with key to numbering
10
243
DUPPLIN
DISCOVERY The Dupplin Cross stands by Bankhead Farm on the Dupplin estate, in the parish of Foneviot, Penhshire (NGR NO 0505 1897). From its position just below the summit of the Gask Ridge, at an elevation of 85m OD, it commands an imposing view south across lower Strathearn and including the village of Foneviot, less than a mile away. The stone may or may not be in its original location. Traces of modern monar in its socket show that it was re-erected in the nineteenth century, though confirmation of whether or not it was moved prior to this will have to await archaeological investigation of the plinth. There is, of course, no basis to the persistent tradition, [e.g. Andrew Lang (1929) History of Scotland VoLl. p. 244, quoted on OS record card], that the cross was erected to commemorate the viaory of Balliol over the Earl of Mar at the battle of Dupplin in August 1332. The cross is a scheduled monument in the care of Historic Scotland. In the midst of much controversy, a public enquiry was held into the future care of the monument at Penh in November 1995. At the time of writing, the Secretary of State for Scotland has not yet announced his decision, though it seems likely that the cross will be moved, either to the NMS or, perhaps more likely, inside the church at nearby Dunning. It is hoped that this removal will allow further study of the monument, in panicular providing the first opponunity for a thorough examination of the ogham-inscribed base.
The earliest mention of the Dupplin cross comes in Thomas Pennant's account of his tour of Scotland in 1769 [Pennant 1772] (it is called the 'Bankhead cross' in the OSA of 1798 [Vol. 20 p. 124]). In Pennant's illustration, the cross is listing severely, as it still was in the early nineteenth century according to a drawing in the NMR [reproduced Walker & Ritchie 1987:134]. Intervention sometime before 1857 when Stuart published his drawings rectified this situation and the cross now stands true to the venical. The first detailed illustrations were those of Romilly Allen [see fig.], but his figures are uncharacteristically deficient. The general level of detail is poor and he omitted most of the interlace on the front face of the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
244
cross-head, which is in fact still relatively clear, and one of the animals in the lowest panel on the same face. The inadequacy of the figures in ECMS is all the more serious since no other drawings have been published since. For a monument of such artistic importance, Dupplin has suffered surprising neglect from modern scholars. It is mentioned in passing in a number of studies [e.g. Henderson 1983:243; 1986:89, 90, 102], but has received no thorough study in its own right (but see now Forsyth 1996). An initial attempt to interpret the iconography of the cross and its relation to the royal site of Forteviot was made by Alcock and Alcock [1982], further developed in the report of their excavations at Forteviot [1992].
Despite the prominence of the monument in the landscape, only very recently was the presence of two inscriptions recognized. A very high-quality fibre-glass cast of the cross was made by the NMS in 1990 for inclusion in their exhibition The Work of Angels: Masterpieces of Celtic Metalwork, 6th-9th centuries AD'. At the exhibition preview the sharp definition of the cast and the strong directional lighting of the display led curator Dr Alison Sheridan to notice the ogham inscription carved on the cross's base. Attention was focussed also on the worn panel at the top of the front face of the shaft. In his illustration, Romilly Allen had left this panel blank, while Stuart rendered the carving as simple interlace [1856 pi. 58]. Intensive study by the author and Dr Michael Spearman of the NMS confirmed that the carving was indeed an inscription in insular letters, a discovery referred to by Alcock and Alcock in their article on the Forteviot complex [1992:282-3], and published in full by the author [1996] (see below).
LOCALITY There was a probably Bronze Age burial cairn a distance to the north on the summit of the hill behind the Dupplin Cross [Stuart 1856:17], but no other archaeological remains have been discovered in its immediate vicinity. Its context, however, is clearly the remarkable focus of remains at Forteviot in the valley below. The archaeological, sculptural, and historical evidence for the Pictish and Picto-Scottish royal centre of Forteviot has been ably summarized by Alcock and Alcock [1982; 1992:218-42] and will not be rehearsed here. There is a growing consensus that in the seventh or eighth century a royal centre was established
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
245
in an area already imbued with prehistoric ritual significance [on the analogous re-use of prehistoric monuments in early medieval Ireland and Northumbria see Aitchison 1994 and Bradley 1987]. Associated with the secular settlement was a Christian cemetery of dug graves and square- and penannular-ditched barrows (a distinctively Pictish type) and a rectangular enclosure, possibly a church, reflected today in crop-marks east of the modern village [see map]. The tiny bell, still kept in the village church, dates to Pictish period [Bourke 1983, Purser 1992:39], as, no doubt, do some of the sculptured stones along side it [see fig.]. This royal and ecclesiastical complex, which may well have been the caput of the Pictish kingdom of Fortrenn, was taken over by the ascendent Gaelic dynasty 'as a going concern' in the mid-ninth century [Alcock & Alcock 1992:238]. According to the Scottish Chronicle, Cinaed mac Alpin died in 858 inpalacio Fothiurtabaicht, but attempts to locate the structure at its traditional site on Haly Hill proved fruitless [Alcock & Alcock 1992:228-34].
It has not yet proved possible to assign with confidence the extant sculpture at Foneviot specifically to the late Pictish period rather than the early Picto-Scottish period, but either way, the remaining fragments attest to a remarkable degree of sculptural patronage in the ninth century. The most impressive item is the famous Foneviot arch, recovered from the Water of May and now in the NMS. This architectural feature is considered 'sufficient evidence for the erection of a masonry church above the May in the early decades of the Picto-Scottish kingdom, that is the middle or later decades of the ninth century' [Alcock & Alcock 1992:236] (though I feel a date in the late Pictish period cannot be ruled out). Alcock and Alcock have pointed to the prominence of the cross and paschal-lamb on the arch, an iconography particularly appropriate to Easter (and, one might add, further reflected on the Dupplin Cross). This they tentatively suggest, may indicate that Foneviot was where the king 'was accustomed to celebrate the major festival of the Christian year' [ibid.]. Other items include a fragment of a free-standing ring-headed cross, and fragments of at least two upright cross-slabs.
These remnants provide a technical, artistic, and
iconographical context for the Dupplin Cross. The position of the cross relative to the main site has not yet been satisfactorily explained (nor has it been either proved or disproved to be in its precise original
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
location).
DUPPLIN
246
Where it stands it may have marked the extent of the protected space around the
ecclesiastical/royal centre, or perhaps a significant point on the route towards the site, but it must be considered in connection with the similarly massive Invermay Cross (a.k.a. the Dronachy Cross) which formerly stood overlooking the complex a similar distance to the south (NGR NO 166059). Its huge base, badly damaged when the cross was destroyed 'not many years before 1772* [Stuart 1856:17], still stands in (or near) situ, but only three other fragments survive (preserved in Foneviot Church), the remainder having been thrown away in the nineteenth century.
DESCRIPTION OF OBJECT A free-standing, ring-less cross in a base, sculptured in shallow relief on four faces with figurative, animal and geometric ornament. Incised with a text in ogham on the base and one in the roman alphabet on a panel on the shaft. Stone:
Old Red Sandstone
Dimensions:
Height-VV (2.62m); Width - 3' 1" (0.94m) across arms, 1'6" (0.46m) at bottom of shaft, V 3 xh" (0.40m) at top of shaft, 9lA" (0.55m) at apex of cross; Thickness - I T (0.33m) at bottom of shaft, T (0.18m) at apex of cross (measurements as given in ECMS with approximate metric equivalents following in brackets)
Condition:
Carving comparatively well preserved, though both inscriptions severely weathered.
Unlike many of the free-standing crosses from southern and western Scotland, which are composite monuments employing construction techniques ultimately derived from carpentry, the Dupplin Cross is carved from a single block of sandstone. It stands in its base a little over 2.5m high. Free-standing crosses are rare in eastern Scotland and the distinctive shape of Dupplin is unique (cf. Romilly Allen's 105A, Cramp 1984 Dll). The form of its cross-head and the distinction between cross and pillar is essentially Anglo-Saxon [Kelly 1993:223], but the elongated top arm and tegulated 'cap-stone' effect recall
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
247
cenain crosses of the Irish midlands. Further Anglo-Saxon influence is displayed in the use of vine-scroll ornament all-over the cross-head [Henderson 1983:266]. The figurative panels incorporate Paschal and David iconography [Henderson 1986] and appear to be making statements about kingship and its divine and material sanctions [Alcock & Alcock 1993:238-41]. Despite the eclecticism of its artistic and iconographic influences, and the inspired interplay between Irish and Anglo-Saxon elements, what is striking about Dupplin is the lack of recognizably 'Pictish' features. Despite this, Henderson suggests a date in the early ninth century, which appears to be supponed by recent work on the inscription (see below).
DESCRIPTION OF CARVING - DECORATION The decoration of the cross is carved in a flat but substantial false relief but the central bosses, especially the one on the front, stand proud several centimetres [see figs.]. The lettering is more shallowly cut using an incised technique. The base has suffered fairly severe weathering but is apparently undecorated except for the incised ogham inscription (see below). The labels 'front' (west) and 'back* (east) are assigned on the basis that the front boss is by far the more prominent, the inscription is on the front, as is the more ovenly ecclesiastical imagery (the paschal lamb, David and the bear), while the explicitly martial imagery is reserved for the back.
On each face the shaft is organized into three panels of varying proponions each panel outlined with a plain narrow moulding. The four top panels are divided from the four middle panels by a horizontal band of geometric ornament (according to Allen: front - square key pattern No.899; south side - diagonal keypattern No. 926; back - key-pattern No. 887; north side - square key-pattern No. 890). Because of the differing proponions of the panels on each side these bands are neither continuous nor level.
The three squarish panels on thefrontare roughly equal in size, the lowest being slightly larger, and the highest slightly smaller, than the middle. The bottom panel contains an eclectic jumble - the paschal lamb
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
D U PPL I N
248
and flag, David rending the jaws of a bear, and two other animals one apparently carrying a pole or flag. The middle panel consists of a central ring containing a circular interlace knot (Allen's No. 789) surrounded by four pairs of birds, legs and beaks crossed and interlaced. At the top is the text panel discussed below. The top panel on the reverse of the shaft, which contains a single mounted warrior, seems to parallel the text panel, and it is tempting to imagine it contains a representation of the king mentioned in the inscription. In the elongated middle panel are four foot-soldiers wearing tunics and headgear, armed with round shields, and spears. A thin border of key pattern (No. 890) divides this panel from the possible remains of a hunt scene with pouncing dog below. Two foot-soldiers, similar but more embellished than those on the back, occupy the middle panel of the south side. Above them is a pair of opposing beasts on hind legs, fore-legs intertwined 'sitting on their haunches facing each other with their paws crossed over on each other's necks' [ECMS], and below them a simple triquetra knot. On the north side the top panel is taken up with a beast biting its own tail which is interlaced beneath its body. The bottom has an interlace knot of six-cord plait with a single break in the middle, and in the middle is a remarkably well-preserved figure, possibly David, seated in an ornate chair, singing and playing a large harp.
The cross-head is separated from the shaft by a thick border around all four faces above which there is a marked step in of the silhouette. The edges of the head are surrounded by roll-moulding (single spiral curves at each of the points of the cusped hollows between the arms). The four 'arms' (top, bottom, right, left) are of equal size but the top panel has been elongated by the addition of a further panel. Above this is a badly weathered portion with traces of what might have been tegulation, recalling the top of Irish High Crosses, themselves skeumoiphs of house-shaped shrines. On each face the cross-head forms a single field of ornament. On the back are two stems of vine-scroll foliage, arranged in simple spirals and interlace, on the front interlace knots in a cruciform pattern around the central boss, and beyond that on each arm, spirals (no. 1054). In the centre of the cross on both faces there is a large circular boss surrounded by a sort of ribbed border. Each side has two panels above the arm, a panel on the end of the arm, one under
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
249
the arm, and one in the section above the top of the shaft. These are filled with interlace and square key patterns, and with single beasts [for further details see ECMS].
ICONOGRAPHY A full iconographical study of the Dupplin Cross has not yet been attempted, but Alcock and Alcock are not wide of the mark in declaring that 'all the figural - as opposed to the decorative - panels are statements about kingship' [1993:238]. They point out that the principal figure, the single horseman, stand, not in the dynamic trotting posture typical of Class II Pictish cross-slabs, but rather with all four of the horse's feet on the ground. This, they say, 'is not a naturalistic image, but a symbol of royal permanence' [240]. The Alcocks have argued that the Cross is a political statement about the take-over of the Pictish overkingship by Cinaed mac Alpin [283], a position surely undermined by the recent interpretation of pan of the roman alphabet inscription (see below). This need not, however, affect their more general point about an imagery of royal power derived perhaps from contemporary Frankish models and, more remotely, from the iconography of Roman Imperial triumphs [240]. Henderson has identified at Dupplin two scenes from the David cycle of Christian iconography [1986], David the harpist (I Samuel XVI 16, 23), and David wrestling with the bear (I Samuel XVII 34-7). Whereas elsewhere in Pictish sculpture (Kinneddar, St. Andrews, etc.) the creature is clearly a lion, the Alcocks are surely right to identify the Dupplin beast as a bear [238], an observation which may have implications for the Cross's iconographic relationship to other Insular examples of the David cycle. David, of course, is important as an ancestor and pre-figurer of Christ, but in this context is more strongly an icon of the divinely sanctioned king [Henderson op. cit.]. ~ In its totality, the Dupplin Cross constitutes a powerful statement about royal power, and both its divine and material sanctions, a statement of particular resonance in sight of Forteviot, the focus of both secular and ecclesiastical might.
THE OGHAM INSCRIPTION The Dupplin ogham is not visible on the original monument as it currently stands because back-filled soil
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
250
is now level with the top of the base (the dimensions of which are not yet known). The surrounding earth was pulled back to a depth of a several centimetres in the course of the making of the cast, thereby revealing the upper portion of the ogham. It is not clear from the cast how much further down the inscription extended. Towards the lower edge of the cast the strokes are becoming clearer, which suggests that below this point, where the carving has been better protected, the text may be more fully preserved. I am most grateful to Dr Michael Spearman for facilitating my examination of the cast and subsequent inspection of the original in the field.
If there was ever any carving on the horizontal upper surface of the base, this has been worn away. A few incised lines have, however, survived on the venical faces. Those on the left-hand venical and upper margins of the front/west face are the clearest [see fig.]. A horizontal line runs parallel to the arris at a distance of about 50mm and between the two are several groups of perpendicular and oblique strokes, the ogham letters. The adjacent south face exhibits a lesser number of strokes towards its upper margin, and faint traces of what may be farther strokes are apparent on the east and nonh faces also. It is possible that the text originally extended round all four sides of the base, which would make it one of the longer Scottish oghams, though on the evidence of the cast alone, it is impossible to be cenain. It is unclear also how many of the venical arrises were utilized.
The most pressing problem is to determine the role of the horizontal incised line. Virtually all ogham inscriptions of this period have a drawn-in stem-line, and it would be natural to interpret the horizontal as such. There do not, however, appear to be any strokes to the right of/below the line, nor does it appear that any have been lost through wear. On reflection, the horizontal lines appears more likely to mark, not the mid-point of the letters, but their lower edge (with the letters depending on a notional stem-line along the arris). Traces of a companion line survive on the upper surface and these are perhaps best taken as intended to suggest a broad corner moulding, an angled band to bear the ogham, cf. Golspie's deeply incised perimeter line delimiting the ogham-inscribed area on the arris. An indication of how such a
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
251
feature might have been arranged on a three-dimensional block is provided by the moulding on the High Cross at Arboe, Co. Tyrone [Harbison 1992, figs.30, 36].
The cast has confirmed die presence of an extensive and monumental ogham inscription along at least two and possibly all four sides of the base, but beyond that little can be shown for certain since only disjointed letters or parts of letters have been uncovered so far. The relative position of stem and strokes is crucial to interpreting the letters and until this matter is decided it is impossible to attempt a reading of the text. The only way to progress study of the ogham is by excavating the area round the base and revealing the unexposed portion of the inscription. On the evidence of the cast, it seems likely that sufficient will have been preserved to ascertain beyond doubt the extent and layout of the text. Once the beginning, end, and direction of the inscription have been established, and most important of all, the stem-line been identified, it should be possible to establish the typology of the lettering, and ultimately to interpret at least a portion of the text.
For now, sadly, the Dupplin ogham is too fragmentary for linguistic analysis of any sort, not even the language of the text can be identified. The mere presence of ogham at Dupplin, however, is of great significance.
The carving of such an inscription on a cross base is completely unparalleled, and a
horizontal alignment, though not unique, is rare in stone. This should not, however, cause undue anxiety, since die Scottish ogham corpus is characterized by its diversity and inventiveness, and the Dupplin artists were innovative in many other ways. The carved ogham strokes of Dupplin are substantial and regular, they are not a casual graffito, and there is no reason to doubt that they are part of the original schema. Nor is there a problem with the two scripts appearing together. The juxtaposition of both ogham and roman lettering is paralleled on die Newton stone and on the recently discovered fragment from Lochgoilhead. Both scripts appear at Dunadd and at St. Ninian's Isle, though on separate objects.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
252
The motivation for using two scripts at Dupplin is not clear. There is no need to assume that the two texts are transliterations of each other. In fact, the disparity in length seems to prove the contrary. The choice of the vernacular for the base inscription may have entailed the use of ogham, but the reason behind such bilingualism is not clear either. The lower text may have had a different kind of message, or a different intended audience. Alternatively, the choice may have resulted from aesthetic or technical considerations. Alcock and Alcock have suggested that ogham was used for its Irish associations, but their assertion that 'its symbolic reference to the coming of the Scots is plain* [1993:238] seems overly dependent on their mid-ninth century date for the cross and their determination to link it to Cinaed mac Alpfn. The Dupplin ogham was carved several centuries after the initial introduction of the script and it is debateable whether specifically Irish associations, if any, would have persisted this long. In fact, the use of ogham was so thoroughly integrated into the Pictish sculptural tradition that, by the time Dupplin was carved, it may have been the book-hand letters of the shaft inscription which appeared 'Irish*.
BIBLIOGRAPHY OS Record card - NO 01 NE 6 Pennant 1772; Stuart 1856:17 pl.57-8; ECMS 319-21, fig.334; Curie 1940:107-9 fig. 16; Henderson, 1983:243 No.ll, 1986:89, 90, 102; Alcock & Alcock 1992:5, 227, 238-42, 282-83; Forsyth 1996.
THE ROMAN ALPHABET INSCRIPTION1 The inscribed panel is on the west face of the cross at the top of the shaft, in other words, in prime position. It is extremely worn, but this appears to be the result of natural weathering rather than deliberate defacement. All the primary work on the text panel was done on the NMS cast, but results were later checked against the original in the field. The relative comfort, convenience and accessibility of the NMS
1
The following is a modified excerpt from my article, 'The Inscriptions on the Dupplin Cross', in From the Isles of the North: Medieval Art In Ireland and Britain (Proceedings of the Jlxird International Conference on Insular Art), ed. Cormac Bourke (Belfast: HMSO, 1996), 237-44.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
253
Granton store, where the cast is housed, were a great boon. More importantly, however, a cast, unlike the original, could sustain the application, removal, and re-application of chalk and charcoal without risk of damage. The ability to mark the surface in this way was crucial because what could be discerned was not yet distinct letters but rather a confusing mass of ridges and furrows [see fig.]. It took quite a while, for instance, to be able to work out even how many lines of text there were. It would have been impossible to hold these disjointed lumps and bumps in the memory until they cohered as letters. The technique developed was a painstaking one. It involved feeling the undulations of the stone inch by inch, chalking-in the depressions which were the components of the letters, gradually building up the picture, until eventually everything else was accounted for and only the ridges between the letters (in other words the original surface) remained.
After many hours of work, spread over several visits, it became clear that the square panel, 290x295mm, delineated by a simple band, is inscribed with a mixture of minuscule and majuscule letters, arranged in seven registers of continuous text. There are no rule lines between the registers, nor do tliere appear to be any medial points or crosses dividing sections of text. The number of letters per line is approximately eight or nine, which gives a total of just over sixty characters. The separate letters, however, are not always distinct, and some of the characters may be ligatures. Nonetheless this makes Dupplin the longest Roman alphabet inscription from early medieval Scotland.
READING AND INTERPRETATION OF TEXT The clearest section is at the beginning of the second line. The first six letters comprise the only completely unambiguous word in the whole inscription: Jz/z/tf [see fig.]. This felicitous survival indicates not only that the language of the text is Latin, but also alerts us that we should expect the preceding and succeeding words to be personal names. The name immediately following it, can, in fact, be made out with comparative ease: Fircus, with the s the first letter of the third line. This is a reflex of the Celtic male personal-name *Uor-gustus, 'chosen one1, which is attested in the P-Recension of the King-List in
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
254
more Pictish guise as Uurguist [Anderson 1973:263; Jackson 1955:140, 142, 165, 161, 163]. Dupplin's /rather than w, and lack of final r, shows it to be closer to the Gaelic Forggus, which appears in 'purely Old Irish form' as Forcus [O'Rahilly 1946:170] on the Drosten Stone at St. Vigeans, Arbroath [Okasha 1985:59-61]. The / in the first syllable is difficult to account for. In Irish, Forcus was conflated ultimately with the more common Fergus, reflex of Celtic *uiro-gustus 'chosen man' [O'Rahilly 1946:368 n.3]. John Koch has posited a similar confusion in Pictish whereby Celtic * Wo- and *Uor- fell together as Uur-y usually interpreted in male names as' man' [pers. comm.]. He would thus interpret the / in Fircus as a hypercorrection, an attempt at a more 'correct' and conservative spelling of the Celtic element *uiro'man' (the preservation of the c (/g/) in the second element indicates that it was not subject to lenition, and therefore must be from original -rg-, hence Firgas is indeed a reflex of *Uorgustns]. If Prof. Koch's interpretation is accepted, then the contaminated Dupplin spelling provides an insight into the thinking of the Gaelicizing orthographer, who appears to be operating to a certain extent independently of the orthographic traditions of both the Drosten stone and the P-Recension King-List.
One worry widi the reading Firms is the apparent lack of a genitive ending: we should expect Fircussu. Although the third line begins sfjuy the second letter is clearly different from the preceding s, and appears instead, identical with the penultimate character on the bottom line. In both cases, the minim supports not one but two short strokes, one horizontal, the other oblique, and thus the letter is closer to an/. The lack of a genitive is not an insurmountable problem since Koch has argued that Pictish lacked declension [1983:220], but a satisfactory resolution of the problem must await the elucidation of the rest of the line.
Returning to the top line we would expect it to be taken up with the name of the son of Fircus. The first and last letters are the clearest, cu[~]ntiny the illegible portion comprising three or four characters. There may be -st- in the middle but the rest is rather confused [see fig.4.1]. Notwithstanding the lacuna, sufficient remains to justify the reading Gistantin or Custentin, a version of Constantinus, the male
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
255
personal name borrowed from Latin into all the Celtic vernaculars [Jackson 1953:86,595]. Custantin is an Irish form, attested in the Annals of Ulster, s.a. 819, in the death notice of the individual in question, if correctly identified (see below). It occurs also in Pictish Regnal List B [Bodl. MS Laud 610, Anderson 1973:263] referring to a later king, alongside the forms Coustantin and Constantin. In the equivalent positions, Regnal List A [Poppleton MS, Anderson 1973:249] has Castcmtin and Constantini (gen.). On the basis of these, and forms such as Alpin (Welsh Elffin\ Jackson concluded that Pictish had not participated in the Brittonic sound-change of vowel affection [1955:162]. In all the Neo-Brittonic languages the result of internal i-affection of a before long / was e (i.e. -aCt- > -eC-), hence the Old Breton Custentin [Cartulary of Redon, AD 869, Jackson 1953:595]. It would perhaps be surprising if Pictish had not followed the other Brittonic languages in this development, and Jackson does concede that the form Elpin is attested in Pictish [1955:162]. The Castcmtin of the Regnal Lists may not, however, be conclusive on this question since the influence of the Latin equivalent cannot be ruled out, nor can a touch of Gaelic orthographic influence (indicated elsewhere, perhaps, by the use of / to indicate palatalization [Jackson 1955:165]). If we could determine that the fifth letter in the first line of the Dupplin text was an e then we would have evidence for i-affection in Pictish. If, however, it turned out to be an a, this would not be conclusive either way, since, as indicated above, Gaelic influence is already indicated in the spelling Firms. Unfortunately the relevant section is very worn. Both Cornish and Breton preserved the cluster-/;/- [Jackson 1953:501], as in Old Cornish Custentin [Bodmin Manumissions c. AD 1000, Stokes 1871]. In Welsh the stop was nasalized, with the result that already by the early ninth century -nt- had become -nnh- [Jackson 1953:506], hence the Old Welsh spelling Custennhin [Book of Llandaff, Davies 1979:92]. Even if one were to expect Pictish to follow Welsh in this development, rather than Cornish and Breton, the apparent retention of nt at Dupplin cannot necessarily be taken as an indication of date, since Welsh nt continues to appear for several centuries in 'traditional and archaicising spellings' [Jackson 1953:506].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
256
As for the rest of the inscription, its interpretation resembles, at times, a game of 'Hangman'. We can discern about forty-two more characters, some of which can be tentatively identified. On analogy with the Latin texts of other Insular crosses, we might expect the principal name to be followed by a title, e.g. rex, or by a verb meaning 'put', 'made', or 'raised' [Higgitt 1986; Okasha 1990].
However, nothing
immediately suggests itself, nor are any of the standard formulas, such as hanc crucem, pro anima eius, and in nomine dei, instantly apparent. It is always possible, of course, that the filius is a red herring and the language of the main text (other than names) is not Latin. Since the orthography appears to be at least partially Gaelicized we might expect the language to be Irish, but none of the common Irish formulae, oroit do, do rigni in chrois etc., can be discerned [Macalister 1949], The Welsh monuments, of course, are written exclusively in Latin, so there are no further Brittonic language comparisons to be drawn (with the important exception of the Towyn inscription) [Nash-Williams 1950:172, No. 287]. If the text is in Pictish then we are reduced to a single parallel. The only other intact and unambiguous example of the Roman alphabet used to write Pictish is the famous Drosten stone from St. Vigeans, the text of which consists of a list of possibly connected personal names, also in Gaelic orthography [Clancy 1993]. In the final line, the Dupplin text emerges from the gloom and a couple of letters can be made out. These, however, are rather perplexing, for they seem to be efg, possibly even cdefg, though the alternative interpretation est cannot be ruled out at this stage [see fig.].
Certain portions of the Dupplin text, especially around the edges, are clearer than others, in the middle, however, very little has survived. There are a couple of reasonably distinct letters in the third row, the final row, and at various points in-between. The markedly worse condition of the central portion is somewhat baffling.
There may be an underlying flaw in the surface of the stone, resulting in the
differential weathering of areas of weakness. Alternatively, human agency, either in the Early Historic period, or more recently, may be to blame. Dómhnall Ó Murchadha has identified several examples of what he believes to be the deliberate defacement of inscriptions on Irish high crosses, including the West Cross at Durrow [Ó Murchadha & Ó Murchu 1988:55], the 'battered panels' of the Cross of Scriptures,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
257
Clonmacnois [Ó Murchadha 1980:48], and the panel at the base of the west face of the shaft of the South Cross, Clonmacnois, which he believes was 'uniquely singled out for mutilation at some later point in its history1 [Ó Murchadha & Ó Murchu 1988:53].
Lines 1, 2, 3, and 7 of the Dupplin text are fairly regular and straight, the letters are extremely worn, but comparatively distinct. In contrast, the middle three lines, 4, 5, and 6, seem crushed and confused, and significantly less substantial. Dr. Michael Spearman has raised the interesting possibility that this middle section is actually a palimpsest, and the reason it is so confused is that behind the worn letters are the feint background traces of earlier letters partially effaced [pers. comm.]. When the panel was less worn there may have been more of a distinction between the two layers, and if the cross was painted there would be no problem with legibility. While it is possible that this portion of text might have been re-cut at the outset because of an error in the initial carving, more likely, perhaps, is that it was altered at a later date to reflect the changing political scene. An undulation in the panel can be discerned in this area, which may indicate that the original surface was indeed cut back, though it is always possible that this is simply a natural variation in the quarried surface.
HISTORICAL CONTEXT - CUSTANTIN FILIUS FIRCUS What about the portion of text that we can interpret ? Who was CustantinfiliusFircas, and what was his involvement with the erection of the Dupplin cross ? The first question is answered easily enough. According to the documentary sources, Custantin mac Forcussa was the first individual to hold the kingdoms of both Dál Ríada and Pictavia simultaneously. From the Irish Annals it appears he won the kingship of the Picts after a great battle in 789, gained power in Dál Ríada in 811, and thus ruled both concurrently for nine years, until his death in 820 [Anderson 1973:174,192-4]. Custantin/Constantine was the first of nine kings to be drawn from three generations of one family which held the kingship for over fifty years and lost power only with the ascent of the Meic Alpin. Very little is known about him beyond the Irish chronicle entries for his accession and death. A note in the Q-Recension of the Pictish king-list
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
258
credits him with founding Dunkeld [Anderson 1973:194, 266, 273], an association worthy of further investigation, despite the lateness of this particular source [Macquarrie 1992; Spearman 1993].
The identification of the person named in the text as the illustrious king Constantine is not in doubt, but what was the nature of his involvement with the Dupplin cross ? From the prominence given his name, filling the whole of the top line, he is clearly the person honoured, but in the absence of an understanding of the rest of the inscription it is unclear if this is a contemporary or posthumous invocation. An history is as yet unable to provide an unequivocable solution. Although a broad consensus has emergedfora date at some point in the ninth century, opinions vary as to whether Dupplin is to be placed in the late [Curie & Henry 1943], the mid-to-late [Stevenson 1959:55], or the early part of that century. Most recently, Isabel Henderson has adduced close sculptural parallels from northern England in support of a date in the first half of the ninth century [pers. comm.]
Cecil Curie linked Dupplin to other members of the Forteviot 'school* of carving, which in her opinion, 'represents the art fostered by Kenneth Mac Alpine and his successors' [Curie & Henry 1943:270,271]. The Alcocks too have argued, on historical and other grounds, in favour of a mid-ninth century date for the Dupplin cross, and the involvement of Cinaed mac Alpin. Given the circumstances of his rise to power, Cinaed would have faced a problem of legitimation which could not have been overcome by military might alone. The Dupplin Cross, with its stress on the iconography of kingship, could well be the creation of someone, such as he, anxious to invoke whatever authority was to hand. The most obvious source of legitimation, of course, was the Church, hence the patronage of an important cross, and perhaps, more specifically, the choice of David iconography. Royal patronage of crosses is attested in contemporary Ireland and 'may be seen to document not only co-operation between Church and ruler at the time, but also a king helping to commission a High Cross to bolster up his own importance' [Harbison 1994:104]. The other great authority to which Cinaed might appeal was history and, specifically, genealogy. This is the strategy employed by his contemporary Conchenn, king of Powys, in erecting the inscribed monument
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
259
known as The Pillar of Eliseg* [Nash-Williams 1950:123-4, No. 182]. Conchenn died c. 854 after a long reign and invoked the memory of his great-grandfather Eliset, who flourished in the mid-eighth century. The relative chronology of Conchenn and Eliset, therefore, is comparable to that of their younger contemporaries Cinaed and Custantin. Furthermore, Eliset is credited with uniting the kingdom of Powys, as Custantin was with first uniting the two kingdoms of Picts and Scots.
The main problem with this analogy is that, of course, Conchenn was the direct descendent of the man he commemorated: most probably Cinaed was completely unrelated to Custantin. It may, however, be because of, rather than in spite of, this feet that Cinaed would have been keen to invoke the memory of the founder of the previous dynasty. It is surely significant that he named one of his sons Constantine [Anderson 1973:78). If the general principal of this theory is acceptable, there is no compelling reason why we should stop at Cinaed. Might not his son Constantine attempt to harness the prestige of his illustrious name-sake ? Or what about his grandson, the mighty Constantine II, who ruled the kingdom for over forty years at the beginning of the tenth century ? Of course this is conjecture, and, all things considered, the Welsh comparison is far from conclusive. The same arguments might apply equally well to Constantine mac Forcussa himself, who may have faced legitimation problems as great as Cinaed's. Constantine is one of three Pictish kings to be commemorated in the Durham Liber Vitae [Gerchow 1988:109-54,304-20). The close links with Northumbria implied by this would provide a suitable context during his reign for the strong English influences apparent in the art of Dupplin [Curie & Henry 1943:272).
To conclude, the identification with Constantine mac Forcussa seems secure, but the nature of his involvement with the cross, whether as patron or honorand, is uncertain. Unless more of the text can be read, in the absence of irrefutable art historical evidence to the contrary, Occam's razor requires the simplest conclusion: either Constantine was the living patron of the cross, or it was erected at his death, in other words the cross dates to the beginning of the ninth century. Although the alternative hypothesis may be attractive, it lacks hard evidence, and the burden of proof surely rests on those who would argue
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
DUPPLIN
260
for a post-820 date. Of course, it would help if we better understood the historical and artistic context of the Dupplin Cross. A thorough analysis of its Irish and Anglo-Saxon affinities would certainly be of benefit, but of prime importance is an examination of its relation to the rest of the rich body of material from the surrounding area [Henderson 1978] including, in particular, the crosses at Mugdmm and Dunkeld [Allen & Anderson 1903:367, 284, 317, 342], and the fragments from Forteviot itself [Allen & Anderson 1903:321-7; Alcock & Alcock 1993:222-7]. According to Peter Harbison 'regal involvement in the erection of High Crosses does not manifest itself [in Ireland] until the reign of Maelsechlainn [mac Máel Ruanaid, reigned 846-862] in the middle of the ninth century' [Harbison 1994:104]. If Dupplin is indeed the product of the previous generation, the cross takes on added significance as a precocious example of 'manifest regal involvement' in Pictland, several decades before the practice was established in Ireland.
DUPPLIN - Location of the Dupplin cross in relation to Forteviot. D, Dupplin Cross; H, Haly Hill; S, scarp of the former course of the Water of May; I, site of Invermay cross. Hatching, areas of crop-marks [Alcock & Alcock 1993 illus. 12]
DUPPLIN - Sculptured stones from Forteviot (not to scale).
^ocms
Fragment of cross-arm [Alcock & Alcock 1992 illus.S]
Front» side, and rear view of cross-slab Forteviot 1. Height 0.6m [Alcock & Alcock 1992 illus.4 (Tom Gray)]
f
llfeSflS
^C^5fesís2li^í Front/west face.
Back/east face.
DUPPLIN - Free-standing cross. [Alcock & Alcock 1992: illus. 13-14]
East and south faces.
'FrontVwest
Side/north
'BackVeast
Text panel
DUPPLIN -
RomUly Allen's drawing of the Dupplin Cross [ECMS
fig.334]
Side/south
DUPPLIN - The base of the Dupplin Cross showing traces of ogham lettering [® The Trustees of the National Museums of Scotland].
DUPPLIN -
Inscribed panel on the west face of the shaft of the Dupplin cross [© The Trustees of the National Museums of Scotland].
DUPPLIN - Text panel (cast) with lettering marked with charcoal - lines 1-2.
DUPPLIN - Text panel (cast) with lettering marked with charcoal - lines 1-3, 7.
261
FORMASTON
DISCOVERY This fragment of a cross-slab was discovered some time before 1874 in theformerkirkyard of Formaston in the parish of Aboyne, Aberdeenshire (NGR NJ 5412 0014). The stone was at Aboyne Castle (NO 522993), the seat of the Marquis of Huntly, by 1883 [Rh?s 1892:271], and was still there when Padel visited in the early 1970s. It is now in the possession of the North-East Scotland Libraries Service Museums Service (catalogue number I 6904) and housed in the Carnegie Library in Inverurie, to where it was moved in 1974 [n.a. 1974:5]. A cast of the stone was presented to the NMS in 1874 by the Marquis of Huntly [Skene 1874:602].
SITE The old Kirkton of Aboyne lies two miles to the east of the present village (' Charleston of Aboyne') which is a modern creation. Not far from the side of Aboyne Loch, and a little back from the North Deeside road, the old kirkyard is surrounded by a wall which is perhaps too rectilinear to reflect the line of an old monastic vallum. Under the turf can be seen the remains of what is presumably the medieval church, but no early remains have been discovered except for the ogham-inscribed fragment. The name Formaston (stressed on the first syllable) is obsolete, though 'known to have been applied to the neighbourhood of the old church at the east end of the Loch of Aboyne, ... that place according to local informants, has always been called the Kirkton within living memory' [Alexander 1952:280]. The churchyard is known as St. Adamnan's [in print for instance in Simpson 1943], though the earliest reference to an Adomnán dedication is a mid-nineteenth century local tradition of a St. Eunan's well and tree [Watson 1926:271, quoting Robertson 1843:633]. In the High Middle Ages the pre-eminent church on Deeside was St. Mary's at Kincardine O'Neill, a few miles to the east [Simpson 1943:116-21]. In the early medieval period there appear to have been important monasteries at Banchory (St. Teman) and Tullich (St. Nathalan), both of which have produced Class IV cross-inscribed stones. Immediately to the north and west of Aboyne is the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
262
Howe of Cromar with early ecclesiastical sites evidenced by cross-slabs at Migvie, Kirkton of Logie Coldstone, and Kinnord [for a general discussion see Simpson 1943 and Inglis 1987]. The artistic connection between the cross-slabs from Migvie, Kinnord, and Formaston is discussed below. In some early accounts the stone is referred to as the 'Aboyne ogham1 because of its long period of residence in the grounds of Aboyne Castle. Despite the obsolescence of the place-name, 'Formaston' as a label for the fragment has been gaining ground in more recent accounts. It is unfortunate that the stone has left Deeside. At the time of its removal to Inverurie there was no appropriate local alternative, but the subsequent building of Aboyne Community Centre has provided an ideal location for its safe display, and surely it is time that the slab were returned to its parish of origin.
DESCRIPTION OF OBJECT A fragment of a Class II symbol-inscribed cross-slab carved in relief on one face with the remains of an interlace cross, the mirror symbol, and two vertical lines of incised ogham, one on the moulding at the outer margin of the slab the other, parallel to it, up the face of the slab [see fig.]. Stone:
Granite
Dimensions:
1124mm tall, 443mm wide at base (c.335 at top), 120mm thick
Condition:
Fragmentary, roughly one quarter of the original slab survives, the carving, however, is well-preserved and the inscription is very clear
Macalister claims that the fragment is split and that whatever was on the back is lost. As currently displayed the stone is against the wall and it is hard to get a good view of the reverse. From what is visible, there is no compelling reason to think the back was once carved. The closely related cross at Kinnord and Migvie is carved on one face only. The extant portion represents the bottom right-hand quadrant of a cross-slab which has fractured horizontally immediately under the arm. The vertical fracture begins by following the inside of the shaft's left moulding but about a quarter of the way up begins to cut diagonally across the shaft to the right. Thus at the bottom the entire width of the shaft (minus the moulding) is preserved but at the top little less than a third of it survives.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORM A S T O N
263
The two lines of ogham are of roughly equal length. The outer occupies the whole of the flat, raised band at the outer edge of the slab. The inner line is roughly parallel to it but wavers rather unsteadily up the background surface of the slab between the mirror and the margin. While it would be possible to construe the outer ogham as part of the original design, the inner one, crammed into the available space looks very much like a later addition. Since both appear identical in form and carving technique it is most natural to assume that they are contemporary with each other. One would imagine that if the monument had been intact when the oghams were carved then the lettering could have continued up the entire length of the outer margin (cf. Brodie). In assessing the relation of the ogham to the rest of the carving most previous commentators have taken the inscriptions to be secondary additions and, despite the fact that the inscriptions on raised bands or corner moulding at Brodie, Golspie, and Whiteness are probably primary, it is difficult to disagree with their judgement on Formaston.
DESCRIPTION OF CARVING - DECORATION The Formaston cross is carved in very deep false relief such that the rings and interlace are almost threedimensional. Between the shaft and the outer margin, the original surface has been entirely cut away. The shallower false-relief of the mirror makes it appear to stand out a little, but in fact it has been cut back from this recessed surface. The disparity in relief may be a technical consideration (to have the mirror at the same depth would have required considerably more work), or may be an artistic effect designed to prevent the symbols from detracting from the prominence of the cross. The outer edge of the slab, which is clearly original, is not quite parallel with the shaft but tapers slightly towards the base. If this had been duplicated in the other quadrants, the over-all impression would have been slightly kite-shaped, like Kinnord, though to a lesser degree [see fig.]. Unlike Kinnord (and Migvie) the Formaston stone has a very clearly defined moulding all round its edge. This is integrated into the total design by being joined to the corners of the shaft and arm by means of spirals.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
264
The cross is of a form very popular in Pictish sculpture, with circular hollows at the intersection of the arms completed by small rings. Formaston has the added detail of a small penannular indentation either side of the shaft, slightly higher than halfway up. This indentation, which comes at the point where the lower arm of the cross gives way to the shaft proper, recalls the indentation in an analogous position on southern Pictish cross-slabs, such as Eassie ANG, St. Vigeans 7 ANG, and Dunfallandy PER, a crossshape derived ultimately from Northumbrian forms [Henderson 1993:210-2]. The Formaston cross marks the beginning of its shaft with a change in the interlace only, the southern examples have a horizontal moulding clearly marking the boundary. The Kinnord cross lacks such indentations [see fig.].
The whole of the shaft is outlined with a circular moulding. At the bottom of the shaft there is a hook mirrored by an identical hook at the bottom of the outer margin of the slab. These two are linked by an indented-C-shaped hook. In two-dimensional terms, these meet in spirals, but since the whole arrangement of cross and decoration has a rather plastic feel reminiscent of metal-work, it is possible to see these hooks as two partially obscured circular rings joined by a figure-of-eight ring. The remains of a similar arrangement is visible at the junction of the arm and outside margin, but too much has been lost to be sure of its details. Migvie has very similar hooks at the bottom of its shaft (it is made entirely out of interlace and there is no border moulding) and rings at the top of its upper arm, almost as if the whole thing was a piece of metalwork which could be suspended from the loops. The Kinnord Cross is closer to Formaston, and like it has border moulding. At each corner it has a spiral, except for the bottom left corner of the shaft which has a pierced ring. The related Dyce ABD cross-slab has spirals in the four outer corners of the horizontal arms only.
The shaft of the Formaston Cross is filled with a continuous piece of interlace analyzed in detail by Romilly Allen [ECMS 188]. He describes how below the indentation the pattern consists of figure-of-eight knots arranged in three vertical rows, the pattern (No.572) being repeated three times in each row. At the indentation the laces flow into two spiral knots (twist and ring), above this, as the shaft widens, there is
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
265
a variation of interlace No. 576. A very similar arrangement is found on the intact Kinnord Cross, which also features a variation of pattern No. 572 in three vertical rows at the base of its shaft [ECMS 194 'Aboyne Castle']. The similarities between Kinnord and the reconstructed original of Fonnaston are so close, both technically and artistically that Isabel Henderson was moved to suggest that they may be the work of the same carver [1972a. 173]. The Migvie Slab is more distantly related. Given the paucity of cross-slabs, especially Class II cross-slabs, in the North-East it is quite remarkable that three very striking ones should be in such close proximity as they are at Migvie, Kinnord, and Formaston.
The symbols on Class II cross-slabs are typically kept separate from the cross, either by being placed below it, as at Monymusk ABD, or, most commonly, by being restricted to the opposite face. The latter is the most common arrangement south of the Mounth. In the north, however, there is a prominent sub-type of Class II slab which has the symbols arranged around the cross. Migvie has one in each quadrant, Ulbster CAT has a pair in each. We should imagine Formaston as being of this type, perhaps with one symbol in each of the upper quadrants. Mirrors (with or without combs) only ever appear in combination with standard symbol pairs, on intact stones they never combine with other symbols to form their own pairs. They always appear below the pair they modify, commonly a little to the right. Thus the bottom righthand quadrant is exactly where one would expect to find the mirror if the components of a typical symbol statement were arranged around a cross [see Forsyth 1996b for why mirror is unlike other symbols].
Usually the mirror and comb symbol are immediately adjacent to one another, one of the features which encourages the view that they are to be taken together as a single unit. On some of the later slabs this convention of proximity is not so strictly adhered to, but in only rare instances are the two actually separated, as at Kingoldrum ANG. Typically die mirror is to the left of the smaller comb which is often placed slightly lower. The conventional layout of the mirror-and-comb unit cannot be squared with the extant fragment from Formaston suggesting the slab lacked a comb from the outset. A significant minority of Class I mirror symbols appear without combs. The ratio of extant mirror-and-combs to single mirrors
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORM A S T O N
266
is 4:3, Elizabeth Alcock lists 24 and 18 examples of each respectively. Single mirrors are much less common on Class II slabs, Romilly Allen lists only two in addition to Formaston. One, St Vigeans 2, is dubious since the area where one might expect the comb is very worn and the mirror itself is only partially preserved. The other case is the reverse of Glamis 2 but this side of the slab should really be treated as Class I (it has been argued that the cross and other carving on the other side are later additions, though this is not proven). Mirrors (with or without combs) appear on a smaller proportion of Class II than Class I stones. If, as has been suggested, the mirror and comb reflect female involvement in the erection of monuments, this decline may reflect a marginalization of women in public roles after the establishment of Christianity.
Romilly Allen thought that stones with symbols on the same side as the cross represented a class transitional between Classes I and II and thus were early [FCMS 11.35]. Stevenson [1955] and Henderson [ 1967, 1972a: 172-3] have argued conclusively against this view. Stevenson argued that the Dyce slab was mid- to late-ninth century, with the Kinnord and Formaston crosses probably slightly earlier [1955:126]. To me this seems too late. Stevenson dated the Dupplin cross to the second half of the ninth century [loc. cit.] but recently discovered inscriptional evidence points instead to a date in the first two decades of that century [Forsyth 1996a]. The North-East seems to have followed a slightly different sculptural trajectory to the rest of Pictland, almost certainly because of different patterns of patronage there. It appears that Class I symbol stones continued to be erected longer than elsewhere, very few Class II and III monuments were carved, and the region did not at all participate in the boss style which had such an impact on late Pictish an elsewhere. Crosses such as Kinnord and Migvie incorporate stylistically late features but are carved on roughly hewn boulders - a form of monument elsewhere restricted to the early period [for general discussion see Henderson 1972a]. In her discussion of southern Pictish crosses related in form to Formaston, Henderson adduces parallels with eighth century objects from Britain and abroad. But it is unclear to what extent chronologies relevant to southern Pictland apply to the North-East. While the Formaston cross-slab cannot be any earlier than the eighth century, and, in fact, is more likely to be
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
267
towards the end of that century, a date in the earlier ninth century cannot be ruled out. Of course, since the oghams appear to post-date the fracturing of the slab this date is no more than a terminus post quern.
DESCRIPTION OF OGHAM INSCRIPTION Both lines of ogham are deeply incised with a shaq) V-section which makes for a neat and crisply legible text. Writers such as Browne have contrasted the 'irregular and inartistic lines' of the ogham with the smooth and substantial interlace [1921:105], but this is to make too much of a difference of technique. The oghams on the Bressay and Golspie slabs present a similar contrast, and like them, the Formaston oghams are neatly and competently carved. The wavering of the inner stem-line may be taken as an irregularity, though Pictish artists appear to have avoided rigidity and ruler-straightness. In scale and treatment, the ogham on these three slabs and on Burrian is like manuscript writing on stone. The lettering on Brandsbutt is at the opposite end of the spectrum for it is rendered on a monumental scale and the same substantial technique is used for both oghams and symbols. The differences between Brandsbutt and Formaston do not, however, mean that the latter is not a well-rendered inscription.
The outer ogham occupies the whole of the flat band at the margin of the slab. It is carved on the flat surface, as at Brodie, not on the angle, like Golspie. The stem is 550mm long, and underlies the letters, which occasionally over-shoot it. The long strokes are c.35mm, the short c.20mm. This is the maximum length possible in the available space, and die band of moulding is completely taken up with the inscription. There is some minor abrasion at the stem with portions of certain strokes lost through flaking; none, however, are thus put in doubt. The stem of die inner ogham is of very similar length, c.560mm, though here the oghamist had a little more space at his or her disposal. Since it is not constrained by the edge of the moulding its letters are on average a little longer than those of die outer inscription, c.50mm for the long and c.40mm for the short. In carving technique and general impression, the two oghams are, however, identical.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
268
Both oghams are complete and intact as they stand. If they had been carved before the slab was broken up, it is possible that there might have been further oghams in the other quadrants, but on balance it appears that the inscriptions are secondary, thus what we have is probably the entire text. A key question is whether the two lines constitute separate texts or two halves of a single text; and if the latter, in which order are they to be taken ? There seems no basis for deciding this. There are five more letters on the outer stem than on the inner, despite their similar length. Padel thought the cramping toward the end of the outer line suggested it was carved second, but since the allocation of letters appears to follow worddivision this need not be decisive, as he himself acknowledges [47]. He preferred to take the outer line first, since this put the MAQQ in the middle of the text, as per the Irish formula. But this is scarcely a compelling argument (he did not push it himself), since Fonnaston does not conform to the 'X MAQQI Yf formula in any case. Whatever it means, the outer line cannot be considered a single name.
The lettering is very clear indeed and the only real difficulty concerns the interpretation of the handful of forfeda present. Both inscriptions read from the bottom up, as one would expect. This is confirmed by the slope of letters of the b- and h-aicmi and of the cross-stroke on the hammer-head A (outer 11). For the sake of argument I follow most previous commentators in taking the inner ogham first, but this is not meant to prejudice the question of the relationship between the two lines. The inner stem begins just before the first letter. The carving continues as follows [see fig. for key to numbering]: 1
A single, long oblique stroke, across the stem, sloping forwards - M.
2
A single long stroke, perpendicular across the stem - A.
3
Five short strokes, to the left of, and almost perpendicular to, the stem - Q.
4
As above - Q. The slope on these two letters is very slight
5
A circle intersected at its mid-point by the stem, and occupying a little less than the middle third of the ogham band. This character is known from the Irish manuscript tradition in both smooth circular form, as here, and in angular, 'diamond-shaped' form. The angular form is present in only two inscriptions, at Bressay and at Killogrone, Co. Kerry (CIIC 235), but the round form
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
269
here and, apparently, at Maumanorig, Corca Dhuibhne (CIIC 193). Its possible phonetic value is discussed below 6
Three short strokes, to the left of, and almost perpendicular to, the stem - T.
7
A single long stroke, perpendicularly across the stem - A. This stroke is longer than the preceding vowel stroke and occupies a little under the full width of the ogham band.
8
Two short strokes, to the right of, the stem, sloping forward - L.
9
As above - L. The gradient on these two strokes is much more marked than the previous h-aiane strokes. The two are very generously spaced.
10
Three long strokes, perpendicularly across the stem - U.
11
Two long strokes, perpendicularly aross the stem - O. These two vowels are longer than 2 and 7 and occupy the full width of the ogham band, their distal tips being parallel with those of the following RR.
12
Five long oblique strokes, across the stem, sloping forwards - R. Neatly parallel, evenly spaced, and of common length.
13
As above - R.
14
A single short stroke, to the left of, and the stem, sloping markedly backwards - H. Padel notes that this stroke is considerably longer and deeper than any of the others and, following Rh?s [1892:271] suggests that, rather than have phonetic value, it may be intended as a continuation or termination mark [48]. There are no precedents for Rh?s's 'hyphen', nor any need to suppose a non-phonetic value, since the reading H makes sense in context. The depth of the stroke may be explained as an attempt to ensure that this final singleton was not lost or overlooked, and in any case, as Rhys points out [loc. cit.], it is no more deeply carved than 2. The slightly greater length may be due to the increase in available space as the top of the mirror curves away from the stem. The preceding letters are constrained by the mirror edge and at 14 the inscription is free to open out a little. That the gap between 13 and 14 is a little more than might be expected may be explained to a certain extent by the need to allow room for the backward-sloping stroke, but
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
270
is perhaps more the result of a desire to stretch out the inner inscription so that it does not end too much below the outer ogham. As it stands the outer inscription projects by one letter. Which gives the reading: MAQQ(?)TALLUORRH
Continuing with the outer line of ogham, very shortly after the beginning of the stem-line there are: 1
Five short strokes, to the right of, and perpendicular to the stem - N.
2
Two opposed angled strokes with their mid-points on the stem but a very clear gap between them > <-. This character recurs at 15 and is found also at Burrian and Lunnasting. Its possible phonetic value is discussed below.
3
A single short stroke, to the left of the stem, sloping backward slightly - H. It is very clearly spaced from both the preceding and succeeding letters. It continues right to the very edge of the moulding.
4
As above - H, after a generous gap. The stroke over-shoots the stem slightly.
5
Three strokes, to the left of, and perpendicular to the stem - T. A small portion of the proximal ends is lost in a patch of flaking at the stem.
Up to this point the letters and their component strokes have been very generously spaced, hereafter they become much more tightly packed. 6
Three short strokes, to the right of, and perpendicular to the stem - V. The area is worn. Working from a cast, Skene read the third stroke as continuing across the stem, thus LA. While there is the shadow of a mark to the left of the stem at this point it is insignificant in comparison with the clearly incised lines of the rest of the inscription. If artificial it may have resulted from a slip of the chisel, but it may be nothing more than wear. To the right of the stem the three letters are clear and evenly spaced. If the second and third are separated by a very slightly greater gap the difference is indeed minimal. Since all the other letters are emphatically separated by unambiguous gaps it is certain that the three are to be taken together as one group of bvicnie strokes. This is the stance adopted by Padel and all previous authorities except Skene.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
7
FORMA S T O N
271
Five long oblique strokes, sloping forwards across the stem - R. Abrasion at the stem has carried away the proximal ends of the first three strokes but they are not in doubt.
8
Two long 'herring-bone' strokes across the stem, angled at die mid-point to point forwards, > > 0 . The left distal end of the first stroke over-shoots the moulding slightly, and much of die left half of the second stroke is lost to wear, but its distal tip is still visible and it is not in doubt.
9
A single short stroke to the right of, and almost perpendicular to the stem - B.
10
As above - B. A little more sloped. The two are more widely spaced from one another than from the surrounding letters.
11
A single long stroke perpendicularly across the stem with a horizontal bar at its right tip (J_) a 'hammer-head' A. On die left the stroke extends to the very edge of the ogham band, to the right it covers only two-thirds of it and stops well short of 10, Since, however, the stem is closer the inner margin of the moulding the stroke is actually balanced on its mid-point. The possible phonetic nuance of this character is discussed below.
12
Four short strokes to the left of, and perpendicular to die stem - C. Only the distal top of the second is visible but it is not in doubt. Skene and Brash read this letter as three strokes (T), but may have been influenced by their desire to read ROBBA(I)T. Thefourthstroke is a little faint but it is not in any doubt.
13
After a gap sufficient for one stroke tfiere is a group of four short strokes to the left of, and perpendicular to the stem, identical to 12 - C. There is a slight knobble in the surface of the stem which in photographs makes the final stroke appear more angled than its predecessors, almost parallel with the left-hand portion of the first stroke of 14. The spacing is far too tight for a reading 3 + 1 (TH), especially since succeeding letters of the same aicme are so carefully distinguished elsewhere on die stone. All previous authorities have read C.
14
The same character as appears at 2 > <. The mid-point of both strokes is below (to the right of) the stem which at this point is closer to the left edge of the moulding. Thus the letters are medially placed on the moulding, if not on die stem. There is a clear gap between them, though
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORM A S T O N
272
not as substantial as that between the two components of 2. Padel suggests that it is strange that the oghamist is using the expanded form of E just as space is beginning to run out [48]. In fact the angled strokes probably take up marginally less room than the four strokes of a standard E, especially since extra space would have had to be left to differentiate it from the straight strokes of the surrounding letters. The reason is far more likely to be an orthographic one, even 'stylistic variation', (Padel's 'showing off of erudition') seems unlikely. 15
Five short strokes to the right of, and perpendicular to, the stem, the final one over-shooting it slightly - N.
16
As above - N. The last two strokes over-shoot the stem a little.
17
Four long strokes perpendicularly across th e stem - E. The component strokes of all three letters, 75, 76, and 7 7, are neatly, evenly, but tightly packed. The letters are separated by gaps larger than between component strokes though not as great as the gaps between letters in the inner ogham or earlier in this one.
18
Three short strokes to the right of the stem, sloping slightly forward - V.
19
As above - V. The stem projects a little beyond the final stroke but the presence of the ring at the junction of arm and margin means the inscription must have ended here.
Which gives the reading:
N(?)HHTVROBBACC(?)NNEVV
FORM OF SCRIPT The over-all impression given by the Formaston inscription is of a piece of manuscript ogham writing committed to stone. Its diminutive scale, the slimness of the strokes, and the perfea curve of Inner 5, are all reminiscent of the oghams found in Irish manuscripts. In this it is similar to Bressay, Buckquoy, and Burrian, to Cunningsburgh 3 and to Lunoasting. The Formaston is written on a stem-line with long strokes for vowels. The longest of these (eg. Inner 70, / / , Outer 17) occupy the full-width of the ogham-band. The b- and h-aicme consonants all slope in the expected direction but at greatly varying angles ranging from a raking 45° to the completely perpendicular. Within single letters and between successive pairs of
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORM ASTON
273
letters parallelism is, however, strictly observed. Component strokes are neatly parallel, evenly-spaced, and of more-or-less equal length, there is also ample space between letters with successive letters of the same aicme being particularly clearly spaced. The letters of the inner ogham are all very generously spaced with, in the cases of 8/9 and 10111% as much as the equivalent of a missing stroke. The outer line begins with generous spacing but becomes tighter towards the end, especially after 14, in order to lit into the finite space offered in that quadrant of die margin. This is not so marked as to detract from the over-all effect which is of a clear and simple inscription. There are no pronounced mannerisms, no bind-strokes, and nothing to indicate word-division. There are four forfeda, one of which is repeated, but all of them are also attested elsewhere in the Scottish ogham corpus.
Supplementary letter forms Inner 5: -&-
In both smooth and angled forms this character appears regularly in the manuscripts as
the second of the five supplementary letters. It goes by the name Ór (later Óir). Along with Ébad it appears to have been added at an earlier stage than the other forfeda and is taken along with the members of the SL-aicme as one of the seven vowels of ogham, na sechtfeda [McManus 1991:143 §7.14], The round form occurs in the tracts Auraicept na nÉces and De dúilib feda naforfid
[McManus 1991:142
§7.13]. The straight-sided, lozenge-shaped version occurs in the Book of Ballymote copy of In Lebor Ogaim and also in the eighth or ninth century Codex Bernensis 207, f.257r, leading McManus to argue that this is die earlier form [1991:135 §7.8, 183 n.7.37]. The Formaston slab is contemporary with the Berne Codex, but, ironic though it is that it is the epigraphic version which is round in this case, the general preference for straight-lined characters means that the straight-sided version of Ór is still more likely to be original.
Patrick Sims-Williams has discussed this character at some length [1992:45-62], arguing that the first two supplementary characters,-x- and-6; originally had consonantal value and were introduced in response to
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
274
the 'Second Spirantization1. This sound change, also known as lenidon, occurred in the fifth century, and caused the voiceless stops /k/ and III to become l%l (Olr. ch) and /6/ (th). The first forfid is epigraphically attested with this consonantal value, Sims-Williams acknowledges that a consonantal value for -6- is theoretical only, but discusses why it might not be present in the corpus [49]. Another method of distinguishing lenited and non-lenited consonants was by doubling the latter but not the former [Harvey 1987], but this was not done systematically (it appears in some Scottish examples). In time, however, as the digraphs ch and th became established and the inventory of Irish vowels expanded, the two supplementary characters came to be ascribed vocalic force. Sims-Williams summarizes their subsequent development as follows: some inscribers experimented with using them for /eá/ and, later /£(:)/. Other ogam theorists adopted them as symbols for Primitive and Old Irish /£:/ and /?:/, helped by the development of *etih 'salmon' to é and by the visual resemblance between -&- and O. -Inevitably this pair of long vowels would come to be compared with eta and omega, which were known to have been added to the Greek alphabet, but there is no reason to suppose that Greek was the primary inspiration.-- Then in due course diphthongal values, éo and oilóe, were assigned.... Moreover, because the value /ea7 for x was not entirely forgotten, it was assigned to the Middle Irish digraph ea; hence 0 was assigned to the corresponding digraph oi. The long-vowel values of x and 0 are usually regarded as purely theoretical, and it is certainly true that no certain instance of the / 0 : / value is attested in inscriptions [1992:62]. This gives us a range of possible options for transliterating Formaston's example. The consonantal value can be discounted straight away as obsolete long before the Formaston ogham was carved. The roughly contemporary Berne Codex shows a contemporary value of 'oo\ also given as such in the Auraicept, which lead McManus to reconstruct the value /o:/. Sims-Williams, however, shows why no such character was necessary and demonstrates instead that Ór ( < Lat. aumm) had the value ID :/ [58-60]. The diphthongal value ollóe given in De dúilibfeda na forfid [McManus 1991:138 § 10] is generally considered later but may have been current as early as the ninth century and cannot be ruled out. The digraph oi is the latest value and the least likely, but neither can it be ruled out.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND Outer 2, 14:
FORMASTON
275
The origin and development of the first forfid, Ébad, is closely linked with that of the
letter discussed above [see Sims-Williams loc.cit. for detailed discussion]. Its original value was /g/, subsequently it represented some kind of E. Sims-Williams has argued that it was used to represent Primitive Irish /eM/ and its reflexes, é /e:/ and later the diphthong éo. In time it became equated with the Middle Irish digraph ea.
The character appears in several guises on Scottish ogham inscriptions. Newton has a fully flourished % which may or may not have consonantal value. The others appear more probably vowels. Buckquoy has a small neat x as does Cunningsburgh 3, from context both of these are vowels. Lunnasting has a similar character in initial position, again clearly a vowel, but later in the inscription there are both-^and ^ - £ It is far from clear whether these are variants of the same character or distinct letters. Sims-Williams treats them as all the same [56]. Burrian similarly has letters with and without a substantial gap between the two angled strokes >
< and X - . There is a fundamental distinction in how they are drawn, the one being
composed of intersecting diagonals, the other by opposed brackets which do or do not meet. Golspie has further examples of the latter. Sims-Williams suggests that if Formaston's N*HHT is the well-known personal name element seen in Irish as Ne(a)cht then it provides confirmation of his reconstruction of the value Id before a palatal consonant [56].
Outer 8:
' Angled-vowels' consisting ofherring-bone groups ofarrow-head strokes pointing forward
in the direction of reading (^>% - > - > ^ e t c . ) are employed in five Scottish ogham inscriptions but not in any other epigraphic context. They appear in both bound and unbound variants. One of the whimsical ogham alphabets in the Book of Ballymote includes similar bound angled-vowels in an alphabet which has b- and h-aicme letters comprising groups of straight strokes across the stem surrounded by boxes left open on the upper edge (b-strokes) or the lower (h-strokes). In this alphabet, m-aicme strokes are bound arrows pointing against die direction of reading ( < < < < < = R) and vowels point the other way (•> > > > >= I). The copyist has M = - < - * and A = ^->-which is surely a mistake. Adjacent to it is a similar
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
276
alphabet which has m- and Zraicme strokes consisting of arrow-heads, above the stem (M = ^ , A = i ) . It is possible that in constructing his alphabet the compiler of the Ballymote tract took inspiration from the angled vowels of the Scottish oghams, or a common exemplar. Since however there is a certain geometric logical to the permutations of the different varieties in Ballymote, auto-genesis is equally likely, especially since only the vowels are identical.
That angled vowels were known in Ireland is clinched by an
uninterpreted ogham in the lower margin of a page of the fragmentary Lebor Gabala bound into the Book of Fermoy, RIA MS 23 E 29 (f.5v).
The hand of the main text is probably 14th century, but,
unfortunately, the ogham is 'largely illegible owing to staining' [Catalogue cf Irish Manuscripts in the Royal Irish Academy p.3097]. Macalister's sketch [1937] shows b- and h-strokes perpendicular to the stem, two x-forfeda, a unique hook-stroke, and three angled vowels - A, O, U, but the text, as it stands, is unintelligible.
Birsay 1 has only angled vowels (and a hammer-head A) but is too short for any significance to be attached to the lack of contrasting 'standard' vowels. The other four, Burrian, Cunningsburgh 2 and 3, and Lunnasting, all have standard vowels juxtaposed with the angled ones. orthographic distinction is being made here.
We may suppose that an
What it might be, however, is unclear.
The two
Cunningsburgh fragments, with angled E and U + , are too short to be of any assistance. Lunnasting's angled E occurs in a word which has not be satisfactorily interpreted and the angled U(?) of Burrian occurs in an area of wear and is doubtful. It is Formaston, therefore, which provides the clearest evidence for how these letters are to be construed. The angled O (8) of the outer ogham contrasts with the standard O (II) of the inner, and with thtforfeda 5 (see below) of the same line. Other vowel/orfafa appear either to represent diphthongs or to mark the quality of a succeeding consonant. The extant examples of angled vowels cover O, U, and E. Given that the sample is small it may be that the lack of examples of A and I is pure chance (but for A see discussion of >
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
to rule out diphthongs (which in any case seem well catered for by the other forfeda).
277
Since both front
and back vowels have angled variants it also appears to rule out the marking of quality, unless the angle meant something like 'opposite quality1, i.e. palatal for O and U, non-palatal for E A third possibility is neater; it could be that the angle is indented to marie length, so that+|- « o a n d - » - » 6. This, of course, is assuming that the language represented is Gaelic. If the texts are written in Brittonic then the game is thrown wide open since little is known for certain about the Pictish system of vowels.
Outer 77:
The 'hammer-head' A consists of a standard A stroke across the stem with a cross-stroke
at its lower/right distal end, J^. Padel calls is a 'seriffed' A, but I prefer to revert to earlier usage and stick with 'hammer-head1 since this term is descriptively less ambiguous. Also in the manuscripts many of the strokes have the wedge-shaped serifs typical of the minims of the roman alphabet in contemporary Irish script. The length of the cross-stroke varies but is usually quite short. In a few examples the crossstroke intersect the vowel-stroke before the end -J-, but these are all casually carved and the more precise examples always have the intersection at the very endj_. Epigraphic examples of this stroke are found only in Scotland. It is not found in 'practical' use in any of the manuscripts and though some of the letters in the whimsical ogham alphabets of the Book of Ballymote have cross-strokes none of them are, strictlyspeaking, analogous to the hammer-head A.
The hammer-head A appears in five Scottish ogham inscriptions. In two of these, Burrian and Lunnasting, all the letters are bound and the stroke on the A might be thought to be merely a bind-stroke added to prevent the A being over-looked (though it is not applied to the singleton Hs on Lunnasting). In neither inscription are there any standard vowel-strokes to contrast with the modified A, so it may be that that character is to be interpreted as having the same phonetic value as an ordinary A. To the extent to which the Burrian text may be interpreted the transliteration 'A' is the preferred one. In contrast, there are no bound letters on the other Scottish oghams with hammer-head As, Latheron, Buckquoy, and Birsay 1. At Latheron all the other vowels are standard except the other A (which is of different form having flanges
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
278
at both ends). At Buckquoy there is also a standard A. In Irish examples, epigraphic and manuscript, supplementary letters are incorporated only if there is a phonetic distinction to be made» There is no reason to believe that Scottish oghamists were any different or varied the forms of letters merely for stylistic variation or to show off their learning. It thus seems reasonable to suppose that the contrast in forms reflects a contrast in orthography. If a phonetic or orthographic distinction is being made at Latheron and Buckquoy, the value AI would be appropriate, ainim being an attested variant of anim and -nait being the familiar personal-name suffix (if this is how -NNAT is to be interpreted)» In these cases the / indicates the palatal quality of the following consonant. The evidence of the Birsay 1 ogham is of no assistance since the text is fragmentary and not satisfactorily interpreted. Since none of the letters in the Formaston inscription are bound, it is appropriate to interpret it along the lines of Latheron and Buckquoy which leaves open the possibility that character 11 of the outer line is to be transliterated AI as Skene suggested.
Dating The form of the cross and its decoration suggest a date in the late eight or early ninth century for the slab. The ogham appears to post-date the breaking of the monument but need not do so by much. An eighth century date has been put forward for Buckquoy, which shares with Formaston the distinction of a 'hammer-head' A. The first forfid is attested in the classical ogham corpus, though in its vocalic form is restricted to the later end of the period (sixth and seventh). The only clear epigraphic example of the second forfid (CIIC 235) displays other late features (i.e. late sixth or seventh century). As discussed above, the earliest manuscript attestation of the character is the eighth or ninth century Berne Codex. The Formaston ogham could thus be as early as the late eighth century, though is perhaps more likely ninth century, though there is nothing to preclude a date in the tenth century or even later.
INTERPRETATION OF TEXT The transliteration of V depends on whether the language of the text is Irish or Pictish, fliI for the former u /w/ for the latter; Ne(a)hhtfrobba(i)cce(a)nneff I Ne(a)hhtwrobba(i)cce(a)nneww
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
279
No word-division is indicated so we are at the mercy of our wits in trying to segment the above sequence. The first five characters, Ne(a)hht, are immediately suggestive of the well-known male personal-name element Nect-, attested in all the neo-Celtic vernaculars [see Ackergill for discussion of this name]. There is some variation in the vowel in the extant ogham examples; Ackergill has Neht-, Lunnasting Nehht-, Latheron and Bressay Nahht-f only Formaston spells it with the x-forfid. According to Sims-Williams's analysis [1992:56] this spelling is closest to the Olr* Neacht-. If the following V is not part of the next word then it presents some problems, though note Bressay's NAHHTVVDDAddS, Latheron's NETU+, and Inchyra's (s/n)ETU.
Turning now to the end of the line. Skene was the first to equate the final seven letters Ce(a)nneff with the place-name Kenneff (modern Kinneff in the Mearns). In this he has been followed by all subsequent authorities and the identification is very compelling, The etymology of the name is dubious [Dr. Simon Taylor, pers. comm.], though presumably the first element is the very common cenn 'head', 'end', most usually found in place-names in the dative arm 'at die end o f [DIL\. Given the site's location, Thomas Clancy suggests a petrified locative 'at the head-lands' [pers. comm.]. Whilst it is surprising to find placenames in inscriptions at all, it is all the more puzzling here since Formaston is about 28 miles north-west of Kinneff, which lies on the Kincardine coast two miles north of Inverbervie, and in between is the formidable barrier of die Mounth. The church of Kinneff is famous in Scottish history as the place where the 'Honours of Scotland' (the Scottish 'crown jewels') were hidden from Cromwell's forces» but ecclesiastical activity at the site is recorded in 1242 in the Pontifical Offices of St. Andrews which note the church's dedication to St. Arnold by Bishop David of Birnam [Anderson 1922.11.522]. Romilly Allen mentions 'traces of an early Celtic ecclesiastical foundation' without specifying them, but the small relief cross-slab found there in the 1890s, now in the NMS, may be taken as conclusive evidence [ECMS 204-5]. Though Kinneff and Formaston are a considerable distance away, it ought to be remembered that the administrative unit of the Mearns (Kincardineshire), the bulk of which is south of the Mounth, has an odd
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
280
little bulge over the mountains and across the Dee to take in the parish of Banchory-Teraan (thus encompassing the whole of the Cairn O'Mount Pass). While formalized at the diocesan reorganization which followed the Reformation in Scotland [see map 103 in MacNeill & Nicholson 1975], this unit had older roots, if these stretched back to the early medieval period, then it might go some way to explaining the link between Fonoastonand Kinneff, even though Formaston is several miles to the west of the modem parish and pre-1974 county boundary. A more direct link may be provided by a possible dedication to Adomnáu in the form of a St. Army's kill near the church of Kinneff [Skene 1874:603, citing the OSA voL6 p. 197; the later medieval dedication to Arnold being perhaps a rationalization of an earlier one to Adomnán]. Perhaps Kinneff was a daughter church of a Columban foundation at Formaston, or vice-versa.
These two identifications, Neahht and Ceanneff'leave us with a middle SQcáon frobba(i)c. Skene had obtained the reading Neachtla robbait Ceanneff, He took Neachtla as a form of Nechtan/Nathalan, and robbait as Old Irish robáid, ro~báith% the third person singular perfect tense of báidid 'quench', 'extinguish', 'drown* (cf. MW, bawd 'drowns*, ModW, bodd, boddi)* This is the word used, as the vernacular equivalent of Latin immolcmt% in the land-grants recorded in the Book of Deer, not for the act of alienating land, for which Deer usually has do rat 'gives', but of granting immunity from dues owed to secular authority in respect of land [Jackson 1972:120-3], Skene took the whole line to mean 'Neachtla granted Kinneff, thus he understood the inscription to be a record of the transfer of revenue rights in the lands of Kinneff from a certain Nechtla to the church at Formaston. This would be a most satisfying interpretation were it not for a number of problems. Thefirstis perhaps not insurmountable, namely that the word order is SVO rather than the expected VSO (though Dr. Clancy points out to me that SVO occurs in die Book of Deer). The second is the interpretation of 6: Skene's reading LA is forced and F is what the text actually says today, though admittedly the area is worn. If the strokes had continued across the stem, then we would have had Neacfmi* a plausible personal name (cf. Inchyra's SETU/NETU, and Latheron's NETU -f-)> but there is no evidence that they ever did (and if the text is to be interpreted in Irish
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
281
terms with CENNEFF, then we cannot opt for the Tictish' transliteration /w/ here). The third and fetal problem, however, is the interpretation of 12. Skene and Ferguson read it as T, but it is quite clearly C. What robbaic might mean is unclear. Dr. Clancy has made the ingenious suggestion that a scribal error may have crept in if the text was first written out in Insular minuscule and then transliterated into ogham [pers, comm.]. Mistaking minuscule C for T and vice-versa is probably the most frequent errors made. If one is not prepared to accept the possibility of such a mistake, RO is still strongly suggestive of an Irish perfective particle though which verb is meant by bbaic is not obvious, perhaps baccaid 'hinders', 'prevents' (the doubling of the B might be to show the consonant was not lenited). The cognate participle is also found in Welsh, of course, and a Briaonic interpretation need not be embarrassed by SVO syntax, but once again no verb is immediately obvious [for rv see Evans 1964:166-8 §185].
The possibility that Skene might, after all, be right in reading Neachtla is of great interest because of the local cult of saint Nathalan, centred on Tullich. According to the Scottish Calendars Nathalan shares his feast day, 8 January, with Nechtan, a saint commemorated in the west of Scotland and noted in die Irish Calendars as Nechtán nárdeAlbae
'noble Nechtan of Alba1 (Félire Oengussa for 8 January) or, according
to one source, anair de Albain 'from the east of Alba* [Watson 1926:308]. A Nechtan of Nér is commemorated in the annals (AU 679 dormitatio Nectain Neir) and this has been interpreted as a reference to the Abbey of Deer (inDér) [O'Rahilly 1946:373-4 n.l], and more recently as Fetternear ABD (Clancy pers, comm.]. Watson may be justified in taking these as multiple references to the same figure, Nathalan, also spelled Nothtan, and Nethalen in the Scottish Calendars is, according to Watson, pronounced in die vernacular Nauchlan, Nachlan, These he derives from Celtic Necto-iaunos 'pure rejoicing*, 'rejoicing in purity*, giving Brittonic *Nmthlaun, *Naithlon [1926:329-31]; perhaps lano- (cf. Ir. Ian, W.llawnB.leun (Lat. plenus) 'fair [Evans 1967:215] is a more likely second element. If Neachtla might be allowed as a Gaelicized version of a hypocoristic Neithla* (from *Neithlaun) we may have a reference to this saint. Such a reading reinforces the Kinneff connection. Nathalan was also culted at
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
282
Cowie, in Kincardineshire not far from Kinneff [Watson loc. cit.]. The identification with this saint does not depend, however, on die reading NECHTLA, since he appears in numerous hypocorisnc guises, as in the early dedications Eglismonithoc ABD and Eglismonichto ANG [Barrow 1983],
If we begin to experiment with alternative segmentations, then in ROBBA(I) we might have ro bá the first and second person singular perfect of the substantive verb in Irish, but this would leave CCENNEFF, and in ogham double consonants appear to have been avoided in word-initial position. It is very tempting to emend the four strokes of 12 to three and turn C into T. This would solve many problems, though not die one of word order, but die rest of die inscription is so expertly cut that there seems no warrant for it (but see Clancy's suggestion above). Casting around for alternatives, diere is die word róiba (for róibach ?) which appears in the Wurzburg glosses meaning 'fond of drinking' ( < ro 'too much' 4- ib 'to drink') [DIL]. It might be thought an appropriate sobriquet, if not a very flattering one, though it is not attested as such. 'Drunken Neacht' seems possible, but what then to make of Ceanneff! If we are to abandon this segmentation altogether, DIL gives die word rvibchend (ROBBACCENNI) but suggests it may be an error for the personal name Sroibchend [s.v.]. Relinquishing the interpretation Kinneff, cerm here may have a more literal meaning 'head', 'chief, though ceanneff for ceannaib may be too erratic a spelling for the Irish dative plural
Maqqo-Talluorrh We are on much firmer ground with the inner line, MAQQ and MEQQ being the Scottish ogham equivalents of the old MAQ(Q)I. Though C or CC would be more philologically appropriate for the period, in Scotland the word is always spelled with ogham QQ. The 'MAQQ' of Formaston recurs at Latheron. Following the forfid, to which I will return, we have die redoubtable Celtic personal name element Talo- meaning 'a protruberance, projecting part, swelling, e.g. boss (of a shield), forehead, crest of a wave etc.* [Evans 1967:259-61]. The words occurs also as a common noun in both Irish (nrf, taut)
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
283
and the neo-Brittonic languages (tal). Evans lists a large number of personal names formed from this element in Continemal and neo-Celtic languages [loc. cit.], early Welsh examples include Talan (attested also in OC and OB), Talyrth, Taihaearn, and Toliesin. Formaston's Talluorrh presumably represents some form of the very popular Pictish name TalorclTalorg [Stokes 1890:413-5, MacNeill 1939:16, 23; Jackson 1955:145].
Talore, and its diminutive Talorcan 'with which it is freely interchanged' [Anderson
1922:253], is well-attested in both King-List and annals in a variety of spellings: Talore, Talorg, Tolairg, Talorcen, Talorcan, Talorgen, Talorgan, Talorggan, and Tolarggan, see also Dubthalorc 'black Talorcf [see indices of Anderson 1922 and Anderson 1973]. As Jackson explains, the root is *Talorgagnos, which would give OW *Talorgan, thus the alternative Talorgen represents a specifically Pictish variant [1955(1980):164]. Fonnaston's U is unexpected since there is no hint of a labial quality in any of the other spellings. The doubling of L and R is in line with widespread Scottish ogham practice.
In WCBM the normal development of -rg was it$l in the fifth century [Jackson 1953:570 §148, and mutatis mutandis for the other voiced stops after r: rb, rd]. Jackson puts forward the evidence of Talorc(an) to show that Pictish followed Gaelic in preserving the stop unlenited [Jackson 1955:164]. As ever with Pictish, the water is muddied by the emphatically Gaelic transmission of most of the sources. It should be noted that the P-version of the King-List twice has Talore which Marjorie Anderson emends to Talore. While scribal confusion is probably the correct explanation, it is just possible that these spellings reflect further weakening of a lenited stop /*7 > /e/ (but cf. theTALORI of CIIC 361 Cynwyl Gaeo (Maes Llanwrthwl), Carmarthenshire, [Nash-Williams 1950:140]). Formaston is particularly germane to the problem since its final -H, may be taken to represent some kind of spirant. If so, then Jackson's assertion that Pictish did not participate in this common Brittonic development may need to be re-assessed.
Both MAQQ and TALLUORRH are unproblematic, but what then of 5 the forfed 6, oílóe, or oi ? The personal name MAQITTAL, 'devotee of Tál\ is attested on an ogham pillar from Coolmagort, Dunkerron
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORM ASTON
284
North, Kerry (CIIC 200), and found in Old Irish as Mac-Táil, and Latinized Mactaleus [McManus 1991:109 § 6*14], cf. the Irish personal name Tál [CGHI. Formaston, however, is Ear too late for one to expect die preservation of a composition vowel If the word is not a compound personal name, but 'son of T, * then, a fortiori* the inscription is far too late for there to be any suggestion of a pre-apocope form, Bac Mhic Connain may have MAQU, and Blackwaterfoot 1 MEQI» but these may well be considerably earlier. An alternative possibility is that the word is some form of Olr. macculmoccu* This is the word which appears in Irish ogham as MUCOI and is used to ascribe an individual to membership of a gens [for a detailed discussion of the history of this word see Charles-Edwards 1993:141-60], The spelling MUCOI is standard in ogham [McManus 1991:119-20 §6.27], the oldest manuscript spelling is moat (as seen in VtK but the Annals of Ulster and die Book of Armagh consistently use mac(c)u [D&]. The word was becoming obsolete in the seventh century and after c.700 survives only as a relic, The last instance noted by DIL is 789. By the Middle Irish period the word had ceased to be understood and was changed by writers to mac ul or mac 6a* The word is indeclinable and there is doubt as to whether the final -u standard in Olr, spellings is a fossilizarion of ogham MUCOI (presumably genitive), 'or derives from an original unattested nominative' [McManus 1991:119], DLL lists the variant spelling macco [citing Meyer I914Ji 17 §4], perhaps Formaston's MAQQO for /?makko:/ is a further example.
The etymology of maccu is unknown but it appears only in apposition to a preceding personal name (or occasionally a place-name, ?cf. Cluain macm Ncis > Clonmacnoise) and followed by another personal name in the genitive [see Hogan 1910 s.v. moccu]. While the ogham syntax fX MAQI MUCOI Y* is not followed at Formaston, die inner line may be interpreted as a statement in apposition to the Neaht of the outer line, i.e. ,Neacht(?0 of the gens of Talorg (robbaic) Kinneff.
The extant examples have
MUCOlhnaccuimocu followed by the name of an eponymous ancestor, either a diety or a prehistoric dynast. The P~version of the Pictish King-List places a legendary king Talore/Talorc son of Achiuir in the fourth century and has him reign for the seventy-five years before the Drust son of Erp who reigned
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
285
a hundred years and fought a hundred battles (who was in ami succeeded by a Talorc) [Anderson 1973:246-7].
DISCUSSION In Auquhollie, the North-East has one of the earliest, if not the earliest, ogham inscription in Scotland; in Formaston it has one of the latest Auquhollie is a classic ogham pillar in every respect, but Formaston has a manuscript air» almost as if it were written in the margin of an illuminated gospel 'carpet' page. If these two represent the beginning and end point of a continuous tradition, it is one which was in continuing contact with Irish ogham practice since Formaston exhibits palaeographical features first attested in the contemporary manuscript, the Codex Bemensis. That another feature, the angled vowels, may be first attested in a much later Irish manuscript, the Book of Fermoy, emphasizes that Scotland and Ireland were part of a common tradition of ogham use. Sitns-Williams emphasizes the Munster associations of the first two forfeda [1992:58], which appear at Formaston, but given the exigencies of the data too much should not be made of this.
Not only is the text of Formaston written using an Irish script, but it appears that the language of the inscription is Irish too, though the names Nehht and Talorc are of Pictlsh origin. Since the context of the inscription is ecclesiastical, possibly even Columban, the explanation may be Picts dealing with the Gaelic personnel of their local monastery, though ogham is not a feature of Columban foundations elsewhere in Scotland or Ireland. It has been suggested that certain Class IV incised crosses on Deeside (at BanchoryTernan and Tullich) betray Columban influence [Ian Fisher, RCAHMS, pers. comm,], implying Gaelic influence in the area perhaps as early as the late seventh or early eighth century, certainly within the Picdsh period* That links between Deeside and Ireland were close in the very early eleventh century is indicated by the fact that Domnáll mormaer of MM fought at Clontarf. Chronologically, Formaston fells somewhere in between [see Logie Elphinstone for discussion of Irish connections with Mar and the Garioch, and Auquhollie for Irish connections with the Mearas],
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
286
The Formaston inscription is very different in character to monumental oghams such as Scoonie and Brandsbutt. The latter, in particular, has large letters which can be read at a distance of several feet. In order to read the text of Formaston, one has to be right up close. Why should such a text be added to a pre-existing monument ? The inscription is not a piece of graffiti but a substantially and carefully carved text. Since it is not monumental in the same sense as Brandsbutt or Newton, I wonder if perhaps an analogy should be drawn with the short legal texts written in the margins of precious gospel books, for instance the manumissions in the Bodmin Gospels [Stokes 1872], the charters in the Books of Kells and Deer [Broun 1995], and the records of dispute settlement, manumission, and land granting written into the Lichfield Gospels in the ninth century [Jenkins & Owen 1983]« The impulse would be the same; to lend ecclesiastical authority to a transaction and to ensure the preservation of its record by inscribing it on a holy relic which was of such value as would not be lost or destroyed. If Formaston did record the transfer of land (though such a meaning cannot be derived from it without emending the text), then it would be the annotated gospel books which provide the closest analogy, not the inscribed stones from Kilnasaggan (CIIC 946) and Uanfíhangel/Llanddewi Brefi (CIIC 993), which were erected specifically to commemorate donations.
BIBLIOGRAPHY
OS Record card - NJ 50 SW 1
Skene 1874; Brash 1879:364; Ferguson 1887:139-40; Rhys 1892:271, 1898:349; Nicholson 1896:E; ECMS 188-9, fig. 205; Browne 1921:105-6, pl.xli; Macalister 1940.212-3, pl.V; Simpson 1943:101-3, pK58; Diack 1944:65-6, ill 1; Jackson 1955:140, 141,142; Padel 1972:43,45,46-9; Henderson 1972:1723; Ralston & Inglis 1984:46; Inglis 1987:151-2; Ritchie & Fraser 1994:12,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
FORMASTON
PREVIOUS READINGS
Skene
MAQQQITALLVORRH NE4HHTLA ROBB^IT CÍANNEFF
Brash
MAQQQTTALLUORRH NEAHHTFR OBBAT CfiANNEFF
Ferguson
MAQQQITALLUORRH N£AHH(tf/hgl)AROBB(s/ai)D(oc/cc/ct)CEANNEFF
Rh?s
MAQQOTALLUORR-
Nicholson, Diack, Macalister
NEHHTVROBBACCENNEVV (as above)
ECMS
MAQQQITALLUORRH NAAHHTFROBBACCAANNEVV
Padel
MAQQOTALLUORRh
NEHHTVROBBACCENNEVV
287
FORMASTON - Fragment of ogham-inscribed cross-slab [® Trustees of the National Museums of Scotland, Reg. IB 111]
FORMASTON - Reconstruction of original monument [Macalister 1940:213 pl.vj
»
m
Cross-slab from Kinnord, Aberdeenshire.
Inner
FORMASTON
(key » numbering)
288
GIGHA
1 (Kilchattan)
CURRENT LOCATION The stone now stands atop a small knoll called Cnoc na Carraigh, 'hill of the rock' [NGR NR 6426 4817], on high ground about 90m NW of the ruined chapel of Kilchattan, on the island of Gigha, Argyll [RCAHMS 1971:96]. The site is on the east side of the island racing the coast of Kintyre three miles away and, according to Macalister, if one knows where to look for it, the stone is clearly visible from the mainland. Though this is not the original location of the stone, according to local tradition, it is very close to it.
HISTORY Martin saw this 'square stone' during his visit of 1695 and records that beside it 'the ancient Inhabitants bowed because it was there where they had the first view of the church' [1703:229]. During the eighteenth century, the stone was also described by Pennant [1774] and mentioned in the Statistical Account of the Parish [Fraser 1793], but not until the latter half of the nineteenth was any mention made of the ogham inscription [White 1873]. The first to submit the inscription to critical examination was a party from the Royal Society of Antiquaries of Ireland who, on the recommendation of Rh?s, included the stone in their 1899 archaeological tour of the Scottish Islands. On the basis of their photographs and rubbing, Rh?s published the first reading in the Society's Proceedings for that year [Rh?s 1899].
The stone has fallen and been re-erected at least twice. Thefirstrecorded fall came in about 1843-4 when the stone was toppled 'by some young men for their own amusement' [Rh?s 1901.19]. The piece that broke off the top as a result of the incident was incorporated into a nearby pigsty, and subsequently into the masonry of new houses built on the site [ibid.]. Martin appears to have over-estimated the height of the pillar when he gives it as about 10'(3m); in 1793 the local minister described it as 9' (2.4m) tall [Fraser 1793]. Probably about 2V* feet (0.7m) has been lost but there may or may not have been any
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGHA
289
carving on the missing portion. After the fell, on the instructions of the landowner, the stone was dragged back to the top of the hillock and re-erected on its old site 'with great care'. It was fixed in place by a mason but, because the hillock had been quarried on one side, towards the base of the stone, it fell again some twenty years later, i.e. c.1864. According to local witnesses the condition of the stone was not impaired by the second fall. Rather than re-erect it precisely on the original site, the new owner ordered it be replaced 'at a very short distance* [see Rh$s 1901.19 for eye-witness accounts of the two incidents]. For discussion of a possibly ogham-inscribed fragment alleged to be the missing part of this stone see Appendix: Dubia.
SITE It is unclear whether the monument is a re-used pre-historic standing stone or a pillar specially erected for the bearing of an inscription. Early writers describe it as the middle of an alignment of three stones in a straight line. The first, a 'beautiful plain stone' (c.4.9 x 1.2 x 0.2m), was still standing in the middle of Achadh a' Charra, 'Field of the stone*, in 1793 [Fraser], but it was removed at the beginning of the nineteenth century [RCAHMS 1971:63] and by the 1920s was being used as a drain cover [Diack 1926:4]. The ogham stone was to the north-east on a hillock, and beyond it, on the higher Cnocna Croise* 'Hill of the Cross', there was a free-standing cross which, though upright in 1695, had broken and fallen some years before 1793. No trace of it survives. The tradition of three stones in a straight line persisted long after they had ceased to be visible. In the 1920s the local minister recorded the designation 'The Clocks of Gigha' and the tradition that they 'were so placed in order to get the time from the sun' [Diack 1926:4]. Doubtless the horological explanation is relatively recent, perhaps stemming fromananglicizationof Sc.G. cloch 'stone* > Eng. 'clock*. The association of pre-historic (pagan) pillars and Christian cross is an unlikely one, unless, as has been suggested, the cross was carved from a pre-existing stone. The alignment may be merely co-incidental. Pennant [226] records the presence of cairns in the vicinity, some containing cists, but there is no trace of them now.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGHA
290
The ruined church beside the ogham stone is dedicated to Catín, usually identified as the late sixth century saint, uncle and teacher of Bláán, bishop of Kingarth, Bute. The building itself probably dates from the thirteenth century [RCAHMS 1971:111], but the Kil- place-name is suggestive of an earlier foundation» Further dedications to Catánare found in Bute, Colonsay, Islay, Luing, Kintyre, Ardnamurchan, and Loch Btive and he is also associated with Stornaway and Aberruthven [Watson 1926:277]. According to the Breviary of Aberdeen [ii,3,77], Catáii was of noble Irish descent. An Irish genealogy of a saint of that name describes him as of the Dal nAraide of Ulster, who were considered to be Cruithne [Ó Riáin 1985:102, 662.211]. The remains of a free-standing cross and a slab incised with early Christian symbols in the burial ground at Tarbert, indicate another centre of early ecclesiastical importance on Gigha, but its relationship to die Kilchattan site is unclear.
DESCRIPTION OF STONE A tall, four-sided pillar bearing an ogham inscription on its north-west edge [see fig,], Stone:
Granite
Dimensions:
1.7m tall x 0.25m to 0.31 m at the base, tapering to 0.20m to 0.23m at the top
Condition:
Poor; the ogham, which is worn and in bad condition, is defaced by series of spalls on the arris.
In addition to general weathering, there are a number of more specific injuries, most serious of which is a series of four large, shallow, unevenly spaced chunks missing from the ogham-bearing arris. Though die depressions are themselves worn, they evidently post-date the carving of the ogham, and how they got there is a mystery. The pattern suggested to Rh£s that the pillar had been used as a gatepost, but there was no local knowledge of such a usage, and die stone's position on top of a hillock weighs against it. Rh?s satisfied himself that the chunks were not lost during re-erection; on die first occasion the stone was dragged on a sledge, on the second several men carried it on their shoulders, in neither instance was it dragged by chains or die like. It is possible that the vandals who pulled it down in die 1840s used chains to topple it, but, of course tins cannot be proved. Macalister also mentions 'frets of vertical and horizontal
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGH A
291
lines' on the stone which are 'evidently modern additions' [1945:484], and some modern initials cut by 'that brainless pest the cheap tripper', fortunately, the latter do not impinge on the ogham.
DESCRIPTION OF CARVING The diversity of the published interpretations of this ogham are an indication of the extreme difficulty in reading it. The general condition of the stone is poor, and the numerous marks mentioned above are a serious obstacle to a definitive reading. The stone is not sufficiently well-preserved to make definitive statements about carving technique. As they now survive, the strokes arefinger-scaleand with U-section. Where it can be measured with any certainty the length of h-aicme strokes is about 60mm. I have not examined the stone in person, the following is based on an examination of the plaster cast made by the RCAHMS in the late 1960s and housed in the NMR. conversations with Ian Fisher and Graham Ritchie of the RCAHMS about their examinations of the stone, and a careful comparison of the various published accounts of the inscription. Rh?s, Diack, and Macalister examined the stone personally. Readings by the latter two, however, are often idiosyncratic and their testimony is treated with caution. As far as I have been able to establish, Jackson's account of the stone was based on an examination of the RCAHMS plaster cast (his pencil marks are still visible on it). What would otherwise be a reliable testimony is thus somewhat weakened. It is unfortunate that Padel, whose readings are consistently sensitive and accurate, chose to omit the so-called 'Irish-looking' oghams, including Gigha, from his study. A particular problem with Jackson's account is the alleged gaps in the lettering. It is very unusual for spaces to be left between letters in Type I ogham inscriptions of this kind. It is more likely that the 'gaps' contained vowel notches now lost, as Rh?s and Macalister claim to have seen when they examined the stone. At that time, up to sixty years before the making of the RCAHMS cast, the stone may have been less weathered.
The inscription reads from the bottom up. The portion nearest the ground is the most badly worn and the most disputed. Macalister and Rh?s began at the same point just above the ground. The Irish Antiquaries and Jackson read nothing for the first 0.25m and began with letter 7 (M). On the basis of photographs
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGH A
292
and a rubbing, Rh?s at first concurred, but revised his view after inspecting the monument himself, and read four letters before this initial starting point. Their drawings show that Macalister [1901] and Diack reached very similar conclusions about the number and position of strokes, though they differed on how these were grouped and subsequently their readings (VICULA and VIDDOSA respectively) suggest more disagreement than was the case. Macalister's reading would have been available to Diack, but the extent to which he was influenced by it is not clear. Macalister's 1945 drawing differs significantly from his initial one [1901], the suspicion remains that it was revised to bring it more into line with his suggested reading. Though he could not make out any letters, Jackson acknowledged there were marks in this area, and his drawing agrees in some respects with Diack and Macalister.
The first six letters are extremely doubtful. There are definitely strokes in this area, the indentations can be clearly felt on the arris, but the lines are too indistinct to interpret. Because of the measure of agreement between the threeearliest witnesses, I offer the following tentative suggestions based on a careful comparison of their accounts. The ambiguous indentations of the cast neither specifically suppon nor refute their readings [see fig. for key to numbering]. i
Three strokes on the B-side ? These are not at all clear, the surface being spalled. Macalister claimed to see signs of the ends of three scores (V), but said they could equally be read as L (two strokes).
2
Macalister and Diack were both certain of five vowel notches, the fifth being in a spall - 1 .
3
Four or five strokes of the h-aicme. Again due to spalling, this portion is not clear, the individual 'strokes' run together, and the two readings were based on measuring the available space. Diack thought there was room for five scores (Q) but could see the ends of two groups of two only (D D). Macalister wavered between five and four - C.
4
Two or three vowel notches. Macalister conceded the third point was doubtful but still read U. Diack saw two strokes (0), but admitted the spacing was rather wide. Presumably this is Rh?s's 0 . Jackson acknowledged two notches disappearing into the spall, which could be taken as
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGHA
293
confirmation, though he had the second continue onto the B-side. In his drawing the marks are tentative, but surely the second is not sufficiently angled to be part of the m~aicme. He described them as 'quite likely natural'. 5
Two strokes were apparent on the B-side. Macalister preferred to read them as one letter because 'L seems to fit the context better*, though they were rather wide apart for this and could be BB. He identified two marks like tips of scores before this letter on the h-side as mere nicks since they did not reach the arris. It is probably these that led Rh?s to identify two cross-strokes and read G. Diack took no notice of the h-side marks but claimed to see parts of three strokes on the bside and postulated a fourth (S) on the basis of the spacing of the next vowel. Jackson left the area blank and the whole is very doubtful. If any of the suggestions have linguistic merit, it is Macalister's L.
6
One faint vowel notch (A), which has been absorbed into a broad sloping mark, like the upper half of an M, running out of it on the h-side. Macalister described it as 'the most conspicuous mark in the inscription... certainly artificial* yet Diack did not mention it. Macalister suggested that the alleged score was a first attempt at the following M 'made too near the A and accidentally splintered into it', but this is unlikely. He did not mark it on his 1945 diagram. Jackson included the mark, acknowledging that it does not reach the arris, but expressed the opinion that it 'may be natural', 'if not, it is presumably H \ Rh^s read M A, but Jackson denied there was anything at all on the b-side. There might be another m~aicme stroke before 7, though they are surely too widely-spaced to form a pair.
From this point on the carving is clearer and there is much greater agreement between the various authorities: 7
One cross-stroke - M. On the b-side, the end of the stroke has disappeared, but the letter is not in doubt. Diack said the stroke was bent in the middle, Rh$s that the middle portion was lost, though the ends were visible.
8
There is a large hole, which Jackson thought held one stroke of an A. He discounted the apparent
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGHA
294
stroke protruding to the left. Diack thought the space (125mm) too wide for a single stroke and read O. 9
Five clear strokes - Q,
10
There follows a 100mm gap which would be expected to carry vowel notches, Both Macalister and Jackson left it empty, Rh$s claimed to count four of die notches of an I 'small and neatly cut', the fifth being lost in wear at the edge. I is the most appealing reading, since it points to the well-known formula 'X MAQI Y \ however, from the published diagrams, there does not appear to be sufficient room to accommodate five notches. Three would fit more comfortably (U), Diack claimed to see a ring-shaped character, like ones on the Latheron and Formaston stones, on the h-side, which he interpreted as 01. This is extremely unlikely.
11
The next group is disputed. It begins in a spall through which, on both faces, runs a groove, the result of modern damage. Coinciding with the groove is a small stroke which protrudes left-wards from the hole. Jackson thought it could be a letter, though very doubtful. Diack, Macalister and the Irish Antiquaries were happy to include it as the first in a group of four strokes to the left of the stem - C. From a rubbing Rhjs read four strokes, but on examining the stone later changed his mind to three (T)> The three strokes are clear except on the actual arris.
12
There is 110mm before die next consonant. Only one vowel notch is left but it is in the position where one would expect to be the second of a pair. The space seems far too great for a single notch, and since the extant one is not centrally placed it is likely that at least one odier has been lost. Jackson saw one 'rather doubtful' vowel notch (A) just over half-way between the consonant strokes, but admitted there was room for two (O). Diack read O. Macalisterf felt almost equally balanced between reading OM and U after the C; the M being very faint'. Rhjs read I on die basis of his assessment of the space available.
13
Two slightly bent cross-strokes, the right-hand portion of the second being obscure - G.
14
There is 200mm before the next consonant, in which four vowel notches are clearly visible, a fifth, at the beginning is more doubtful. If it is accepted dien the space between the first vowel
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGHA
295
notch and the previous consonant stroke is a third of the distance between the last vowel notch and the next consonant stroke. If this doubtful first vowel notch were disallowed then the gap after the previous letter would be only slightly larger than the gap before the next letter, but only if measured at the stem, the preceding consononant stroke slopes forward slightly. Rh$s [1901] read E and this may be preferable. All other authorities have read I, though Jackson considered the first and fourth notches doubtful, so, in addition to I, offered E, OA or UA as possible alternatives. Rh$s thought there was further room for the five cross-strokes of an R, though admitted he could not place them. His suggestion is very unlikely. 15
There follows a group of strokes to the right of the arris. Jackson saw two (L), the second being the clearer, but thought them rather far apart to be one letter (BB). If there had been a third between them the group would have been quite cramped. The Antiquaries counted four (S), Rh$s read five (N) as did Macalister and Diack. Jackson saw nothing more and admitted that 'the readings of those who claim to discover more seem to me imaginary'.
16
Beyond this Macalister mentions a final I which 'can only be guessed at* [1945], Rh$s thought he could count the notches. Diack described them as 'rather feint, though certain enough*. There are indeed a series of bumps which could be vowels, but equally they could be natural, or the result of damage.
The following tentative reading can be put
forward:
vi(q/c)ulaMAQ(u/ /i)(c/t)0( /m)GI(n/l/bb)i
FORM OF SCRIPT Gigha is written in a classic Type la ogham with no drawn stem-line and vowels which are little more than nicks - the 'text-book' form of the script. The m~aicme strokes slope forward, but other consonants are perpendicular to the stem, as far as can be discerned, given wear, all strokes appear to be equally spaced, regardless of letter boundaries. There are no successive letters of the same aicme. The letter strokes are of moderate width, length, and spacing, giving the letters anuncnannered profile which is neither squat nor elongated. There is more of a distinction between vowel and consonant strokes here than at Auquhollie,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGHA
296
and none of the sloping found at Newton» The closest parallel in Scotland comes from Guroess, but it lacks Gigha's evenness of spacing. Comparable scripts are, however, ubiquitous in Ireland - another reason for considering this one of the earliest oghams in Scotland, if not the earliest.
INTERPRETATION OF TEXT All the above authorities have, to some degree, shaped their interpretation of the carving to accommodate particular readings and expectations* The appearance of MAQ- is reasonably certain, and in all likelihood Gigha 1 conforms to the standard Irish formula 'X MAQI Y\ The only problem with this interpretation is the doubt over the vowel following the Q. If there is nothing there at all, and the reading is simply MAQ, then parallels can be drawn with Irish oghams dated tentatively from the middle or second half of the sixth century onwards [McManus 1991:§5.28-301. On balance, however, even though there are no longer any notches visible, the spacing of the flanking letters is sufficient justification for positing some kind of vowel. Though there appears to be insufficient room for I, anything else would be difficult to account for* U fits the available space, and, in die light of bi-lingual oghams from Wales (such as CIIC 341, 442, 326,428) which have MAQI in ogham but MAC(C)V- in roman letters [McManus 1991 §5.18; 5*32], might be acceptable as an otherwise unattested spelling of an early intermediate stage of apocope and de-labialisation of Q (/kwi/ > /kw/ > /k/, Q > QU > CU > C). Gigha's QU for CU, could perhaps be explained through the tenacity of the orthographic convention of spelling the word for 'son' with Q, a convention which seems to have exerted a remarkable influence in Scotland, if that is how such ahistorical spellings as MAQQ and MEQQ are to be explained»
If the letters before the M are not admitted then the formula MAQ-X is likely, either as 'the son of X\ or more likely, as a compound name with first element MAQ, of which there are many ogham and manuscript examples, for instance - MAQI-TRENI > Mac-Tréin% (MAC(C)V-TRENI); MAQI-ERCA > Mac-Erce, MAQI DECCEDA > Mac-Dmchet As McManus explains, such names *do not denote a filial relationship to the second element ... but probably had the meaning "devotee" or the like' [1991:108-9,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGHA
297
§6.14]. Jackson subscribed to the former view and offered a Very hypothetical' interpretation - 'the son of Coicéile' [RCAHMS 1971:96, for Coicéile 'mate' < com + céile, see Uhlich 1993:207]. Jackson thought both MAQ and the (second) name lacked the genitive ending -/, hence he considered the inscription late, 'perhaps not older than the 7th century', but this is doubtful. As discussed above, the final vowels il-u and 4 may indeed be present and thus the text need not post-date apocope.
Surely, however, there is sufficient evidence to warrant the acceptance of at least some ogham carving before the MAQ even if one is not prepared to accept the letter by letter interpretation. Hence Gigha 1 does appear to conform to the standard Irishformula'X son of Y'. Macalister's suggestion that VICULA represents Fiacal 'tooth', must be rejected since Olr ia is from earlier *, both long and short / remained as /. The only personal name listed by O'Brien as beginning with fie is Fichellach (< Fidchellach) [CGH], and to spell this as VIQULA is indeed eccentric. Macalister interpreted the second name, COMGINI, as Olr Coemgen, a widespread and well-known male name (Cóem + -gen 'fair born' [Uhlich 1993:207]), found variously spelled, including Comgani [Uhlich 208].
DISCUSSION Gigha 1 is one of the most damaged ogham stones in Scotland and consequently presents a number of difficulties, as the above has made clear. Despite this, enough has been preserved to indicate that the monument is of the classic Irish type. In terms of the size and shape of stone, the positioning of the lettering on it and the proportions of the strokes, Gigha 1 is closest of all the Scottish examples, to the Southern Irish norm. To reject a Goidelic milieu for this monument, in favour of some appeal to Pictish, would be perverse indeed, even in the absence of closely paralleled names and orthography.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GIGHA
298
BIBLIOGRAPHY OS Record Card - NR 64 NW 2 Martin 1703:229 (no mention of ogham); Pennant 1774 (no mention of ogham); W. Fraser 1793 (OSA) (no mention of ogham) T, P. White 1873:384 (no reading); T. P. White 1875 pl.IV. (no reading); Brash 1879:364 (mention only); Rh$s 1899:346-349, ill.; Rh$s 1901 18-23; Macalister 1901; Macalister 1902 131-135; Diack 1926; Macalister 1945:484 (CIIC 506), ill,; RCAHMS 1971:96-7, No.244 Fig.107 pl.16.
PREVIOUS READINGS Rh$s 1899
MAQICAGILEB
Rh$s 1901
OGMA MAQITIGERNI
Macalister 1901, 1902
VICULA MAQ COMGINI
Macalister CIIC
VICULA MAQ CUGINI
Diack
VIDDOSAMO QOICOGINI
Jackson (RCAHMS)
HHMAQt/cAGH/bb
GIGHA 1 (Kilchattan) - Ogham-inscribed pillar [RCAHMS 1971 pi. 16]
GIGHA 1 (Kilchattan) - Comparison of previous readings
a)
MACALISTER (1902)
frnM^fapM-Jto^fr b)
DIACK (1926)
•••••
^j&m*d»**//..~. innm~...
"«p
c)
JACKSON (RCAHMS 1971)
d)
Key to numbering of letters
....JHI. 1
9000"
2
3 *
./.IHH...ÍIH..//. 5 67 8
9
10 11 12 13
14
15 16
299
GOLSPIE
DISCOVERY This large, ogham-inscribed cross-slab was transferred to Dunrobin Castle Museum in 1868 (Ace. No. 15/15A) from the parish churchyard of nearby Golspie in the county of Sutherland (NGR NC 837 002). The cross was certainly in Golspie churchyard in 1856 when it was visited by Stuart [1856:12], and may have been there as early as 1780 [Cordiner 1780:72, 1788]. Close-Brooks suggests that Sir Robert Gordon's description of 'a ston curiously carved' standing in 'Golspie Kirktoun' in 1630 may be the earliest reference [1989:16]. Stuart records the tradition that it bad previously been at Craigton (NH 7898) two miles south of Golspie [1856:12], though there is no record of any ecclesiastical site there [ECMS 41]. Allen said it had been moved from Craigton in 1840 [ECMS 48]. Close-Brooks claims this is a mistake, perhaps arising, she suggests, because the stone, which had previously been lying horizontally, was reerected at this time. Allen says that the Class I symbol stone from Craigton, which still stands in die Castle grounds today, was removed to Dunrobin in 1840 [ECMS 41] and perhaps that is the source of the confusion. The tradition that, at whatever date, die cross-slab had previously been at Culmaily (NH 8099), the only early ecclesiastical site in the area, has persisted. According to the OS this site is indeed its 'most likely provenance' [OS record card NH 79 NE 5]. The chapel of St. Andrew at Golspie became the parish church only in 1619-20 [ECMS 41], At some point before the middle of the eighteenth century, and probably in the seventeenth, the cross-slab was re-used as a grave-slab. The cross side is defaced with a late inscription which reads: HEIR IS THE BURIAL PLE AC [sic) TO ROBERTGORDON ELDEST SON TO ALEX GORDON OF SUTHERLAND]
This must refer to the Dunrobin family sometime between 1514 and 1766, though farther work would be required to pin it down more closely. The deceased cannot be the Sir Robert Gordon, 'historian of die house of Sutherland1 [DNB], mentioned above (1580-1656), since he was the fourth son of Alexander Gordon, eleventh (or twelfth) Earl of Sutherland.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
300
The stone is now on display in the museum in the grounds of Dunrobin (the seat of the Duke of Sutherland) along with a superb collection of local Class I Pictish symbol stones and Class III early Christian Sculpture, and an eclectic array of local antiquities, historical, ethnographic, and natural history specimens [Close-Brooks 1989]. Cordiner published a ratherfancifuldrawing of it but Stuart's was the first accurate representation [1858 pi. xxiv].
LOCALITY The old name for the parish of Golspie was Kilmaly (Culmaliun 1275) < citil Mháillidh 'Maillie's retreat' [Watson 1926:290]. Nothing is known of this saint who is not mentioned in Scottish or Irish calendars. With the exception of one possible outlier in Dunbanonshire, the sites associated with Máillidh have a rather 'Pictish' distribution. He is commemorated in the name of the largest parish in Scotland, Kilmallie, on Loch Eil INV, and not far away on Loch Arkaig there is Glen Mallie, the river Mallie, and Invermallie (all these just to the north of the Great Glen). Dalmally at the head of Loch Awe ARG is on the major route east into Pictland through Loch Lochy and Strath Fillan. Of particular interest are two eglis names incorporating the name of this saint, Eglis-maldiis (Inglismaldie, North Esk) KIN, and Eglismaly near Burntisland FIF. Though in his discussion of this Brittonic early ecclesiastical place-name element, Barrow does not mention the latter example [1983]. Not far away from it is the church of Inchekkerie, which is recorded as being dedicated to St. Maling whom Watson identifies with Máillidh [Watson 1926:290; see also Redford 1988:243].
A number of very fine Pictish Class I symbol stones have been recovered from this area of Sutherland, most of which are in the Dunrobin or Inverness Museums [Ritchie & Fraser 1994 (full refs.); ECMS Close-Brooks 1986: ch.9, ill.; Close-Brooks 1989, ill.; Harden 1986]. Some are regarded as being close to the symbol prototypes and among thefinestof all Class I stones [Stevenson 1955:110: Henderson 1958, 1967:106]. One of these was recovered during Joanna Close-Brook's excavation of Pictish graves in the grounds of Dunrobin [1980], and other Dark Age cists in the vicinity have been explored [Woodham &
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
301
MacKenzie 1957]. One of the most unusual Pictish finds fromthe area is a remarkable gilt-bronze pin now in the NMS. Decorated with a fine human head, it has been dated to the eighth or ninth century [CloseBrooks 1975; Laing 1993:80-1 No. 131; ill. Ritchie 1989:5]. Just to the north of Dunrobin is the wellpreserved broch of Cam Liath whence came the unusual silver fibula decorated with a double-disc symbol. It has been identified as a native version of a well-known Roman type and dated to the beginning of the Pictish period, fourth-fifth century [Ritchie 1989:51 ill.].
DESCRIPTION OF OBJECT A large Class II Pictish cross-slab sculptured partly in relief, partly incised [see fig.]. Stone:
Purple (Old Red) sandstone
Dimensions:
1.83 x 0.82 x 0.17m (to plinth) (Allen's measurements are 6* tall 2'8" wide at base, tapering to 2' at the top, and 6" thick)
Condition:
Cross face defaced by modern inscription and badly weathered. Reverse in generally very good condition. A small portion of the carved surface has laminated off around the beast symbol. The ogham is clear except for two short sections where the surface is very worn.
The lectangular slab is carved on all four sides. The ogham occupies a roll-moulding on the right-hand edge of the reverse face. It starts at the bottom, reads up the long edge, turns the top corner and continues across the upper edge. The bottom right-hand corner of the reverse face is missing, but since the carving makes allowance, it is clearly an original feature. An exact parallel can be drawn with the Class II crossslab The Maiden Stone1 ABD, and with the shaft of thefree-standingDupplin Cross PER. This device may have been merely stylistic, or may have had a practical purpose, making it easier to slot the slab into its base while raising it to the vertical [R. M. S. Spearman, pers. comm.]. A number of the Class II slabs still stand in their original bases, though others have been moved, it appears that at least some of the earlier ones were set into the ground directly. If the Golspie slab ever had a base the two were separated when the slab was taken to be reused as a grave-cover.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
302
DESCRIPTION OF CARVING - DECORATION From an an historical perspective, the Golspie slab is of interest because it appears to reflect a transitional period between the purely incised monuments of the earlier period and the higher relief of the later. Henderson singles out the intertwined snakes on the reverse of the slab as one of her examples of the ' first movement towards relief, and includes the stone as one of the earliest relief monuments. As she points out, 'if the incision is sufficiently deep, then the area contained within two incised lines is virtually in relief [1967:128]. The scrolls on the narrow faces of the slab are so deeply cut as to be almost threedimensional. Stevenson considered the designs on the rectangle and the crescent symbol to be 'highly developed* and placed the Golspie stone at the end of the series of early relief sculptures, before the development of the early boss style and the higher relief monuments. He suggested a date in the early ninth-century (in general, Stevenson tended towards a date for Pictish sculpture later than most other authorities). Linking Golspie with the, in his opinion, slightly earlier Ulbster stone, he described both as 'aberrations of chiefly local interest as for as stylistic developments are concerned, for the far north had fallen behind' [Stevenson 1955:116]. Stevenson's dating would put Golspie earlier than the magnificent cross-slabs of Hilton of Cadboll. Nigg, and Shandwick, slightly further south. Henderson has argued strongly for a date in the second half of the eighth century for Nigg [1994:76] which would push Golspie back further, if it is, indeed, to be placed earlier in the series.
'Front' To bear the modern inscription, a band several cm. wide has been shaved off the 'front' or cross-face of the slab along its two long edges and across the top. Romilly Allen says it continues round all four edges [ECMS 50] but the lower one is currently obscured by the plinth. In addition to this deliberate defacement the effects of lying flat exposed to the elements for many years are clearly visible. The stone has suffered considerable weathering and, in some pans has been worn smooth by the passage of feet. Chunks of the carved surface have been lost to lamination, and much detail has been lost.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
303
The central and right-hand portions are still substantially clear, however, and Allen was able to offer a detailed analysis (which is the basis for the following account). The decoration, all carved in shallow (false) relief, consists of an interlace cross surrounded by nine panels of geometric ornament. The cross is long-shafted and has circular arm-pits. It is of a shape (Romilly Allen's No. 101a) which was extremely popular with Pictish carvers. Romilly Allen lists thirty-eight crosses of this design in Pictland, including Brodie and Scoonie [ECMS.II:51]. In the centre of the cross is a roundel decorated with double-beaded interlace of a form (No.785) apparently unique to Golspie. The interlace work on the arms is partly defaced, but Romilly Allen was able to identify that on the right as No.662, a known Insular type. The top arm is filled with two interlaced ribbon-creatures with fish-tails. The shaft has, in Romilly Allen's words 'double-beaded interlaced-work (No.690) composed of a small piece of broken plait, arranged in two rows and repeated six times in each row, the pieces of plait in one row being the symmetrical opposite of those in the other with regard to a centre' [ECMS 48]. This particular arrangement is unique to Golspie.
The background ornament is arranged in nine panels separated by plain bands. On either side of the top arm is partly defaced interlace work (No.607). Romilly Allen described this as 'an extremely beautiful pattern, and the cruciform spaces in the centre between the Knots cannot have escaped the notice of the designers. The point is here approached where symbolism and decoration may be said to overlap' [ECMS.II:235]. It appears at Nigg, on the St Andrews sarcophagus, on the Thornhill Cross, Yorkshire, and in the Lindisfame Gospels [ECMS.11:235]. Allen's comments on these particular panels at Golspie are worth quoting in full: 'The ornament on each side of the top arm of the cross is the same, except that the pattern on one side is turned through a right angle so as to make the knots face in a different direction. The effect is to give the appearance of symmetry without the monotony of the absolutely 'cast-iron' repetition which is so characteristic of most modern work. The two panels also are of slightly different size and form so as to suit the shape of the slab. A modern mason would cut away the stone until both sides were exactly alike, instead of using his brain and art feeling to save a needless waste of material.' [ECMS 48-9 n.2] On either side of the shaft there are three panels, those on the left being entirely defaced apart from the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
304
remains of some spiral work in the lowest one. Those to the right are all partly defaced, but in the top panel can be discerned double-beaded serpentine interlace (No.648), an arrangement unique to Golspie. Below that is a panel of diagonally-set square key pattern of a form (No.970a) also found at Nigg and in the Lindisfarne Gospels. The bottom panel consists of spiral work (No. 1084), again unique. Across the base of the slab is a broad panel of key-pattern, a combination of Nos. 969 and 974, both of which are fairly common.
'Reverse' On the reverse of the Golspie slab are no fewer than eight Pictish symbols, which makes this slab second only to the Ulbster CAI slab in number of symbols (the latter has four pairs on one side and one pair on the other). Only three other monuments have more than four genuine symbols: Meigle 1 PER (two pairs plus mirror-and-comb), Rosemarkie ROS (two pairs plus mirror plus mirror-and-comb), Dunfallandy PER (three pairs (one incomplete) plus a set of three smith's tools which appear to be functioning as symbols). The normal symbol syntax is the symbol pair, with or without the modification of the mirror or mirrorand-comb. A close examination proves die Golspie symbols to be carefully arranged in pairs, differentiated by scale and by relative placement. The top two, the 'rectangle' and the 'Pictish beast' are much larger than the others. Arranged vertically as is normal, they occupy the whole width of the slab. This is the only Class II example of the decorated rectangle, and the symbol is not particularly common on Class I (ten examples: Sandness SHE, Firth and South Ronaldsay ORK, Ackergill CAI, Clynekirkton 1-2, Little Ferry Links 1 SUT, Benbecula, Grantown MOR, Old Deer and Newbiggin Leslie ABD), though it does appear in Covesea Cave MOR, and two of the Wemyss caves FIF. With die exception of the latter, this is a markedly northern distribution, and it is interesting to note that almost a third of the monuments bearing this symbol are in die Golspie area. The beast is the third of the three most common symbols, Alcock lists 25 Class I examples alone [1989:9]. Though the beast most commonly faces to the right, there are sufficient exceptions that little significance can be read into the fact that the Golspie beast faces left. Murray considered this example a long way from the Class I prototype [1986:243-9].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
305
The left-hand side of the middle portion is completely taken up with a single male figure in profile. He is bearded, has a pointed nose, wears his hair long, and is dressed in a short (above-the-knee) tunic belted at the waist. He steps forward with a knife in his left hand and a distinctively-shaped axe in his raised right hand. His fingers and thumbs are correctly rendered as they grip the weapons. He is instantly reminiscent of the famous Rhynie Man [Shepherd & Shepherd 1979], who differs from him only in not bearing a knife and in having his axe resting on his shoulder rather than held menacingly out in front of him. Also Rhynie Man's fearsome teeth and specific tonsure are depicted in detail as befits his much greater scale [the two are illustrated side by side in Ritchie 1989:6]. Most commentators have described die Golspie figure as battling with die creature immediately in front of him. Joanna Close-Brooks has gone so far as to suggest that he represents pagan man confronting Christianity in die form of die lion of St Mark and the fish (Christ) [reported in Sutherland 1994:199, see also Ritchie 1989:6]. This seems unlikely. While there are a few examples of a figure struggling with another figure or an animal of some son (e.g. Shandwick ROS, Glamis 1, St Vigean's 1 FOR, certain of die Meigle stones PER) coherent scenes are rare. Henderson has emphasized die eclecticism of Pictish sculpture and die cumulative nature of its message [1967].
It seems far more likely that die separate symbol pairs are to be taken
independently. There are a total of six incised single figures in Scotland (Rhynie 3 and 7 ABD, Collessie Man FIF [ill. Sutherland 1994:174], Balblair INV, Strathmartine ANG, and the newly discovered figure from Cunningsburgh/Mail SHE [Turner 1995]; 1995]). Turner discusses the very close similarities between diem and quotes Isabel Henderson's view that there was an early prototypical 'standing profile man' [Turner 1995:323]. Since these figures do not appear in pairs with other 'genuine' symbols, they are being used in a slightly different way, but they do seem more like symbols than portraits [see Forsyth 1996a for discussion of definition of 'genuine' symbols].
The area opposite him is taken up with a quadruped with extravagantly curled tail above a fish (both facing die man), above a pair of symbols set side-by-side - a 'flower' and a crescent and V-rod. The quadruped is very like die lion of die Book of Durrow [Henderson 1967:122-4]. A similar animal appears in die Fife
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
306
caves and on three Class II slabs (Bressay SHE, Ulbster CAI, and Mortlach BNF), but never on Class I. Golspie, however, has a decidedly canine air, and ought to be compared with the three Class I wolves (Ardross 1 ROS, Newbiggin Leslie ABD, and Keillor FOR, no Class II examples). The mutually exclusive distribution across the Classes, and similarities between the Class I animals and the Durrow lion, outlined by Henderson, suggests to me that the Class II lion' and die Class I wolf may be two varieties of the same symbol. The wolf would have been familiar to the artist: Sutherland notes the last wolf in Scotland was killed just north of Golspie in Lothbeg around 1700 [ 1994:98]. Apart from one of the two fish at Inchyra, all Class I examples face right, as do the fish on the Class II slabs from Ulbster CAI, Glamis 2 FOR, and Meigle 1 PER. The only exceptions are the left-facing fish on St Vigeans 1 FOR, and Latheron SUT, both of which are gripped by left-facing birds of prey. Golspie is unique in having a lone fish face left. The fish occurs also in both Covesea MOR and the Wemyss caves FIF. There are three Class I examples of the 'flower' (Craigton 1 SUT, Pabbay, andDunnichen FOR) and four on Class II (Ulbster CAI, Golspie, Glamis 1 FOR, Gask PER). It too occurs in the Fife caves. The label 'flower' is entirely conventioanl, no satisfactory explanation has been put forward for the meaning of this symbol. The Crescent-and-V-rod is the most common of all the symbols. Alcock lists 52 Class I examples, the majority of which are north of the Mounth [E. Alcock 1989:9]. The Class II examples are more evenly divided [for a discussion of the various forms, see Murray 1986:226-35],
At the bottom of this face is a large 'double-disc' above a pair of intertwined snakes. The double-discwith-Z-rod is the second most frequent symbol. Without the Z-rod it is much rarer. It occurs in several of the Fife caves, and is more common on Class II (Ulbster CAI, Shandwick ROS, Monifieth 1, St Vigean's 4, Woodwray FOR, Dunfallandy, Fowlis Wester, Meigle 3, 6 PER) than on Class I (Drambuie 1 INV, Fyvie 2, Logie Elphinstone 1, Newton ABD, Inchyra). While the single snake without Z-rod is not uncommon (Class I Kirtomy SUT, Knockando MOR, Aberlemno 1 FOR; Class II Ulbster CAI, Glamis 1, St Vigeans 2 FOR, Inchyra), to have two snakes together is unprecedented, though they do seem to be functioning here as symbols rather than as merely decorative interlace. One of the serpents has a zig-zag
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
307
down its back, the other has a straight line. The head of the latter is rendered in profile, that of the former, from above with both eyes bulging. Both have ears and fish-tails and their jaws around the other's tail.
The Ulbster and Golspie slabs have already been linked as both having an exceptional number of symbols. Given that some of these are not particularly common, it is striking that both the symbols on the front of the Ulbster stone, and five of the eight on the back, appear also on Golspie. The only one of the Golspie symbols which does not appear on Ulbster (taking the serpent pair as equivalent of Ulbster's lone snake) is the 'rectangle'. A 'stepped rectangle' appears on Ulbster but the two are separate symbols. It is notable that neither slab has the mirror or mirror-and-comb.
Sides The two narrow sides of the slab are carved with a spiral border pattern (No. 1045), composed of one row of double spirals connected by S-shaped curves, i.e. an inter-locking chain of S-shaped curls. This simple arrangement is found on a number of Irish and Pictish crosses, and indeed much further afield. It is very deeply set between continuous roll moulding along both edges. On this plain moulding that the ogham is carved.
DESCRIPTION OF CARVING - OGHAM INSCRIPTION Macalister described the Golspie ogham as 'one of the most difficult of the entire series to decipher, the difficulty being due to the carelessness of the craftsman who, quite clearly, did not understand what he was doing' [1940:206]. He contrasted the 'admirable cross' with die, in his opinion, slap-dash inscription in which 'relative position of letters become progressively more careless' [208]. Macalister is, however, quite misconceived in this opinion. Not only is the inscription very well carved indeed, but the letters are wellspaced, generally explicitly placed, and, with a few exceptions due to wear, very legible. Padel is quite right that the text 'can be read with some surety' [1972:94]
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
308
Individual strokes are carved with assurance. The technique has produced deep and well-defined strokes, broad at the surface but pointed at their base. The inscription begins 420mm from its current base (the plinth), about 270m higher than the other carving. The preceding portion is intact and absolutely clear so there can be no doubt that the inscription starts at / [see fig. for key to numbering]. The ogham reads leftto-right vertically up the stem. This would have been straightforward to read if placed on the left edge of the reverse face, the reader facing the reverse (symbol-inscribed) face with head turned down to the left (the usual stance). But placed as it is on the right edge, this is awkward to read unless one comes round and stands at the narrow edge, in which case it is perfectly straightforward until one comes to the final portion along the top of the slab. The choice of the right over the left arris is baffling, the left is clear and would have served perfectly well. The inscription turns the comer at the top of the slab and continues across the upper edge of the slab, stopping about 60mm from the end. From the top of a ladder (or the upper gallery of the museum) this presents no problems to read, but it is not at all easy when standing in front of the stone. The strokes on the h-side of the final portion are clearly visible, but, because the inscription runs continuously up-and-across, to someone standing facing the reverse face, they appear to run 'backwards' (right to left), and upside down (i.e. they appear to be below the stem, in the b-position). The b-strokes lie horizontally on the upper face of the stone and are thus invisible to all but the exceptionally tall! It would be wrong to over-emphasize the obscurity of this portion of the text, even a short person would have seen enough (h-strokes, vowels, spacing for b-strokes) to have had a clue what the text read, especially if they had some idea of the likely meaning beforehand.
A deeply-cut line has been incised about 35-40mm from the vertical arris on either side. This has created the impression of a slightly raised roll-moulding. Along the upper edge there is no extra line, simply the top of the 'rectangle*. The curlicues which sprout from each corner of this symbol neatly delineate the curved comer of the moulding, a perfect solution to the problem. The fish's tail and the tail of the wolf impinge a little on this band, but otherwise it is devoted to the ogham. The b- and h-consonants are sloped and thus their length, c.45-50mm, can be accommodated within the band. M-aicme strokes occupy the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
309
middle two-thirds of the ogham band, the vowels are short cross-strokes, perpendicular to the stem. All the consonants slope and in the expected direction. The whole thing is very nicely spaced. Of particular note is the lack of a drawn-in stem-line. Instead the strokes are arranged either side of the arris which, though not sharp, is ridged in some places. With the exception of the strokes on the bend of the corner (25), there is no difficulty in telling one aicme from another. The inscription reads as follows [see fig. for key to numbering]: 1
A hammer-headed A.
2
Two strokes below the stem, sloping well forward - L.
3
As above - L.
4
One stroke above the stem, sloping forward - H.
5
As above - H. These two (4 and 5) are surely too far spaced to be a single letter (D), especially when compared with 13 and 14 which are beyond doubt as DD. The inscription is generally carefully spaced so precision is to be expected here.
6
A single angled vowel stroke pointing in direction of reading ( > ) - A.
7
Two strokes below the stem, sloping forward - L.
8
As above. These two letters slope a little more acutely than 2 and 5.
9
Two short vowel strokes perpendicularly across the stem - 0 .
10
Five oblique strokes across the stem, sloping forward - R.
11
As above - R. The b-portion of the second and third strokes is damaged but the letter is not in doubt.
12
This letter consists of two opposed angled vowel strokes (•> <•). This character appears also on Burrian, and Formaston, and probably on Lunnasting. Since Burrian also has an x-foifid this •> < with a pronounced gap is probably a separate character. What it's value might be is not clear, though in each case a vowel is more likely than a consonant, given context in each case. A character ->*>• would be 0 , -x-or >^r would probably be E, so, at a guess->—<-may be some modified form of one or the other.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
310
13
Two strokes above the stem sloping forward - D.
14
As above, at a slightly greater slope - D.
15
One long oblique stroke across the stem sloping forward - M. The distal end is lost on the b-side.
16
This area is contused. The b-surface is worn away, and the h-surface is damaged though not as severely. There could be single short vowel stroke across the stem (A). Padel thought perhaps a V-shape touching the stem, which he suggested might be the upper half of an x-forfid (E).
17
Five strokes above the stem, sloping strongly forward - Q. The first stroke of the group is longer at both ends. It definitely protrudes beyond the edge of the moulding and also crosses the stem slightly.
18
As above, and at same angle - Q. The first stroke is elongated in exactly the same way as 77 which indicates that this is a deliberate device, almost certainly, as Padel suggests to demarcate the two groups.
19
Five strokes below the stem, sloping forward slightly - N.
20
Six short vowel strokes perpendicularly across the stem. Padel and Macalister both thought them evenly spaced, and an error for five (I), as at Burrian. Macalister thought perhaps AI. Both noticed a slight indentation between the third and fourth strokes. Macalister thought this flaw natural, or due to weathering [208]. Padel was more inclined to think of it as a deliberate device added to separate two letters originally carved too close together (cf. Lunnasting 16117) and this idea has some merit (UU). My initial impression was that the six stroke was quite clearly separate from the first five, which would imply IA. Padel is right to say this stroke lacks a middle-part, but this is clearly just the result of wear on the arris, rather than any mistake by the carver.
21
Three strokes below the stem, sloping forward - V. Padel thought the tip of the first stroke was bent perhaps into some form of a hammer-head A. In fact the horizontal ponion is just wear. At this point the ogham bends round the corner of the slab. Though the comer flattens out a little, there is no real doubt over the reading.
22
At this point there is a single stroke above the stem which may cross it slightly - H or A. It may
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
311
be a letter in its own right, though H would be difficult to account for unless it had vocalic force. Since there is potential confusion between two successive letters of the same aicme (21123), especially since they are squeezed in round the curve, it may be that this is simply an little extra stroke designed to separate the two, without phonetic value of its own. 23
Five strokes below or across the stem - N or R. Padel swithered between the two options and decided in the end that these were m-strokes (R) and that the shortness on the h-side was due to the misalignment of the inscription after turning the corner. Most previous authorities have read 23 and 24 as RR. My initial impression was N, but on balance, since the arris is fairly flat in this corner area, R is perhaps the more likely.
24
Five strokes on the stem. If these were perpendicular to the stem they could be I, but since they clearly slope forward it is probably best to follow Padel in reading these as R. They are very similar to 27 which is to be interpreted RR. As Padel points out, the b-side of the last two strokes is lost to wear.
25
Four strokes above, and perpendicular to the stem - C. The proximal end of the first stroke is lost in wear. Given their gradient it is understandable that Padel thought these vowel strokes (E), however, they scarcely cross the stem, certainly aren't balanced on it, and the distal ends of the b-aicme come right to the very edge of the moulding (contrast 20).
26
In the following portion the surface is flaked and worn and then comes the metal clamp which holds the stone upright in its base (an ingenious arrangement suspended from a girder running between the opposite balconies). Three or four strokes have been lost, but since this portion was already worn before the clamp was put in position (as is clear from Stuart's drawing) little would be gained by moving it.
27
The final section consists of nine strokes above the stem preceded by the possible remains of single stroke. Padel reads ten, which I'm sure is correct. We have here, then, two five-stroke letters of the same aicme, the question is, which aicme ? Padel thought them too far over to be NN, the preferred reading of all previous authorities and this is definitely the case. He considered
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
312
QQ impossible because of the direction of slope, again this is probably the case. For him the most likely interpretation was RR, which, Padel notes, was rejected by Southesk to accommodate his reading -ANN.
Pans of the first two strokes and the b-side distal tips of the final four are lost
due to wear. Which gives the reading: ALLHALLORREDDM(a/e)QQN(i/ia/uu)V(h/a/)(rr/ni)(c/e)(?)(rr/qq) or, less tentatively, ALLHALLORREDDMeQQNiaVarrc(e?)rr or Vanicerr.
FORM OF SCRIPT One of the most interestingfeaturesof the Golspie ogham is that it is written without drawn-in stem-line on the arris of the stone. While this is the norm on otherwise undressed and uncarved pillars, it is unique on a highly decorated cross-slab. Since Golspie cannot be as early as the seventh century and is probably no earlier than the middle of the eighth, this is an important reminder that the lack of a stem-line is not necessarily indicative of an early date.
The consonants are straightforward sloping strokes, with no special characters. The vowels show more variety. In addition to the simple straight stroke vowels, there is a hammer-head A, an angled A, an xforfid, and, if, as seems likely, it is a separate character, the > <-forfid. Successive letters of the same aicme are well differentiated, in the case of 7 7 and 18 the initial stroke of each group is prolonged to further demarcate the individual letters. Successive letters of different aicmi are not particularly spaced, and the last two, 27, are run together, since however 5+5 is the only possible reading, this is not a problem. Except for the absence of stem-line, the script of Golspie is mostly closely comparable with that of Formaston and, to a lesser extent with Birsay 1 (seventh century), but also, though they lack angledvowels, with Latheron and Buckquoy (second half of eighth century?).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
313
INTERPRETATION OF TEXT Padel complains that Golspie appears 'long and complex', though at twenty-eight characters it is roughly the same length as Newton and Formaston, and considerably shorter than Bressay and Brodie. It appears to conform to the familiar syntax X MAQQ Y (cf. Latheron and St Ninian's Isle). The presence of MAQQ or MEQQ gives a clear indication as to preliminary word division, and most authorities are agreed in taking ALLHHALLORREDD as a separate word. The portion following QQ may be a personal name, that of Allhhallorredd's parent. The sections both before and after the MEQQ are, however, rather long, and it may be that further word-division is required.
ALLHHALLORREDD ALLHHALLORR has been interpreted by many (e.g. Nicholson 1896:7) as Norse. Padel suggested a late Icelandic form all 'very' + hallr 'leaning' [1972:98]. This seems chronologically impossible. He comments that the syllable -EDD occurs elsewhere in the Scottish ogham corpus, and, following Macalister, suggests that it could be a genitival ending, 'though one should not necessarily look for IE categories in a non-IE language' [1972:98]. If the -edd is to be separated off, it may be compared with Ackergill's Nehtet, Brandsbutt's Iratadd, Bressay's Nahhtwddadds* and Latheron's Dunnodnat, but since it is a common enough conclusion to both Goidelic and Brittonic personal names, it seem scarcely necessary to look for an otherwise unattested declensional ending. If the segmentation is, rather, -red one wonders if what we have here is an Anglo-Saxon personal name. Perhaps Allhhalloredd (Alhalored) is an attempt to render Aethelred.
A handful of Anglo-Saxon names appear in North British sources in Celtic
orthography, e.g. Golistan (Uulfetan) in YGododdin [John T. Koch pers. comm.] and Dectotric (Theodric) in the Pictish King-List [Anderson 1973:246 a 72].
The suggestion is not so far fetched when one
remembers that Northumbrian dynasts chose on occasion to spend their exiles in Pictland, presumably some did not live to return home. The strong Anglo-Saxon influence visible in Pictish an cannot be solely the result of the Northumbrian masons sent north by Ceolfrid at the beginning of the eighth century [Bede HE V21].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
314
It need not, however, be necessaiy to go outside the Brittonic world to find an explanation of ALLHHALLORREDD. The final two syllables, -orredd (-ored), may be compared with the common Brittonic male personal element -glared (guored, uoret), * protection, bulwaric, defense' attested in Welsh, Cornish, and Breton [Lloyd 1888:47]. The element also occurs on its own, OB. Uuoret, OW. Guoret, Pictish Uoret (Drosten Stone) and Uurad (King-list). The only thing to be accounted for at Golspie is the lack of an initial labial element. Taking the previous syllable with this yields Allorred, (ALored\ which may be a rendering of the well-attested Elguored (var. Elguaret, Elguoreu Elwared) formed with another common personal name element, the intensifying prefix El [Lloyd 45-6, no etymology given], Al for El may be explained by Pictish vowel affection [Jackson 1955:162] or Gaelic influence, cf. MpinlElfin. The name Sulworet is attested in Breton sources [Lloyd 47]. Indo-European s- survived in early British, but became h- in Welsh Cornish and Breton [Jackson 1953:513 § 113]. This still leaves the first syllable All. Some kind of reduplication is possible, if unlikely. Perhaps it is meant as a separate word in which case Irish ail 'rock, boulder, grave-stone, monument' [DIL] springs to mind. This noun also appears as a personal name element [Uhlich 1993:148-51].
A Brittonic name in an Irish language inscription is not improbable in this ethnically mixed area. Dorian claims that the phonology of East Sutherland Gaelic shows that the Dornoch area was Gaelic-speaking at the time of the arrival of the Norse [1981:11] which suggests that Gaelicization was well underway during the ninth century. If the text is in Irish and the first word All is ail 'monument', then we would expect the following name to be in the genitive. Since Pictish had lost declensional endings by this time [Koch 1983] the oghamist could have supplied a plausible Irish genitive, or simply left it as it was. The latter appears to have been the case.
MAQQ NIA We come now to M(a/e)QQ. It is a great pity that 16 is lost to wear and the vowel is in doubt. MEQQ
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
315
occurs unambiguously at St. Ninian's Isle, and possibly at Blackwaterfoot 1, and is best explained either as a dialectal variant of this formula word, Irish MAQQI, or an attempt to render the genitive make, 'son'. The next word is of particular interest. Nia occurs in Irish with two meanings, 'warrior, champion' and 'sister's son' [DIL, see also Vendryes 1960 s.v.]. These are either two distinct words which have fallen together [Mac Neill 1909], or, perhaps less likely, one word which has given rise to two separate meanings [Thumeysen 1912:185]. In his discussion of the nephew-uncle relationship in Irish literature Ó Cathasaigh demonstrates the semantic overlap between the two words [1986:141]. Both words, NIOTA and NET(T)A, occur in Irish ogham inscriptions as the first element in compound personal names (cf. Mac- names), for instance CIIC 271 NETA-VROQI > Nad-Frolch [see McManus's discussion 1991:109-10 §6.15]. The form Nia- also occurs as the first element in a number of Old Irish personal names [CGH s. v.], and when followed by a genitive or adjective Nia occurs frequently uncompounded as a sobriquet or even personal name [DEL]. We are left then, with a number of options. Given what follows, it seems unlikely that Golspie's NIA is the first element of a compound name, we may then be justified in taking it literally to mean 'champion' or 'nephew'.
McManus gives one example of ogham NIOTTA as a formula word
meaning 'nephew', CIIC 252 DUMELIMAQIGLASICONAS NIOTTA COBRANOR[... [110]. That an uncle-nephew relationship should be commemorated on an Irish monument should allay fears that Golspie's text might constitute evidence for Pictish so-called matriliny (though in the case of the latter it would be a grand-uncle-grand-son connection). In discussing the aforementioned Irish ogham, Charles-Edwards [1971:120] and Ó Cathasaigh [1986:144-5] have put forward the interesting theory that GLASICONAS is to be interpreted as cú glas, an Irish legal term for a foreigner, and that, as the son of an immigrant Dumel had to express his kindred membership through his maternal uncle [McManus 1991:110]. If Golspie was erected in a Gaelic-speaking milieu Hallorredd's Brittonic-looking name might indicate that his father was a Pictish-speaking outsider, in which case he might wish to express kin-group affiliation as 'the son of the nephew of X' (for an example of a father with a Goidelic name and his son with a Brittonic name recorded on a bi-lingual ogham-roman monument, see CIIC 488, Tavistock 3, Okasha 1993:278-81, Thomas
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
316
1994:265-6]). There is an alternative explanation. Ma also appears as the second element in the male personal name Macc-Ma. According to the Annals of Ulster (s.a. 1034) MaccNia Ua Uchtcun, lector of Kells, was drowned en route to Scotland. Whatever the explanation, once again, the apparent lack of a genitive is troubling, the expected form would be nioth.
VARRQJRR There is an element of doubt about the final section of the text, and several interpretations are possible. VARRC[.]RR is the most obvious reading, if transliterated FARRCARR which may be a form of the Irish male personal name Ferchar ( < fer 'man' + cha(i)r 'loving* = 'friendly') [Uhlich 1993:242,245 for the separate elements]. This name 'occurs occasionally in the early period' [Ó Corráin & Maguire 1990:96] and was panicularly popular among the Cenél Loairn. Fercar is the name of a number of figures in the genealogies, including kings, such as the famous son of Connad Cerr [see index of Anderson 1922 for further details of Fercars with a Scottish connection]. The Cenél Loairn had increasingly strong connections with Moray, which makes the use of the name by the dynasty particularly pertinent to an inscription just north of the Dornoch Firth. Is there any hope of identifying the Fercar of the inscription with a historical figure ?
The name was popular in the region (in the thirteenth century it was born by Ferchar Maccintsacarit, Earl of Ross, active in 1235 [Anderson 1922.H: 233, 404, 458, 496]) but its use in medieval Scotland was widespread [Dorward 1995:91-2], The name is still in use in the Highlands, Mod.ScG. Fearchar, anglicized Farquar, from which comes the well-known surname Farquarson. According to Watson the disused burial ground opposite Shiel school in Glenshiel ROS is called Cill Fhearchair, though no saint of that name is recorded in the Calendars [1926:304] and it is always possible that the Ferchar commemorated was a layman whose patronage endowed the ecclesiastical site.
Given the meagre
documentary material available, there is little hope of identifying the Ferchar mentioned at Golspie. Though it may be worth noting that a Ferchar m. Feradaich m. Fergusa appears in the Genelach Clainde
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
Lulaig in the Books of Ballymote and Leinster [CGH 162 e 11].
GOLSPIE
317
He is none other than Fercar Fota
('Fercar the long'), who was king of Lorn and, for a year, of Dál Ríada, and who died in 697 [Anderson 1973:111-2, 179]. He had sons and grandsons who appear to have been kings of Dál Ríada. One of these, Selbach, died in 730 [Anderson 1973], so a son of a nephew might be expected to last into the second half of the eighth century, about the right date (and the right status) to be honored with a monument at Golspie (after his death or during his lifetime after an act of patronage). At what date before the ninth century the Cenél Loaira first began to take an interest in Moray is not known [Crawford 1995:3], nor exactly how far north their influence extended. At the time, 'Caithness' included modern Sutherland, fairly near the southern border of which is Golspie. While we would expect a figure of considerable status to have commissioned the Golspie stone, and the ogham clearly points to ultimately Irish influence, so few Fercars made it into the historical record that it would be unreasonable to seize on one of the few who did purely because he did.
One problem with the interpretation Fercar is, of course, the A rather than E in the initial syllable. The name should appear in the genitive, but since the ante-penultimate letter is lost we cannot be sure if it does or does not. Assuming the missing letter or letters to be a vowel, there is too much room for a simple straight, or hammer-head A, though an angled A would fit in nicely. There would just be room for a straight E, though not enough for the six strokes that would be needed for AI, the anticipated genitive form. If the reading FARRC[.]RR is rejected in favour of something along the lines of FINC[.]RR, then we may have the Irish male personal name Finchar < fine + char 'loving one's fine (kindred)' cf. Venicarus [Uhlich 250]. The problem of the apparent lack of genitive case still applies.
Putting these elements together, how are we to interpret the text ? If the formula is X MAQQ Y, then we day have an Irish text with one Brittonic and one Goidelic personal name ail AloredMac Nia Fercar 'the nibnument of Alored son of the nephew of Fercar'. Though nia does appear as a formula word in one Irish ogham, the formula 'son of the nephew o f is not otherwise attested. Perhaps more likely is that Nia and
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
318
Fercar are pan of an otherwise unattested compound personal name NiarFercar, in which case the text is simply 'the monument of Alored son of NiaFercar'. If Mac and Nia are to be taken together as a personal name then we have a string of three ail Alored MacNia Fercar, and the text reads, either 'the monument of Alored, MacNia, and Fercar* or 'the monument of Alored: MacNia and Fercar*.
The latter
interpretation would alleviate the problem of the lack of the genitive, and calls to mind the three names on the Drosten Stone (St. Vigeans 1 FOR), Drosten/ipe Uoret/ett For/cus [Okasha 1985, Clancy 1993]. A number of Welsh monuments with Latin inscriptions carry single names, e.g. Nash-Williams 1950 Nos. 206, 159, 193, and 194. Others explain that X put up the cross for his soul and the soul of Y (and Z, A, B .„), Nash-Williams Nos. 220, 223, 231, 233. One from Merthyr Mawr, Glamorgan [No. 239] adds the name of the craftsman a me preparatus + Sciloc, and No. 222, from Llanilltud Fawr, has a main text and then four other names each in a separate panel, with no fiinher explanation of the role of the individuals concerned. The three names at Golspie could be three individuals commemorated equally, or the name of the patron (secular), patron (ecclesiastical), and craftsman, as is found on later Irish crosses such as the Cross of the Scriptures at Clonmacnoise [CIIC 849] or the Market Cross at Tuam [CIIC 523].
DISCUSSION Despite the doubt over a few of the letters, and the perennial problem of how to interpret non-standard vowels and the first character of the h-aicme, the Golspie inscription is one of the more legible and penetrable of the Scottish monumental oghams. It is unusual in having no drawn-in stem, and in being placed up the right hand edge, though the inscription on Scoonie, which does have a stem-line is similarly placed. All the ogham-inscribed Class II cross-slabs have the inscription on the reverse (i.e. non-cross) face, with the exception of Latheron which is carved on one face only. In scale and workmanship, Golspie is perhaps most closely comparable with Brodie, in both cases the ogham is separated from the rest of the carving and placed on a special band. There is no reason to doubt that these bands were intended to bear ogham from the outset and that the inscriptions relate to the original purpose of the slab. It is, of course, impossible to rule out the possibility that the Golspie text was added later, though there is no positive
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
319
reason to believe so; it is a monumental inscription, carefully conceived and executed, not a piece of graffiti.
Nothing conclusive may be said about any possible relationship between the inscription and the rest of the carving. Attempts to link the figure of the man to any of the people named in the text appear to me entirely misconceived. It has been argued that Pictish symbol statements are most likely to refer to personal names [Samson 1992, Forsyth 1996], and if there were three pairs to go with the three names, then it might be worth trying to match them up. But it is no more than an assumption that these are bilingual monuments such as occur in Wales and Dumnonia, it could equally be that the ogham is used to express things which could not be rendered within the native Pictish system.
BIBLIOGRAPHY OS Record card - NC 80 SW 13 Dunrobin archives - DUOCM ARC 527 Cordiner 1780:72, 1788:n.p.; Stuart 1856:12 pl.xxxiv; Brash 1879:363; Southesk 1884:193, 1886:22-30; Ferguson 1887:151-3; Rh?s 1892:288-90; Nicholson 1896; Rh?s 1898:368; ECMS 48-50; W. Baiinerman 1908:347-52; RCAHMS 191 la: 101 No.295; Curie 1940:85 pLxxxviii; Macalister 1940:206-9, pl.iv b; K. H. Jackson 1955:140, 141 pi. 12; Padel 1972:93-8; Close-Brooks 1989:14, 16 ill.; Ritchie & Fraser 1994:18 ill.; Nicoll 1995:139 fig.27 (ill. only).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GOLSPIE
PREVIOUS READINGS
Southesk
ALLDALLDQQA(a/r)DDMQQNUUFFHRRIaNN
Rh?s
ALLHHALLORrEDDMa(hc/q)(hc/q)N(o/u)UVv(h/a)Rr(eirf/iann) [='Beast and MacN's conflict']
ECMS
ALLHALLORREDDMa(qq/hchc)NUUVVHRR(e/i)(i/a)(rf/nn)
Macalister
ALLHHALLORR EDDARR NAIVALARRENN
Padel
ALLHALLORREDDM(a/e)QQN(uu/i)V(v?a)LA(hr/n)RE[-](rr/nn)
Close-Brooks
ALLHHALLORREDDMeQQNiVvhrrErr
320
c p
•I 55
i CO
"2
GOLSPIE - Ogham-inscribed cross-slab [Close-Brooks 1989:cover, 9]
GOLSPIE - Ogham-inscribed cross-slab. Side view showing ogham [ECMS 48A]
*^ ^H ~ ^*
1
fe
22 21 20 19
18
Kf 11
GOLSPIE - Ogham inscription (key to numbering)
321
GURNESS
DISCOVERY The Gumess knife was found in 1931 during excavation of the Broch of Gumess at Aikerness on the mainland of Orkney [NGR HY 3818 2685]. There was no exact record made of context or stratigraphy but it is known that the knife came from the post-broch occupation layers in the southern part of the site in the area of the dwellings known as the 'Shamrock' and annexe* [see fig.]. The knife was given a cumulative site number of GA 31.277. Along with all the other finds from the early years of the Gumess excavation the knife was sent to the National Antiquities Museum where it is now on display (NMS Cat. No. GAA). A brief description of the site was published in the form of a Ministry of Works Guide [Richardson 1948], but it was over 50 years before the Gumess finds were properly published [Hedges 1987|. Hedges pieced together his account from contemporary field-notes and records and from an examination of the extant finds. In his catalogue the ogham-inscribed knife is numbered 252 [Hedges 1987.11:96 fig.2.22 & 118 fig.2.44]
SITE The Knowe of Gumess at Aikerness was listed as a broch in Petrie's 1872 list [1890:93 No.26| but it was not explored until 1929. Over the following ten summers the grass-covered mound was excavated, first under the direction of Hewatt Craw, then, following his unexpected death in 1933, by a number of excavators under the nominal direction of J. S. Richardson [see Hedges 1987.11:1-14 for a full account of the excavation). The site is on a promontory on the shore of Eynhallow Sound which divides the Mainland from the island of Rousay. Gumess commands a hinterland of good farmland. On excavation the mound was discovered to have a 900-year history of occupation. A discussion of the broch village is outwith the scope of this entry, but it appears to have been abandoned by about AD 400 [for a brief, well-illustrated survey, see Fojut 1993:8-15]. After an interval of perhaps two centuries, the site was re-occupied and new habitations, in a distinctive style, were set into the rubble of the old structures. This post-broch settlement
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
322
coincides with the Pictish period [see fig.]. Norse activity at the site is indicated by a tenth century Viking grave.
Gurness is typical in displaying continuity of settlement locus from the Early Iron Age into the Pictish period, and continuity of artefact types between broch and post-broch layers, and between post-broch and Viking. A number of nonetheless distinctively 'Pictish' items were recovered, most noticeably a small slab (0.27 x 0.19 x 0.29m) incised with unusual geometric Pictish symbols [see fig.; Clouston 1937; J. N. G. Ritchie 1969; RCAHMS 1985:14]. In 1967 a small fragment of a second Pictish symbol stone, incised with the mirror symbol, was discovered nearby at the Sands of Evie [HY 3723 2640]. It is now in the Tankeraess House Museum [OS record card HY 32 NE 31; J. N. G. Ritchie 1969]. The lobate form of the post-broch building known as the 'Shamrock' is of a diagnostic type, the closest parallel being the cellular 'figure-of-eight' structure at Buckquoy [Hedges 42]. The 'Shamrock' appears to have been a dwelling of combined stone and turf construction, behind which there was evidence of bronze-working, including moulds for pins and brooches of seventh to eighth century type. Fojut describes it as a lowroofed, 'diminutive building', and emphasizes thefourth-andfifth-centurydecline in die status of the site [1993:10]. The precise nature of the post-broch settlement at Gumess is not clear. A small cross-marked stone implies adherence to Christianity, but it was discovered unstratified and may possibly date to the Norse period [Hedges 306, ill 126 fig 2.52].
In the official publication of the site Hedges was restrained in his criticism of die standard of the excavation, elsewhere described it as 'wincingly bad' [1990:26]. The expressed aims of the excavation were to clear out the ruins, consolidate the structure for display to the public and to recover finds. Very little attention was given to stratigraphy or proper record-keeping and it is clear that aspects of the site were only imperfectly understood at the time. For this reason we have only a general indication of where the knife was found - in the area of the 'Shamrock' and 'Annexe' - though it is clear that it came from the post-broch layers.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
323
Hedges suggested that an unstratífied cetacean bone casket lid [Hedges 1987:207 No. 174], and a whetstoneshaped pebble [Hedges 1987:245 No.802] might be ogham-inscribed [43] but this seems unlikely [see Appendix - Dubia]. Furthermore Hedges identifies 'putative runic inscriptions' on two short orthostats [GA 77.568, Hedges 1987:245 No.802], but this identification has yet to be confirmed.
DESCRIPTION OF OBJECT A small iron knife with bone handle, inscribed with two lines of ogham but otherwise undecorated [see fig.]. Material:
Handle - bone (species unspecified). Blade - iron.
Dimensions:
Handle - 72mm long x 15mm wide x 12mm thick. Blade - 88mm long x 18mm wide x 6mm thick (max).
Condition:
Inscription well-preserved and clear except for small portion lost near haft.
Better preserved than the Bac Mhic Connain example, this knife is substantially intact, though the tip of the blade has gone and a chip in the haft has carried off a short section of the clear and otherwise complete ogham. The blade is very corroded and, according to Hedges has 'broken off flush with [the] socket' [1987:213]. Hedges described the handle as a 'slightly tapering flattened cylinder' but in fact it is somewhat triangular in section, with two long curved sides meeting at a slight point, and the third a flat narrow base. The trimming of the base has provided two well-defined ridges, but for whatever reason, these were not utilized as stem-lines. Instead the longer of the two lines of ogham is inscribed across the indistinct ridge between the two broad curved faces; and the shorter in an unusual arrangement across the flat base. Thus the two parts of the text are on opposite edges separated by the broader flanks. When the knife is held as if about to make a cut (with the width of the blade vertical), the longer of the two lines is uppermost and clearly visible.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
324
CARVING TECHNIQUE Like that of Bac Mhic Connain, the Gurness ogham is carved on a minute scale. The longest strokes are a mere 10mm in length, and most are only a few milimetres long. They have been lightly cut with a sharp blade which has produced a fine V-section score. Several individual letter-strokes have been cut and re-cut, especially on theflattersurface of the longer text. Macalister suggests the knife is a palimpsest, but I see no reason to believe this. The re-cut strokes are clearly intended merely to improve the definition of strokes imperfectly cut with the first strike. Despite this occasional re-cutting of strokes the inscription is fairly legible. The strokes are evenly and closely spaced and of uniform length. There are clear gaps between letters and the over-all impression is neat. The cramped spacing of some strokes and the apparent need to run-on to a second line suggest a lack of planning.
DESCRIPTION OF OGHAM INSCRIPTION Though individual strokes are clear, there is doubt over certain letters because of the cramped spacing, and in some cases overlapping, of strokes. A further difficulty is in ascertaining the direction of reading of the two pans of the text and their relationship to one another. The longer section covers the full length of the blade and consists of twelve letters, with a possible thirteenth lost with the chip. There is no slope or distinctive letter-form to indicate the direction of reading, though the slight decrease in spacing towards the blade may indicate it was carved, and read, in that direction, as with the knife from Bac Mhic Connain. The shorter section opposite, consists of a further four letters squashed in hard against the socket. These are in the same 'hand' as the longer text and it has been assumed, because of their position, that they represent a 'run-on' line. The question remains, do they read in the same direction as the longer text (towards the blade) or boustrophedon (away from the blade). The latter is unlikely since boustrophedon in ogham is found only on classic Irish pillar stones when the ogham line is continuous and runs up-acrossdown the stone. At Gurness there are two quite separate lines so I think it more likely that they read in parallel. For what it is worth, a more pronounceable sequence is retrieved if the shoner text is read towards the blade. More importantly, it would be hard to account for the over-lapping of these four
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
325
letters, presumably to conserve space, were they not being carved towards the blade.
In what follows I take the longer text first and read from the butt towards the socket. All strokes, unless otherwise stated, are perpendicular to the stem [see tig. for key to numbering]. 1
Five short vowel-strokes on the curve - L These cover approximately a sixth of the width of the ogham band.
2
Five long strokes to the right of the stem ~ N.
3
Four strokes on or above the stem - E or C. It is not clear whether this letter is a member of the a- or the h-aicme since the individual strokes are longer than the vowels / and 4 but not as long as the consonants 2, 5 or 7. They are closer in length to the vowels and at a glance appear to belong to this group. A closer examination reveals that their proximal ends are on the stem, i.e. at the mid-point of the other vowel strokes, but this observation is perhaps reading too much into a cursively carved text
All previous authorities have taken this group as E and this is the
preferred reading. If not merely accidental, the disparity in the length of consecutive vowels may be deliberate, intended to clearly differentiate the two groups. 4
Five very short strokes on the curve - 1 .
5
Three long strokes to die left of the stem - T. Padel notices 'little false starts' on die first two strokes but concludes that these are too 'cramped and haphazard* to be actual letter scores.
6
A single very short stroke on the curve, If this nick is intended as a letter it is A. It has not been mentioned by any previous authority and may be no more than casual damage. Letters 5 and 7 would in any case have to be generously spaced to avoid conflation, so die gap need not indicate a missing stroke.
7
Three long strokes to the left of the stem - T. As Macalister and Padel point out, the third stroke has been re-cut.
8
Four short strokes on the curve - E.
9
A single very long oblique stroke sloping forward across the stem - M. From here to the end of
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
326
the line there is some doubt over the reading since certain strokes appear to overlap. 10
Two short strokes on die curve - O. The first stroke has been duplicated in re-cutting but the two attempts are far too close together to be taken as separate strokes.
11
A single stroke across the stem sloping forward - M. This stroke has its top end level with the previous stroke and its lower end flush with the next stroke, making it shorter by a third than the other M stroke, letter 9. Macalister had difficulty explaining this 'long score' and thought it possibly the second score of an earlier D, the first score being partially effaced. I can see no evidence in support of his palimpsest theory. Padel acknowledged the doubling of the first stroke of 70, but preferred to interpret the 'long score' as two separate strokes, one a third vowel stroke (possibly followed by a fourth), the other a sixth (mistaken) b-stroke, giving a reading EN. His doubt over the number of vowel strokes is demonstrated by his alternative readings UN and ON. There is no obvious solution to this problem. The 'long score' is continuous. It is clearly not itself a vowel, as indicated both by its length and its slight forward slope, which is demonstrable, though not as marked as 9 r s. It is clearly spaced from 12 which in any case has a full compliment of five strokes. If Padel is correct and 11 is a conflation of a vowel stroke and a b-stroke then the reading is more likely to be UB, which is almost as unhelpful for an intelligible reading as the most obvious interpretation - M. It is possible that the stroke is simply a carver's error, an elongated vowel. After a gap the size of those between the preceding strokes there is the faint trace of a line. Padel suggests it may be another vowel stroke, though it seems too insubstantial to be admitted, and, since it joins the stem between the first and second strokes of die next letter, too open to confusion.
12
Five long strokes to die right of the stem - N. Macalister alleged that five vowel points (i.e. I) were 'clearly to be traced' at the bottom ends of the scores of this letter. I could see no trace of them. So, the various possible readings of 10, 77, and 12 are, in order of increasing alleged incompetence on the part of die carver - OMN, OBN, UN, ON.
The fifth stroke is partly lost to the fracture in the surface of the bone. The portion which is missing
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
327
would have been sufficient for another letter, but too much of the surface is lost to ascenain to which aicme it would have belonged. If the shorter section is indeed a run-on of this line then it might be assumed that maximum use would have been made of this last bit of space, in which case we should indeed posit a missing letter. But if letters 13 -16 constitute a separate word, the oghamist may have preferred to leave the space unused and, since it was by now clear that he or she had to start a second line in any case, to begin it with the first letter of the next word. Padel suggested that letters 10-12 were carved overlapping because the oghamist was suddenly aware of the approaching end of the available carving area. It must surely have been apparent to the carver that it would be impossible to fit letters 13-16 in the remaining space. A possible explanation might be that the carver wanted the start of the new line to coincide with a word-boundary and was thus attempting to squeeze in 10-12. It is not certain, however, that these letters do indeed overlap.
The shorter line of ogham is squeezed into the closing millimetres of the handle with considerable overlap of the individual letters. These remain distinguishable because of the unusual three-surface arrangement across the breadth of the flat base. It is unclear why the letters should have been so cramped when the whole of that side was available. If the oghamist did not even allow sufficient room for the final four letters it does indeed imply, as Padel said 'a total lack of planning'. Rather than be written on two surfaces meeting in a notional point (the arris/stem), this short section is written across three surfaces, the two sides being the h- and b- surfaces respectively, and the broad flattened base being a band of 'stem' area. This makes it possible for letters 75 and 16 to over-lap by two and a half-strokes worth without confusion. The vowel-stroke 14 occupies the whole of the width of the base. The only irregularity in these scheme is that the m-stroke, which by analogy would be expected to begin on the b-surface flush with the distal ends of 76, cross the base, and continue on the h-surface to a point level with the distal ends of 75, in fact covers only two-thirds of this distance beginning at the lower ridge of the base (as if it were taking as the stem the arris between the h-surface and base). 13
A single long oblique stroke sloping forwards across the stem - M.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
14
GURNESS
328
A single short stroke across the bottom edge - A. Padel saw a possible, very feint, second vowel stroke giving 0, but I could not see it. The A is entirely beneath the overhang of the preceding M.
15
Three strokes to the left of the stem - T. The last two strokes overlap with the first two strokes of 16. Padel points out that the first stroke of the T has been re-cut.
16
Four strokes to the right of the stem - S.
This gives a reading of: IN(e/c)IT( /a)TEM(o/om/ob/u/e/?)N( /.)MATS
or
IN(e/c)IT( /a)TEM(o/om/ob/u/e/?)N( /.)CFAM or, in the unlikely event that the two sections were read from blade to butt: CFAM( /.)Q(o/om/ob/u/e/?)MEF( /a)FI(e/s)QI
or
MATS( /.)Q(o/om/ob/u/e/?)MEF( /a)FI(e/s)QI
FORM SCRIPT No stem-line is indicated, instead the uneven ridges of the handle are used as an arris. The letter groups are differentiated by position relative to the hypothetical stem-line and by length.
All but the m-aicme
are perpendicular to the stem. The vowels are very shon strokes across the stem, the m-aicme consonants are particularly long. Thus the Gumess knife is inscribed in a form of the script which differs in nothing but scale from the Type I script of the classic ogham pillars of Munster. Among the Scottish oghams, the script of Gumess is most closely similar to that of Gigha and Poltalloch. It differs from Bac Mhic Connain in having shorter vowels and sloping m-aicme consonants, furthermore Bac Mhic Connain is carved in a far finer 'hand* than that of Gumess, or at least with a far finer blade. Gumess is unlike any of the other Orkney oghams and on typological grounds, if on no other, may be the earliest of the group. It is a moot point whether or not the simple form of the script is relevant for dating. There is evidence that simpler forms continued in use long after the introduction of more developed forms so simplicity may be more a function of the humble register of the inscription than anything else. In this regard the lack of a stem is
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
329
perhaps not very significant being dependent instead on the possibility of using a pre-existing ridge in the bone.
INTERPRETATION OF TEXT The question of the direction of reading has already been discussed. Two remaining issues are the order in which the two sections are to be taken and where word division, if any, occurs. The two most plausible readings are INEITaTEMO(m/b)N.MATS and INCITaTEMO(m/b)N.MATS.
The latter might be
segmented INCITATE MONMATS, with incitate the Latin adverb Vehemently', 'rapidly'. There should be no particular objection to the writing of Latin in the ogham script if, as appears likely from the Scottish evidence, the alphabet could be applied to the Pictish or Norse langauges. There is in fact a fourteenth century example of Latin written in ogham characters in the margin of a text bound into the Book of Fermoy [Macalister 1937:60]. Semantics is more of a problem, since incitatus is used primarily of speech, there is no obvious explanation of MONMATS, or parallel for an epigraphic message of this son. The preferred reading is, in any case, INEITaTEMO(m/b)N.MATS, and the rest of the discussion will focus on this interpretation.
When faced with an intractable text it is of assistance to state our expectations of the probability of various explanations. The text is most likely to be written in some form of the Irish language, or failing that in the local language of Orkney in the period. The latter was presumably Brittonic, but may have been partly or wholly pre-Celtic. Norse may be ruled out on chronological grounds. A text of this length may be expected to be a personal name, either with or without filiation, referring to the owner, maker or donor. Or alternatively, an invocation to the object to perform its task well, or to the owner (as, presumably, at Buckquoy). Other possibilities are talismanic texts or playful ones, such as the Dublin runic 'Hart's horn' on a piece of deer antler [Barnes IR 12]. Our working assumption is that we begin by trying to explain the text in Celtic terms and move on to other possibilities only if this attempt fails.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
330
Working backwards we encounter an immediate problem - what is the explanation of the final -S ? Since it follows a T it cannot be an old Celtic genitive ending and there is no other obvious explanation for an S in such a position, unless it is an otherwise unattested reduced form of the Irish enclitic demonstrative panicles -so, and -sa. This explanation would imply that MAT is an indefinite genitive or adjective qualifying a previous substantive [Thumeysen 1946:299 §475]. It would be possible to interpret MAT as an adjective meaning 'good' (Olr. maith). The DIL lists two words mat: one a possibly archaic word for 'hand', the other an archaic and poetic word for 'pig' [s.w.]. Since the knife is certainly a small handknife and might perhaps be made of pig bone neither explanation is perhaps as outlandish as at first sight. Classical latin matara is listed as a Celtic javelin, pike or lance. If there is a Celtic word behind this form it is possible that its reflex is found in Pictish, meaning 'weapon' or 'knife'. Another possibility for MAT is some element of a personal name. In his dicussion of the Irish male personal name Matgen 'well/noblebora' or 'born of a bear', Uhlich gives a number of Old and Neo-Celtic examples of compounds with mat 'bear' or mat 'good' [1993:276-7]. Further Irish comparanda are Mát(a) and Mátach [CGH:22 = 118 b 6; 120 = 136 a 26; 376 = 324 e 5], from Welsh there is, for instance, Matoc and Mathus in the Book of Llandaff [Davies 1979:76b 203b 144; 143] and the famous Math of the Fourth Branch of the Mabinogi.
If MATS is indeed a separate word, then we are left with INEITTEMON- or INEITATEMON- or INEITTEMOBN- or similar (cf. Inchyra's INEHHETE-). Whether this is to be further subdivided, IN EITTEMON- or IN EITTE MON- for instance, is unclear. EITTE is reminiscent of EHTE and ETTE of Cunningsburgh 1 and 2 respectively, perhaps representing different spellings of the same word, or different stages in its phonological development, cf. the spellings of Nectan, Neitano, New. This would leave IN which might be interpreted as the Olr. definite article. Alternatively, in Olr. the prefix in- is used to form adjectives with meaning 'capable of, fit for, proper for worthy of [DIL]. So, an entirely different approach would be to read IN-EITTEM ON-, taking in-eittem as a form of the Olr. adjective in-ethaim 'suitable for tillage' [DIL s.v.], a not inappropriate pronouncement on a domestic or agricultural
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
331
implement. None of the above suggestions are in any way compelling and we may have to admit defeat for the time being, in the face of a doubtful reading and the lack of any clear knowledge of the kind of text to expect.
DISCUSSION Obviously Gumess invites parallels with the other ogham-inscribed knife handles,fromBac Mhic Connain, North Uist and Weeting in Norfolk. Like the others it is an informal personal text 'almost certainly ... made during the functional life of the knife* [Holder 1990:57]. The context, date, and meaning of the Weeting inscription are a puzzle, but the Bac Mhic Connain one is slightly better understood. As far as can be discerned it appears to bear a personal name, presumably the owner, but possibly the maker or donor. The Gumess text is a few letters longer though what it says remains unclear. Until the Gumess text is understood and interpreted there is no possibility of identifying chronologically diagnostic linguistic features. Script typology points to an earlier rather than a later date, but as discussed above, is not a necessarily reliable method. Hedges dated the Gumess knife to the eighth century [1987.111:43] but only on the basis of the received wisdom on the dating of Pictish ogham. The discovery of oghamsfromPool and Birsay in radio-carbon dated contexts in the seventh and sixth centuries mean there is no reason that the Gumess knife should not be as old, if not older. The post-broch occupation at the site has been broadly dated to the fifth to eighth centuries, but on no particularly sound evidence other than a terminus ante quern for the Pictish layer provided by a Viking burial. The Gumess ogham wasfoundin very similar circumstances to those of the Buckquoy whorl - lying abandoned in or around a cellular dwelling. The Buckquoy ogham seems to datefromaround 750 or earlier though it is written in a more developed form of the script. In the absence of definitive linguistic evidence to the contrary there is nothing to rule out a date in the fifth century, so while a vague ascription to the seventh or eighth century might be more broadly acceptable a much earlier date should not be dismissed out-of-hand. Whatever its precise date, the Gumess ogham is further important evidence of the demotic use of ogham in Orkney in the Pictish period.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
GURNESS
332
BIBLIOGRAPHY
OS Record card - HY 32 NE 5
Macalister 1940:218-9 pl.IVc; Padel 1972:98-100; Hedges 1987:213, Cat.No.252, ills. 96 fig.2.22, 118 fig. 2.44; A. Ritchie 1987:48 ill.; Holder 1990:52-58; Fojut 1993:11 (ill. only).
Site:
RCAHMS 1946.11:75-9 No.263; Richardson 1948; Childe 1961:46-7, 106 ill.; Hedges 1987; A. Ritchie 1987:45-48
PREVIOUS READINGS Macalister
INEITTEMON MATS-
Padel
INEITTEMEN (M(a/o)TS-) / (CV(a/o)M)
Holder
(i/u)(n/s)EITTEMON MATS (not based on personal inspection)
u X a
•'
18 24 »
44 f»t
Pitm «f post laoch-pértcxf struettm. 0 1
4
U „ ,>, - J
•
I
« 11 m * «
k»j—i—t
GURNESS - Plan of the Post-Broch period structures [Hedges 1987.11:64 fig.2.11]
GURNESS - Plan of the "Shamrock" and "Annexe" [Hedges 1987.11:66 fig.2.12]
GURNESS Carved stones recovered from the site. Above: Pictish symbol stone. Below: Cross-incised slab, unknown date. [Hedges 1987.11:125 fig.151, 126 fig.2.52]
306 (P8j
0 4 8
cms
GURNESS - Ogham-inscribed knife handle [Hedges 1987.11:96 figs. 2.22, 118 fig.2.44, details]
12 11 10 9 8 7 6 5
3 2 1
Top
Underside
GURNESS - Ogham inscription (key to numbering)
333
INCHYRA
DISCOVERY The Inchyra slab was found on 19 February 1945 during ploughing in former parkland on the Carse of Gowrie, 120m south of Inchyra House, in the parish of St. Madoes, Perthshire (NGR NO 1904 2120) [see map]. For some years previously, the field in question had been under pasture and was turned over to crops only towards the end of die War; it was deeper ploughing in the spring of 1945 which revealed the stone [Wainwright 1961:3 n.2]. On striking the stone, the farmer alened the proprietor Harold de Pass and the then Curator of Perth An Gallery and Museum, James Wood, who came to examine the site. Several weeks later 'when all had been levelled and harrowed' [Stevenson 1959:33], Thomas McLaren, the Burgh Surveyor of Penh, examined the stone and made a record of the discovery. This he read as a communication to the Society of Antiquaries of Scotland in die January of thefollowingyear [n.a. 1946]. He died, however, eighteen months later, before publishing a full account. Stevenson's account of the discovery [1959:33-4], which is followed below, is based on McLaren's typescript notes.
Unfortunately the stone was removed before the nature of the underlying finds was recognized, and it is a great pity that the site was not recorded to modern archaeological standards. Despite the incomplete nature of the record it is clear the slab was covering an inhumation burial. The slab, oriented WNW-ESE, was lying flat about a foot (0.3m) below ground level when found, resting on fony-nine water-rolled stones. These stones, which ranged in size from c.200 x 180 x 150mm to c.300 x 230 140mm, were 'above and around' some human bone: 'the upper pan of a skull (at the west end), an arm-bone and shoulder socket... identified by Mr Wood and then re-interred on the spot at the wish of the proprietor' [Stevenson 1959:34 n.l]. It is not clear which face was uppermost when die slab was uncovered. Stevenson suggests that if some 'erratic scorings* on the rougher of the two broad faces are plough-scrapes then this may have been the upper face. Nothing is now visible at the find-site. The Harold de Pass presented the stone to Perth Museum where it is now on display (Ace. No. 5/45).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
334
Wainwright refers to 'details collected by McLaren from persons present at the time of the discovery [which] strongly suggest that the sandstone slab was not decorated to accompany the human remains with which it was found' [1961:3]. What these details were is not elaborated. At the time the grave was discovered, virtually nothing was known about Pictish burial practice. Wainwright complained in 1952 that 'we cannot with confidence affix the label 'Pictish' to ... a single burial' [1955b:87]. In particular there were no certain examples of symbol stones associated with Pictish graves [op. cit. 96]. Our understanding of Pictish mortuary ritual was transformed in the late 1970s when a distinctively Pictish type of burial monument was first identified [Ashmore 1980] and connections between symbol stones and Pictish graves were first proved by archaeological excavation [Close-Brooks 1980]. If this knowledge had been available in the 1940s interpretation of the find might have been very different. McLaren interpreted the burial as Bronze Age and thus, in his view, the association with the Pictish stone had to have been secondary. Observing that 'the stone had obviously stood erect above the ground for some considerable time before it was buried in a horizontal position over the human remains', Wainwright concluded that the slab had been 'laid over them at a time when its original purpose had been forgotten, perhaps when the burial ... was accidentally disturbed in a more recent century' [ibid.]. This seems rather implausible. More likely is that the slab stood upright beside the grave, fell over in antiquity and subsequently became buried under accumulated deposits; or, alternatively, that having served as an upright monument for a number of years the slab was reused later in the Pictish period as a cover for the grave. It is clear from the extant carving that the stone has been reused on at least two and probably three occasions. It is unfortunate that the prevailing opinion that 'neither the nature of the burial nor its precise location are directly relevant to the problem of the stone and its original purpose' [ibid.] encouraged its discovers to neglect to record more fully die circumstances of its discovery.
When the stone was first discovered only the two bolder sets of symbols were noticed and the larger of the ogham inscriptions (B). Following McLaren's presentation to the Society of Antiquaries, Dr James Richardson re-examined the slab and recognized the third set of symbols, and three further lines of ogham.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
335
It was he who realized the relationship to the main slab of several fragments also discovered at the site, and thus was able to complete further sections of both ogham and symbols. Through the offices of Professor Gordon Childe, McLaren sent R. A. S. Macalister photographs, drawings and rubbings of the stone, and Macalister's letter in response was preserved in McLaren's papers deposited in the archives of the Society of Antiquaries of Scotland [Wainwright 1961:4]. The first full treatment of the stone and its ogham inscription was published by Wainwright in the French journal Ogam [1959], subsequently reprinted as a monograph by Dundee Museums [1961]. R. B. K. Stevenson discussed the stone at some length in an article in PSAS having examined it in the company of Kenneth Jackson whom he reported as being 'in general agreement' with his analysis. No interpretations of the ogham texts were offered by either author. Oliver Padel made a number of linguistic observations on the Inchyra texts in his 1972 dissertation, but offered no coherent interpretation.
LOCALITY The second element in the name Inchyra has not been explained, but the first, 'inch' (innis) refers to a haugh, or low-lying piece of ground by the bank of a river, liable to flooding. This is most appropriate to the site, lying as it does by the north bank of the river Tay. No other archaeological remains are known from the immediate vicinity except for a mound, marked 'Witch Knowe' on sheet 37 of the OS first edition, about 110m west of the burial [Stevenson 1959:33]. It has not been explored and might date to any period. More generally, this important area, in and around the mouth of the Tay, is rich in early medieval remains. There are hints of early connections with Ireland, including the famous reference in the late sixth century Irish poem Amra Choluimb Chille to Columba teaching the tuatha Toi, 'the tribes of the Tay' [Clancy & Markus 1995:104]. Nearby St. Madoes, which has given its name to the parish, may represent a dedication to Maedoc [Simon Taylor pers. comm. pace Watson 1926:327]. There are several saints of this name, but the most prominent is Maedoc of Ferns. As Dr Taylor points out to me, a possible dedication to this major Leinster saint takes on added significance in the light of traditions linking Abernethy with the greatest of Leinster saints, Brigit [see entry on Abernethy]. Abernethy is almost
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
336
directly opposite St. Madoes on the south bank of the Tay and Dr Taylor informs me the two were still linked by a ferry as recently as the last century.
DESCRIPTION OF OBJECT A long, thin, tapering, rectangular slab incised on both broad faces, and three narrow faces with three pairs of Pictish symbols and three ogham inscriptions (arranged in a total of five lines). Stone:
Very fine grained local sandstone (mica-rich, containing both biotite and muscovite, according to Dr A. G. MacGregor, Geological Survey [quoted by Stevenson 1959:34])
Dimensions:
Length: 1.6m (5'3") Breadth: narrow end - 0.31m (12"), broad end - 0.43m (V5Hn); Thickness: 70-90mm (234-3V$")
Condition:
Narrow end intact but broad end broken, and a portion missing. Areas of the surface have flaked off, especially along the edges, taking with them pans of the carving, though some of the larger of these fragments have been recovered. The extant carving is generally well preserved. Some sections of ogham have been lost, but only A (ii) is severely damaged.
Richardson noted that one surface was smooth and undulating, die other flaky 'suggesting that it was obtained by cleavagefroma water-worn exposure' [Stevenson 1959:34]. Stevenson points out that all the narrow faces (except the broken one) are flat and smooth, their arris with the smoother broad face being rounded, and that with the rougher face, sharp [ibid.]. The Inchyra slab is entirely atypical of Class I Pictish symbol-inscribed stones in being carved on a very carefully shaped and dressed slab. It is not alone in being reused, there are several examples of that, including Logie Elphinstone 2, Kintore 2 ABD, and St. Peter's, South Ronaldsay ORK. What is exceptional is that it appears to have been reused no less than three times.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
337
The slab is very tall and thin, but not to the extent that it could not have stood upright. Stevenson records that 'part of one side [was] shattered at time of discovery', some of these fragments were, however, recovered, along with a substantial chunk off one edge of the narrow end. There are some horizontal grooves on the surface of the slab, some, including one right across the smooth surface, were probably made when the stone was transported to the museum. Another long groove nearer the broad end 'was unfortunately made to mark the depth the stone was to be set in a wooden stand' [Stevenson 1959:34].
DESCRIPTION OF CARVING - SYMBOLS I will follow Stevenson in designating the smooth surface the 'front', but this is merely conventional and is not meant to imply that the carving on this side is primary. It is also merely conventional to refer, as previous authorities have, to the narrow end as the 'top' and the broken broad end as the 'bottom'. Since the RCAHMS [1994a] illustrate the slab with the narrow end at the bottom, to avoid confusion I will refer instead to the 'narrow' and 'broad' ends. The slab is currently displayed in low lighting on its side in a large perspex box, conditions which made detailed examination of the carving almost impossible. P am, however, most grateful to Dr Michael King of Perth An Gallery and Museum for facilitating my examination of the stone].
'Front'/Smooth face There are two symbol 'statements' on this face of the slab, each at opposite ends of the slab, and each inverted relative to the other. At the narrow end there is one pair oriented with that end as the top, namely a notched double-disc, and a fish. Positioned very close to the top of the slab, they are the most substantial and finely drawn of all the Inchyra symbols. They have been very well carved, 'definite and accurate' to use Stevenson's phrase, using the traditional pock-and-smooth technique. Stevenson thought the smoothing might have been done 'by oblique "chiselling" with the same round-nosed tool' as was used for the pocking [1959:34]. He went on to comment that in the initial pocking 'many strokes landed outside the final quite broad outline. This is particularly clear at the rear of the dorsal fin'. This 'shrapnel' effect
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
338
is not at all unusual, however, and there is wide variation in the extent of smoothing on different stones [Gordon 1956].
At the other end of this face, oriented with the broad end as top, is an indistinct group of symbols which were not noticed at first. The lowest is a mirror and, above that, the remains of a 'tuning-fork' symbol. These carvings have been plausibly interpreted as both incomplete and partially defaced, as if, as Wainwright suggests, the carver had a change of hean before completing the design. The carver has begun to pock the lines, getting further with the 'legs' of the tuning-fork and the upper rim of the mirror, than with the handle, as Stevenson noted, but then appears to have attempted to deface them. Stevenson thought this was aimed at making the symbols as inconspicuous as possible before the half-finished slab was re-used (certainly this is more plausible than his alternative theory that the defacement was the result of a deliberate antagonism). A possible reasonforthe abandonment of the enterprise lies in the flaky nature of the stone. Stevenson hypothesized that the slab's surface had begun to scale away under the pressure of carving and that the carver had therefore abandoned the attempt. Later, when the stone was pressed into service anew, the old half-finished symbols were pocked over to minimize their visual impact, and a different carving technique applied to the reverse. The mirror is in the expected bottom position. While unaccompanied mirrors are common, the 'L' shape surviving on one of the flakes from the left of the mirror may, as Stevenson suggests, be the remains of a very incomplete comb. We would expect a second symbol above the mirror to form a primary pair [Forsyth 1996b]. Stevenson thought one 'probable' to the left of the 'tuning-fork', but that area has largely flaked away and certainty is beyond us. Since the broad end of the slab is broken, a portion of the tuning-fork has been lost.
'Back'/Rough face Though this is the rougher of die two broad faces there has, as Stevenson noted, been some preparatory smoothing of the stone before carving. The carving consists of a single pair of symbols at the broken end of the slab oriented with die broad end to the top. The pair comprises a serpent below a fish, the upper
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
339
pan off which has been lost at the fracture. Though somewhat less well drawn than the first pair described, these do have the appearance of being finished. Their carving technique, however, is quite exceptional for they have been cut with a chisel or knife, to produce a V-section groove, rather than pocked with a point to produce a U-section channel. They were die only example of symbols cut in this way encountered by Gordon in his survey of Pictish symbol carving technique in the eastern Mainland [1956:42],
Discussion of symbols The six symbols are organized into three pairs: notched-double disc above fish, fish above serpent, and 'tuning-fork' above mirror. The first two are classic 'symbol statements', i.e. a pair of two different symbols arranged vertically. As it stands the third group is exceptional because mirrors appear only in addition to normal symbol statement pairs. Since, however, the top of the slab and some of the surface is missing, and the mirror is in standard bottom position, there is every reason to take the extant carving as the remains of a typical symbol statement 'pair + mirror'.
Elizabeth Alcock lists twelve examples of the fish symbol on Class I monuments [1989:9], Allen has a further five on Class II, making it one of die more common of the 'Second Division' of Pictish symbols. Nonetheless to have two examples on the one monument here at Inchyra is striking, especially since, otherwise, the symbol's distribution is markedly northern [see entry on Ackergill]. The combination double-disc plus fish does not occur elsewhere, though a fish with double-disc-and-Z-Rod appears on the Class I stone from Keith Hall ABD (modified by a mirror and comb). The degree to which the rodless double-disc is to be considered separately from the otherwise identical type with Z-rod is not yet clear. The former variety is much rarer, occurring on only four other Class I stones - Drambuie 1 INV, Fyvie 2, Logie Elphinstone 1, Newton ABD - but on ten Class II - Ulbster CAI, Golspie SUT, Shandwick ROS, Monifieth 1, St. Vigeans 4, Woodwray FOR, Dunfallandy, Fowlis Wester, Meigle 3, 6 PER - and in a number of the Fife caves. The example from Newton is of particular note since it has a distinct notch
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
340
taken out of one of its discs, the left hand, in the 'six o'clock' position. Inchyra has an identical notch out of each disc, at 'eleven o-clock' on the left-hand and at 'six o'clock' on the right-hand. Presumably, these notches modify the symbol in some way. Similar notches occur on a handful of other symbols, including the crescent [see Forsyth 1996b].
The serpent without Z-rod is similarly less common than its rodded counterpan, there being only three other examples of Class I (Kirtomy SUT (dubious), Knockando MOR, and Aberlemno 1 FOR) and three of Class II (Ulbster SUT, Glamis 1-2 FOR), excluding all serpents not in proper symbol 'statements'. The combination with fish occurs only at Glamis 2, serpent uppermost, where it is modified by a mirror. Serpents modified by Z-rods tend to be oriented right-left, but unmodified examples tend to be top-down, with the head to the top (Aberlemno being less than vertical, perhaps to fit into the constrained space at the top of the slab). The Inchyra example is conforms to the typical pattern - upper loop to the left, middle to the right and lower to the left with the tail curling round in a spiral - but with the head and neck a little straightened to fit into available space. In common with other rodless serpents, Inchyra's is undecorated, bar its head. The fish in this pair is atypical in having its tail to the right rather than left.
Only the 'legs' of the upper symbol in the third pair survive. These are longer and thinner than would be expected for a 'notched rectangle', and are more plausible interpreted as part of a 'tuning-fork' symbol (what Allen calls 'notched rectangle with curved ends' [ECMS.ii.67] though it is quite different from the 'notched rectangle'). This symbol is seen on a handful of Class I stones with predominantly northern distribution, Craigton 1, Dunrobin, Kintradwell 4 SUT, Roskeen ROS, Kintore 3 ABD, and Abernethy 1 PER, though it is possible that an acephalous symbol from Dairygreen, Collace PER is another 'tuningfork'. The symbol occurs once on Class II, on the cross-slab fragment Tarbet 1 ROS.
RELATIONSHIP OF SYMBOLS AND ORDER OF PHASES The three sets of symbols are clearly the work of at least two, probably three, hands in at least two, and
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
341
probably three» different phases. The order and relationship of the different phases remains, however, unproven. The symbols cannot have all been visible at the same time, since if one end was embedded in the ground only the symbols at the other end would have been fully visible, and if the slab was intended to lie flat only the symbols on the upper face could have been seen. Of course, if the slab were intended from the outset to be buried, then it need not have mattered that symbols were obscured. The difference in draughtsmanship and in carving technique, and the fact that one set is inverted relative to the others does, however, suggest that there was more than one phase of carving involved.
If the slab were meant to stand upright it would be more stable with the broader edge at the bottom. This has encouraged commentators to view the narrow edge as the original top, and the fish and double-disc as the primary symbol statement. Wainwright's interpretation of the mirror and tuning-fork as a false start is appealing, and implies that the fish and serpent represent the third and final phase of use. If the slab stood upright the fish and double-disc would be embedded in the ground and thus obscure, and the imperfect symbols on the front would be unobtrusive thanks to their defacement, thus only the fish and serpent would be visible. This scenario would account for the change of technique for the third set, and the use of the rough face only as a last reson (the narrow end of the rough face is even worse than the broad which is an additional reason for not choosing the narrow end). Since the two sets of symbols on either face of the broad end are different we may take it that they relate to separate phases of use otherwise we would expect a completed tuning-fork and mirror in place of the fish and serpent. It is interesting to note that the two sets of symbols at the broad end are to the same scale, larger than the set at the narrow end, and carved at roughly the same height - the serpent's tail is level with the bottom of the mirror handle.
If the narrow end were embedded sufficiently deep in the ground there need not have been a serious problem with stability, but it may well have been that the later phases involved the slab lying horizontally as a grave cover. The Inchyra slab tapers markedly in width towards one end. If recumbent monuments
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YR A
342
taper they tend to have their broad end at the head. Stevenson records that the broader end of the Inchyra slab was towards the west, which was also the end where the skull was discovered. The only other factor to consider is the breakage at the broad end, which, if it occurred when the slab fell over, most plausibly did so when the narrow end was at the top (leaving the rest embedded in the soil). Since the fracture clearly post-dates the carving of the symbols at this end, it could be taken to imply that the fish and notched double-disc were the third and final phase. The slab, however, is thin and the break could as easily have occurred when the slab was lying horizontally.
RELATIONSHIP OF SYMBOLS TO OGHAMS There are ogham letters in five positions on the slab: across the rough face, along the narrow end, along the adjacent section of one long side, further along the same side, and in the equivalent position along the opposite side. Three different hands using three different carving techniques are apparent. Since there are also three pairs of symbols carved in three different hands it seems natural to associate each ogham with a different set of symbols. In two cases the connection is almost certain, in the third, it is less clear. The shortest ogham (Padel's C, Stevenson's (d), Wainwright's 3) runs without a stem-line up the broad rough face of the slab immediately to the right of the fish's back. There can be little doubt about the association, all the more so since the ogham is chisel-cut like the symbols. With the broad end (and the fish's head) to the top, this ogham reads upwards. The top of the hypothetical stem is a little further to the left than is the bottom, meaning the inscription is slightly to the left of vertical. The tapering edge of the slab makes this appear more marked than it is. Doubtless the strongly diagonal alignment of the fish has exerted a pull on the ogham's orientation. The most substantial ogham (Padel's B, Stevenson's (a), Wainwright's 1) starts midway along one long side, a little below the fish's tail, and runs 790mm (31") up to the narrow end, continuing round the corner and across the top. It thus neatly frames the fish and notched double-disc widi which it shares the same substantial pock-and-smooth technique. With the slab upright, narrow end to the top, and these symbols on the 'front', the ogham runs up the left edge and across the top - the classic ogham position. There is no real doubt about the associations of these two
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
343
oghams, but the remaining two are not so certain. There are two lines of ogham, in exactly the same hand and carved in the same style, along the narrow edges either side of the broad end. These are Wainwright*s 2, Stevenson's (b) and (c), and Padel's A. I will follow Padel's terminology but divide them into A (i) and A (ii).
A (i) starts 45mm (Vhn) below B, runs along the narrow edge for 450mm (\llhn)
in the opposite
direction, leaving 230mm (9 n) blank at the end. On the opposite long edge, A (ii) ends in wear about 90mm (3V£") in from the broken broad edge, and can be traced back for 240mm (9lA") until it disappears in the fractured surface of the slab. Both these slight oghams read vertically upwards with the reader's head turned to the left as one faces die long edge in question. This means that A (i) is on the left when the fish and serpent are the 'front', and A (ii) is on the left when the incomplete symbols are the 'front'. If the slab were lain on its back, rough side down, A (i) would be the right way up, but A (ii) would be upside down. Padel thought these two lines of ogham very tentative and suggested they might have been 'hastily scratched' [103]. He felt they were most naturally associated with the incomplete symbols on the smooth face. Wainwright did not advance an opinion on their association, Stevenson was undecided. He thought the carving of A (i) similar to that of the snake 'and might have been made by a light application of the same tool' [38], but thought a definite association unlikely since C was already linked to that group. On the other hand, he felt 'tempted to associate ... [A(i)J with the unfinished symbols'. He concluded that it couldn't have been a preliminary setting out of a text to be carved more substantially since the strokes were too erratic and close together, but he wondered if it might have been 'a quick note, to be set out better later on'. Padel found this idea appealing, but I think it unlikely. Rather than provide an insight into the working methods of the oghamist, I think A (i) and (ii) are simply informal, casual inscriptions, perhaps even 'graffiti'. I don't think they need necessarily be associated with any particular group of symbols, though their carving in this configuration would have been most feasible when the stone was standing upright with the broad end to the top. I think it likely both were carved at the same time, but this could have been considerably after the carved stone was set up. Stevenson thought A (i) was carved
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
344
after B since the former stans only 45mm below the first stroke of the latter, but I do not consider this conclusive (there is ample space at the far end of A, it could have begun much higher), and, for the sake of argument, it could even be taken the other way round - that B followed A (i), since B occupies all the available space between the end of A (i) and the far edge of the narrow end. A (i) starts at roughly the same level as the bottom of the mirror-handle and the lowest point on the snake's tail.
To summarize, B is associated with the fish and double-disc, C with the fish and serpent, A (i) and (ii) go together and may be associated with either of the symbol statements at the broad end, or both of them, or neither of them, but they were almost certainly carved when the stone stood upright with the broad end to the top. The evidence for the order of carving is ambiguous and the sequence remains unproven.
DESCRIPTION OF OGHAM INSCRIPTIONS
A (i)
(Stevenson (b), Wainwright 2)
This very slight inscription was first noticed by Richardson. It is casually carved with a particularly haphazard stem-line ('extremely erratic1 in Stevenson's words), but is nonetheless legible. The only doubts arise because of lack of neatness. All the consonants slope and the direction of reading is thus quite clear [see fig. for key to numbering]: i
Stevenson illustrates four short strokes across the stem. According to his drawing there is a clear gap between the start of the stem and the first letter, thus the letter can be seen to be complete. There is a very slight forward slope on these strokes, but it is not nearly as pronounced as later xorQicme letters, and is explained by the gentle curve of the stem here (they maintain perpendicularity with it). As Padel makes clear» there is no reason to read 'Z' (i.e. Si) as Wainwright did, the length of the strokes proves them to be part of a vowel - E. Unfortunately the conditions under which I examined the inscription were not ideal, but I wondered if there might in fact be two vowels here - two or possibly three strokes immediately following the crack
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
345
in the surface, then, after a gap, three vowel strokes (UU or OU). This, however, is very uncertain. 2
Fortunately the next section is clear: Three strokes to the left of the stem sloping backwards - T.
3
As above, with longer strokes - T. Letters 2 and 3 are widely spaced.
4
Two strokes to the right of the stem sloping forwards - L.
5
Five short vowel-strokes perpendicularly across the stem - 1 . Padel comments that the third and founh strokes are slightly more widely spaced than the others, but he is quite right that subdivision (UO) would be ftoo subtle'.
6
Five short vowel-strokes perpendicularly across the stem - E. Again, these are well spaced from the previous group.
7
Three strokes to the left of the stem, sloping backwards - T.
8
Five very long oblique strokes sloping forward across the stem - R. The length of these strokes shows that all the previous cross stroke letters are vowels.
9
Four strokes across the stem - E or S. These are definitely shorter than the preceding group, but longer than earlier vowels. They are, however, perpendicular across the stem and all previous authorities have read E.
10
Four strokes to the right of the stem, sloping forward - N.
11
The next character looks like a right angled triangle sitting on its point. It consists of one vowellength stroke across the stem, followed by an angled-vowel-type stroke $>), their b-distal-tips are touching, but not the h-distal. There are at least two ways of interpreting this character, either as one half of the angled-(^character seen at Bressay (see also entry on Formaston for possible phonetic values), or an O made up of one straight and one angled stroke, cf. the similarly mixed E on the back right ogham at Brodie. Either way the character is unprecedented.
12
Five vowel-strokes more or less perpendicularly across the stem -1. That these are rather longer than, say 1 or 5/6, lends further support to the vocalic interpretation of 9. The stem has been sloping down through letters 10 and 11, is virtually invisible at 12 and starts afresh back towards
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
346
the middle at 13. Padel's alternative suggestion of Q makes too much of this droop in the stem. 13
Two short strokes to die left of the stem, sloping backwards - D.
14
As above - D. This section (13/14) is very worn and I am reliant on Stevenson and Padel for interpretation.
15
The next portion is a little clearer. Two short vowel-strokes perpendicularly across the stem - 0 .
16
Five long oblique strokes sloping forward across the stem - R. Padel noted that the first stroke was rather widely spaced from the rest, but rightly concluded that this is not significant.
17
Three or probably four strokes to die right of the stem sloping forwards V or S. Padel commented that the first and fourth strokes were faint almost to the point of invisibility. This appears to be the end of the inscription.
Which gives the following reading: (e/uu/ou)TTL(i/uo)ETR(e/s)N(o?)IddOR(s/v) probably ETTLIETRENOIDDORS
A (ii) (Stevenson (c), Wainwright 2) This inscription is not visible as the stone is currently displayed, thus I have not been able to verify the accounts of earlier authorities by personal examination. The following is based on these previous accounts and on photographs in the possession of Perth Museum [I am grateful to Dr Michael Kingforallowing me to examine these]. Wainwright does not illustrate A (ii), nor comment on the carving, he simply lists his reading. Stevenson quotes Wainwright's reading and provides a line drawing but, according to Padel, it is inaccurate in certain details (a view supported by my examination of the photographs).
Approximately 240mm (9V£") of lettering survives, but both the beginning and end of the inscription appear to be missing and even the extant letters are damaged in several places. Stevenson says there is no trace of lettering on fragments from lower down this edge but does not give details of how far these clear
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
347
patches are from the present beginning. Thus an unknown number of letters have been lost from the beginning. The marked difference in length between vowels and consonants and the fact that all consonants slope gives a number of clues to the interpretation of the extant carving. Unfortunately the stem-line is not particularly clear. There is a very misleading crack running the entire extant length of the lettering but the stem is just visible a few millimetres below. The slope of 2, 3 and 8 indicate the direction of reading as towards the broad end as follows [see fig. for key to numbering]. The beginning of the inscription has been lost and extant section begins in mid-letter: 1
Three short strokes perpendicularly across the stem - U, or any subsequent letter of the aaicme. Padel complains that Wainwright's drawing has these rather shorter than they really are, this is confirmed by the photograph, not that it affects the reading, of course.
2
A single oblique stroke to the left of the stem, sloping backwards - H.
3
Three oblique strokes to ihe left of the stem - T. These are roughly parallel with each other and with 2. The first is longer than the other two. There is a clear gap, more than enough for a stroke, between letters 2 and J, so there can be no doubt that they are to be read separately.
4
Three short strokes perpendicularly across the stem - U. These follow fairly hard on the end of 3. There may be a fourth stroke (E), but it is doubtful.
5
After a clear gap there is another vowel stroke perpendicularly across the stem - A. The next section is worn but at the end of it is another vowel stroke. If this is all one group it might have contained four or five vowel-strokes, judging by the space occupied by 4 - E or I. On the photograph there appears to be two strokes between the clearer ones, of a similar length but sloping forward across the stem. This seems unlikely but the section is too damaged to be sure.
6
Two long oblique strokes sloping forwards across the stem - G. From the photograph it is clear that these do cross the stem, in any case, from their slope they must be members of the m-aicme. Padel comments that there is no trace of Stevenson's dotted score before this group, certainly none is visible on the photograph, it seems doubly unlikely since, as Padel points out, if produced to the full length of an m-stroke it would intersect with 5 (in any case the third member of the m-
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
348
aicme is extremely rare in epigraphic ogham). 7
There is ample room for the four vowel strokes read by Wainwright at this point (E), though the photograph is obscure.
8
The tips of the two h-aicme strokes read by Wainwright are clear (D), it appears as if this group is complete but shortly further on the entire surface of the stone has been lost, and the inscription could well have continued. There is a further 90mm (3 xh") before the broken end, and, of course, the slab originally continued further. By this point A (ii) is already higher than A (i), and may have gone on quite a bit more.
Which gives the following tentative reading:
B
-](u+)HTu(o/i)( /a)GED[-
(Wainwright 1, Stevenson a)
This is the boldest of the Inchyra oghams, the only one to be noticed initially. The letters are rather deeply carved with a pock and smooth technique comparable to that of the symbols at this, the narrow end. The inscription comprises approximately eighteen letters, 1-14 up the long side, 15-18 across the short end [see fig. for key to numbering]. The inscription ends there and does not continue down the other long side. All the extant carving is clear, unformnately several portions of the h-surface have flaked off and there is particular doubt over 13 and 14. The large fragment which contained the h-strokes of 15-18 has been recovered and the letters completed. The first three letters, and especially the first, are less distinct than the rest. Stevenson commented on the 'irregular punching of their margins'. His suggestion that this represents an attempt to render less distinct the portion of B visible when the narrow end was sunk into the ground seem very plausible.
Letters 77, 72, and 15-18 have bind-strokes at their b-distal tips (the h-distal tips having been lost), but, though the b-distal tips of other letters are clearly visible, they are not bound. What distinguishes 77/72 and 16/17 is that they are successive letters of the same aicme and thus in particular need of unambiguous
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
349
grouping. The slope of letters 2, 4, 9, 16 and 77, indicate the direction of reading. The near unanimity of readings between Wainwright, Macalister (quoted by Wainwright), Stevenson (and by inference, Jackson), and Padel is an indication of the clarity of the carving. The lettering begins flush with the end of the stem-line and, occupying almost the full width of the narrow long edge, continues upwards as follows: 1
Five long straight vowel-strokes perpendicularly across the stem, quite widely spaced - 1 . The second stroke slopes a little backwards which means this could be two separate groups, 2 + 3 , OU, but all previous authorities have agreed on taking them as a single unit and the spacing is in favour of this interpretation. One would have expected successive members of the same aicme to have been more clearly differentiated.
2
Five short oblique strokes to the right of the stem, sloping forward - N. More closely spaced than the previous group.
3
Four long straight vowel-strokes perpendicularly across the stem - E.
4
One short oblique stroke to the left of the stem, sloping slightly forward - H.
5
One short stroke to the left of, and perpendicular to, the stem, - H. Presumably the differentiation of gradient is to clarify two successive letters of the same aicme.
6
Four long straight vowel-strokes perpendicularly across the stem - E. Most of the h-surface has been lost here, but there are traces of strokes on that side, especially the fourth.
7
Three short strokes to the left of the stem, sloping backwards - T. The first stroke sloped greatest and the third is almost vertical.
8
Four long straight vowel-strokes perpendicularly across the stem - E.
9
Four oblique strokes to the right of the stem, sloping backwards - S. Quite closely spaced.
From here to the end of the slab, very little of the h-surface survives. The few traces immediately above the stem must be interpreted with the help of the spacing of the extant b-strokes. 10
There are four proximal tips visible on the h-surface, the first being particularly clear, and the bsurface is intact and empty. These must therefore be h-aicme rather than vowels - C. Wainwright
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
350
thought that, while there was space for four strokes, 'balance and spacing* was in favour of three (T). I agree with Padel, however, that C is more likely. 11
Five long straight vowel-strokes across die stem - L Only die proximal tips of the h-strokes survive. There is a very faintly pocked bind-stroke connecting the distal tips of the b-strokes.
12
As above, but only four strokes - E. The final stroke is die most fully preserved on the h-surface. Again, there is a faint pocked bind-stroke.
13
The h-surface is lost almost to the stem and only the distal tips of the strokes survive. Two are definite, there is probably a third, and possibly a fourth. Since the b-surface is intact and empty, these must be h-strokes - D, T, or C.
14
Between here and the end there is room for a few h-strokes, perhaps three, but nothing has survived the damage to die corner of the slab. It is likely that the carving went right to the very edge, since it does so on the narrow end. Wainwright took 13 and 14 as TD, DT, or Q. All are possible, though Q outside the formula word MAQQ would be exceptional.
The final four letters take up die whole of the narrow end face: 15
Five long vowel-strokes perpendicularly across the stem, bound - 1 . Padel was concerned that the fragment which completes this end did not have the h-part of the first stroke. To explain this he thought it must have been Very curved out of keeping with the other scores' [106]. He said that the b-side bind-stroke didn't meet the first stroke, and thus wondered if the latter might have been part of anodier letter. I feel, however, that all of this may be explained by the damage die corner has suffered. There is a clear gap between die end of the slab and the first stroke and the spacing is such that the group cannot be other than a single group of five cross-strokes.
16
Five oblique strokes to the right of the stem, sloping forward, bound - N. Each stroke is a little longer than its predecessor giving the group a slightly wedge-shaped appearance. The first stroke, in particular, is rather short, presumably to prevent it intersecting with the last stroke of 15.
17
As above - N.
18
Four long straight vowel-strokes perpendicularly across the stem, bound - E.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
351
Which gives the tentative reading: INEHHETESCIE(t/c/d)(t/.)INNE
C
(Stevenson (d), Wainwright 3)
The strokes of C are not as distinct as those of B and the inscription went unnoticed at first. Gordon thought them 'trial grooves' [1956:42], but Richardson recognized them as ogham letters cut using the same technique as the adjacent symbols. The letters are quite deep, a smooth chisel-cut groove, not pocked like the letters of B. Though markedly less substantial than B, the letters are, in feet, of a similar size, though the script used is different. The stem-line is not drawn-in but must be imagined, and the vowels are very short strokes. The slope of 1 and 3 indicates the direction of reading is from bottom up (towards the broad end), with head turned to the left, as one would expect [see fig. for key to numbering]: 1
Four long oblique strokes sloping diagonally forward across the hypothetical stem - S. Padel and Wainwright were in no doubt that the reading was S, but there is the possibility of the end of a fifth stroke before this (N).
2
Four very short vowel notches - E. Three are clearly visible, the fourth less so.
Above this
letter the surface of the stone has flaked away, but there can be no doubt that these are complete vowel-strokes. 3
Three longish oblique strokes to the left of the stem, sloping backwards, each a little longer than the previous one - T. Thefirsttwo are clearly visible, the third stroke is almost lost in a flake. Wainwright thought there might have been a fourth stroke now flaked away, but I agree with Padel [106] that this is unlikely.
4
Three short vowel notches - U. As Stevenson points out, flaking has removed the original surface of the stone and 'perhaps one or two scores' [39], as a result these three strokes are faint and worn, and portions of them are missing.
A line, which Wainwright claims post-dates the
discovery of the stone [6] has almost obliterated the third stroke. The inscription could have continued, but there is no trace of any farther strokes.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
352
Which gives the reading: SETU(+) or NETU(+)
FORM OF SCRIPT The three Inchyra inscriptions, A (Mi), B, C, look radically different but, the variation in scripts is not as great as might at first appear. C lacks a stem, and B has some letters bound, but setting these aside for a moment, all are written in fairly simple forms of the script with similar over-all proponions in letter size and spacing. All have b- and h-aicme letters which slope and, with the exception of character A (i) 77, there is a noticeable lack of supplementary letters and other developed forms. Nonetheless, despite these similarities, there are three distinct 'hands' visible - A (Mi), B, and C - and the differences between them or not merely those of carving technique.
A is written in an infonnal and cursive hand which, though not exactly careless, has given little attention to the finer points of length, slope, and spacing. Letters are, however, clearly distinguished, especially successive letters of the same aicme, and legibility is not compromised. The vowels are rather shon strokes across the stem and occupy a minimum of the middle third of the ogham band. The closest palaeographical comparisons are with Ackergill and Buckquoy (excepting the forfeda), but Inchyra A seems somewhat exceptional in having sloping b- and h-consonants and well-spaced letters (both features more common in later inscriptions) but having shoner vowel-strokes and lacking, with one exception, supplementary letters and other developed features, which are generally considered late.
Conversely, with B, the bind-strokes, which elsewhere are correlated with having numerous forfeda and other late features, seem a little incongruous on an otherwise very simple Type Ila ogham. It could be merely coincidental that the text was such that no supplementary letters were required, and certainly, vowel-strokes which, as here, occupy the entire ogham band, are generally later than shon-stroked ones. It is also curious that there are no m-aicme letters, but this could be similarly coincidental, the text is not long. Note the marked and consistent slope of the b- and h-aicme consonants (the only exception is 7
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
353
whose slope diminishes with each stroke). Component strokes are generally parallel but there is some variation in how widely spaced they are, giving the letters an uneven quality. Letters, especially successive letters of different aicmi, are rather closely spaced, though they remain clearly distinguishable. Successive letters of the same aicme are more clearly spaced but by less than a stroke's wonh. In contrast to other inscriptions employing bind-strokes, which have all appropriate letters bound, bind strokes are not used for every letter at Inchyra. They appear to have been reserved for use only if there was room for doubt, though damage to the distal tips of several letters prevents a categorical assertion of this. B is a wellcarved inscription on a monumental scale compatible with that of the neighbouring symbols. It is most closely comparable the ogham from St Ninian's Isle, and to a lesser extent Scoonie and Brandsbutt (though these do not share the marked slope of b- and h-aicme letters). With one exception, all the other instances of bound letters occur in inscriptions from the Northern Isles, written in scripts markedly more complex than Inchyra's, containing rmmerousforfeda and even, in some cases, word division. The exception is, however, an important one, Abernethy. Though it has only three letters surviving, these are closely comparable with those of Inchyra B - sloping h-strokes and long, straight, vowel-strokes occupying the full ogham band. That the only two Mainland examples of bind-strokes should be within three miles of one another is striking.
The lack of a stem-line at C seems incongruous with its position across the face of the slab, but is scarcely as revolutionary as it seems. There are numerous examples of Irish oghams carved on stones so round and smooth that there is no arris, or an arris so flat as that the inscription is carved effectively across a face. Macalister lists a number of pulvinar or 'cushion-shaped' boulders from Corca Dhuibhne with such inscriptions, including Lugoagappul (CIIC 190-1), Burnham (175), Ballintaggart (155-63), and Ballineesteenig (147), but there are others elsewhere in the country, for instance, Tuckamine, Co. Carlow (CIIC 17). The incongruity of Inchyra C is merely that a stem-less ogham was carved on a slab with sharp and angular arrises. This choice may support the theory that C was the third ogham to be carved, since both narrow edges of the broad end would by then be already occupied by A (i-ii). But why leave out the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
354
stem-line ? Perhaps the carver was more familiar with the older conventions, certainly the very short strokes for die vowels are more in keeping with the old arris-cut oghams. But the marked slope of the band h-aicme letters, and the ample spacing of the letters have more in common with the later, manuscriptinfluenced, stem-written oghams.
The relative proportions of vowel and consonant strokes is paralleled at Auquhollie and Bac Mhic Connain (both written without a stem-line on an arris), but both have upright b- and h-strokes, similarly with Gumess. Newton has more of a slope on its b and h-consonants, and, in being split between stem-less arris and stem across the face, underlines the feet that the presence or absence of a stem may in some cases be a somewhat arbitrary indicator of date (similarly with the late but stem-less Golspie). Pool combines short vowel-strokes and sloping b- and h-strokes, and, apart from its stem-line and the exceptionally long thin profile of its letters, is perhaps the most closely comparable to Inchyra C. Previous authorities have commented on the 'Irish-looking* vowels [Padel 1972:106; Stevenson 1959:39], though this is more correctly a feature of date rather than ethnicity. The implication, however, as Stevenson noted, is that C is the earliest of the Inchyra oghams and thus that 'the order of the symbols argued above is in contradiction to the typological order of the inscriptions that seem associated with them' [ibid.]. I think it would be safer to say that the typological development of ogham, and its chronological implications, has yet to be worked out. It appears that within an essentially unified tradition, there may have been several styles used in different areas at a given time.
INTERPRETATION OF TEXT With die exception of A (ii) which is defective at both ends and in the middle, all the Inchyra oghams appear intact and, apart from a few letters which are damaged and difficult to read, are more or less legible. Nor are there many problems with transliteration, only the triangular character, A (i) 11, is obscure. The difficulty comes in knowing how to segment these undivided texts which do not conform to the standardformulas('X MAQ Y etc.). Since we do not know the functions of the monument in its
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
355
various phases of use we have few clues as to the kind of messages to expect. We do not even know if we should be looking for text in Irish or Pictish.
Inscriptions B and C are complete and independent. Lines A (i) and (ii), however, may be pan of a single text. If they are, their order can only be guessed at, all the more so since A (ii) is missing both its beginning and end. That A (i) does not occupy the whole of the available space, and A (ii) continues higher up the slab may indicate that A (i) is the run-on, though, given ample space, word boundaries would probably have an over-riding effect on layout. Even if they are both part of the same text the order of the two sections may not have been important. The problem is largely academic, since so little of A (ii) survives, but it ought to be born in mind that A (i) need not be semantically complete or self-sufficient. The informal nature of the script and carving, may give a clue as to meaning in that the text need not express an 'official* or 'public' sentiment. Neither need it refer to the rest of the carving, it could be no more than a graffito on a standing monument.
A (i)
ETTLIETRENOIDDORS
The reading of some of the vowels is doubtful, but the only transliteration query concerns 11 the unique triangular character. If it is a mix of straight and angled vowel-strokes then we may take it as some kind of O. Even if, on the other hand, it is related to the fh/br/fc/, it is still probably some kind of O. SimsWilliams has posited /6/ for the original value of this character, replaced at a very early date by a vocalic value, perhaps / 0 : / . The diphthong value oí/óe is, in his opinion, probably no earlier than the ninth century, the digraph value oi later still [1992:58-60; see entry on Formaston for discussion].
The first problem is where to segment the text. The sequence of letters is too long for a single personal name yet no word division is indicated. The only principle we have in deciding where to insert it is the tendency to avoid geminate symbols in word-initial position. The initial ETT is paralleled in the roman alphabet inscription from St. Vigeans [Okasha 1984; Clancy 1993] where it appears between two personal
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
356
names possibly as a variant spelling of Latin et 'and'. Macalister's ett between personal names on the roman alphabet inscription from Carew, Pembrokeshire (CIIC 1035) is a misreading [Nash-Williams 1950:184 No.303]. In ogham terms, double T may be explained as an attempt to indicate non-lenition, but it seems a rather outlandish explanation to posit a Latin word made to conform with native spelling conventions. This explanation would entail taking A (i) as following on from A (ii), being die end of a list of names et Lietren dddors.
lietren recalls the Lutrin in the Pictish king-list, which Jackson
interpreted as a Pictish reflex of Celtic *Lugutrinos. As Jackson shows, the expected Brittonic reflex of this would be *Loutrin [1955:164], which does not account for Inchyra's LIE-. According to diis interpretation OIDDORS would be some epithet of Lietren, though what this might be is unclear.
This line of enquiry may be entirely mistaken, however. Padel notes that ETT- is found as the opening sequence of Cunningsburgh 3 and Lunnasting. In die case of the latter it was very tentatively suggested that ETTE might be some form of the copula it-e, taken with the -S at the end of the section to mean, perhaps 'this is'. If the text is to be interpreted in Irish terms, then LIE may be Olr. lie 'stone' \DJL s.v. Ua], This spelling is attested for the nominative singular in the Old Irish glosses [WB 21c6, ML 131cl2, see Stokes & Strachan 1901-3]. While meaning 'stone' in a general sense, the word also has a range of more specific meanings including 'standing stone, pillar-stone, usually of a memorial stone, one marking place of death or burial, ogham-stone', later it comes to mean 'grave' or 'tomb', or a stone functioning as a landmark [DH] (this interpretation seems more likely than lie for lia 'flood', 'spate'). Irish trén is an adjective meaning 'strong', but is also used as a substantive meaning 'strong man'. It appears as the first element in compounds with nouns, adjectives, and verbs, meaning 'strongly'. Perhaps more relevant, is its use as an element in male personal names, e.g. Trénmór and Tréinfer [Uhlich 1993:302], cf. ogham TRENAGUSU (CIIC 428), TRENACCATLO (CIIC 353), TRENALUGGO (CIIC 26), MAQI-TRENI (CIIC 341). The element also appears uncompounded as an independent name, cf. ogham TRENU (CIIC 57). Inchyra's TRENO- could be the genitive of the name Trén, with LIETRENO for lie Tréno 'Trén's
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCH YRA
357
(ogham) stone'. This would leave IDDORS. The closing sequence -ORS is also found at the end of Cunningsburgh 2 but since the text is probably to be interpreted as the name Conmor + sy this may be no more than coincidence.
Alternative segmentations yield NOIDDOR, but the DD seems to preclude interpretation as Irish noithir, the present passive form oinóid 'makes known', 'spreads the feme', or some form of na(i)thir 'snake', either of which might be deemed appropriate comment on the stone, similarly, with OIDDOR for odhar present passive of oidid 'offers' 'grants' 'lends', used in the laws, in relation to the lending of property meaning 'it is granted' [DH]. To be accepted, any explanation must account for the whole of die text, so the above can remain no more than suggestions.
A (ii): -](u+)HTu(o/i)( /a)GED[This text is too short and fragmentary to be made much sense of. It may begin and end in mid-word, unknown amounts of text have been lost from both ends, and there may well be internal word division.
B:
INEHHETESCIE(t/c/d)(t/.)INNE
Unlike A (Mi), this inscription is a formal one, clearly related to a symbol statement.
We might,
therefore, expect a different kind of message. Padel comments that both sections of this text begin with IN(N)E. In Irish inne means 'wealth', or 'private possessions' [DLL], but there is no certainty that this is how the text is to be segmented.
More relevant may be the comparison between the opening
INEHHETE- and Gurness's INEITA- (unexplained). The text is complete and, apart from a few letters, easily legible. This makes our difficulties in interpreting it all the more frustrating.
C:
SETU or NETU
Like A, this is an informal text, but like B, it is clearly related to a symbol. The association with the fish may be the key to interpreting the text, since it seems to be paralleled on several other ogham stones. That
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
358
Ackergill (NEHT-), Golspie (NET-), and Latheron (NETU) all feature fish seems more than merely coincidental. Note also the beginning of the legible section of A (ii) which may be -EHT (for NEHT-?). Netu may be a name in its own right, or a hypocoristic form of Nechtan or similar [see entries on Ackergill and Golspie for further discussion]. The extant visible carving, however, reads SETU rather than NETU. This may be a hypocoristic form of a personal name like Irish Sétnae or similar. Whichever reading is correct, the text has the air of a simple personal name.
DISCUSSION Inchyra is one of a tight little group of oghams in south-eastern Perthshire. The link with Abernethy is particularly close, but Dupplin is only 8Vi miles away. The three, which are isolated from the next nearest oghams (Scoonie to the south and Auquhollie, a considerable distance to the north), may reflect Irish influence in this part of the Pictish kingdom of Fomenn. It is unfonunate that only Inchyra is legible and doubly so, since it provides so many problems of interpretation. The Inchyra texts are in line with other Class I symbol stone oghams in not conforming to the 'X MAQIY' formula. This cannot be coincidental, especially since most of the ogham-inscribed Class II cross-slabs do hav•- 'X MAQQ Y* (of the exceptions, Scoonie has a single name, and Brodie is largely illegible). Inchyra is a galling example of how difficult it is to interpret these texts as soon as they deviate from standard formulas. Text A (i) may provide a further example of the linguistically puzzling -S in what appears to be syntactically genitive position in a Scottish ogham. Inchyra is, in some ways, an anomalous monument and one with a complex history we don't fully understand. It is impossible to date closely, but the slab's regular shape, the use of a chisel for some of the carving, and certain features of the ogham script, are more plausible interpreted at a date later rather than earlier in the sequence of Class I monuments. Why a single monument should have attracted so many inscriptions is a puzzle, but this repeated use over time, supports the notion that ogham was well-integrated into the local culture, not a rare or exotic feature. The different hands and styles used further underline the heterogeneity of the ogham tradition in Scotland.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
INCHYRA
359
BIBLIOGRAPHY OS Record card - NO 12 SE 9 McLaren 1946; Stevenson 1959:33-9, pl.III-IV; Wainwright 1959 (reprinted as Wainwright 1961) (pi. 11 facing p.368); Gordon 1956:42-3; Henderson 1967:214-5, pl.24; Padel 1972:100-107; RCAHMS 1994a:92, 93, ill.; Ritchie & Fraser 1994:23.
CONCORDANCE Forsyth
A (i)
Padel
A
Stevenson
(b)
Wainwright
A(ii)
A
(c)
(2)
B
B
(a)
(1)
C
C
(d)
(3)
PREVIOUS READINGS Wainwright
1:
INEHHETES(t/c/dh)IE(td/dt/q)INNE
2:
(e/z)TTLIETRENo(i/q)DDOR(s/v) ~(u+)HTUoa(g/ng)ED
3: Stevenson
Padel
(2)
SETU (as above)
A:
ETTLIETRENOIDDORS -](u+)HTU(o/i)(a/ )GED[-
B:
INEHHETES(t/c/dh)IE(q/td/dt)[-]iNNE
C:
SETU
INCHYRA Symbol and ogham-incised slab [RCAHMS 1994a:95a]
'Rough'
AC)
INCHYRA - Symbol and ogham-incised slab showing position of inscriptions [Stevenson 1959 pl.Ill]
'^%m? INCHYRA - 'Unfinished* symbols and inscription A (i) [after Stevenson 1959 pi.IV] (key to numbering)
INCHYRA - Detail of damaged edge showing inscription A (ii) [photograph © Perth Art Gallery and Museum]
(key to numbering)
1
2
3 4 5
6
7
8
9
10
11
12
13 14
15
16
17
18
INCHYRA - Inscription B Top: Edge and top of slab showing position of lettering [Stevenson 1959 pL III]; Middle: Detail of inscription (key to numbering) [RCAHMS 1994a:931 Bottom: Detail of lettering on edge [Henderson 1967 pl.24J
5
4
.4* V. #
.
INCHYRA - Detail of broad end of slab showing symbols and inscription C (key to numbering) [Stevenson 1959 pi. Ill]
360
LATHERON
DISCOVERY This stone was found in 1903 built into the interior wall of an old byre at Latheron, Caithness [NGR ND 1981 3315]. The finder, John Nicolson of Nybster, who had previously discovered the Ackergill slab, brought it to Sir Francis Tress Barry who in turn presented it to the National Antiquities Museum (NMS Cat. No. IB 183). The previous whereabouts of the stone are unrecorded.
SITE A second fragment was found built into the same byre: a cross slab 2*3" by T P (c. 0.69 x 0.33m), on which the only carving to survive is the outline of parts of the side arms of the cross and the remains of the splayed shaft (1*7" (480mm) long, 6-8 V4" (150-220mm) wide). The angles of the cross arms are filled with T (50mm) diameter discs [RCAHMS 191 lb:82 No.298, OS record card - ND 13 SE 26]. A third fragment, part of a Class I symbol stone [OS record card - ND 13 SE 31] was found fonning the lintel of a false window in the south gable of the farmhouse of Latheron Mains [NGR ND 1990 3343]. It is incised with a 'crescent and v-rod' symbol decorated with what Stevenson has called the 'dome-and-wing' pattern [Stevenson 1959:40 pl.v.3]. Bothfragmentsremain in situ.
Apart from these three pieces of sculpture, which encompass Classes I, 11, and III, there is no other evidence for early medieval activity at Latheron. The two cross-slabs, however, might be thought indicative in themselves of some kind of ecclesiastical site. The only other pre-Norse remains in the area are afragmentof a penannular brooch from nearby Achavrole, and the possibly Pictish settlement at the Wag of Forse [Batey 1991:55,52]. It is significant that the name Latheron (or Latheron-wheel < Gaelic Latham a'phuill 'Latheron of the hole or poor, attested 1287 as Lagheryn, Laterne) is one of die few Celtic names to have survived in an area with an overwhelmingly Norse toponymy [Waugh 1993:120-1]. It is impossible to know whether the name was originally Brittonic or Goidelic since the root lath is
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
361
common to both, giving Welsh Head 'mire', Irish lathach 'puddle', Scottish Gaelic láthach [Watson 1926:122].
DESCRIPTION OF OBJECT A cross-slab of local stone carved on one face with, one above the other, the lower part of a cross in relief, an incised bird and fish, and two incised horsemen. To the left is an incised ogham inscription [see fig.]. Stone:
Caithness sandstone
Dimensions:
3'x 1 '5wx 4" (c.0.90 x 0.43 x 0.10 m)
Condition:
Left edge intact, all others fractured. Portions of unknown length lostfromtop and bottom. Carving worn, but generally clear.
The left edge of the slab is more or less intact, and is bounded tor the entire length of the carved face by a plain border. The right edge is more seriously damaged and lacks the raised margin, but, judging by the roughly central position of the carvings, it appears that only a few inches have been lost. More difficult to gauge is the size of the pieces lost from top and bottom. Several inches more would be necessary to complete the horsemen at the foot. If the rectangular design at the top is correctly interpreted as the lower part of a cross-slab than a fairly substantial piece would be needed to accommodate the missing arms and upper shaft. This would make the Latheron slab tall and thin, but not implausibly so.
In attempting a reconstruction of the original dimensions, comparison should be made with the two other Class II cross-slabs from Caithness. The Skinnet slab is a grander and more accomplished monument than Latheron, but the two have a number of features in common. Although carved on both sides, Skinnet, like Latheron, has a simple border round its margin, and incised symbols below a relief cross. On its other face, Skinnet has a relief cross above a horse, recalling Latheron's horsemen. The Skinnet slab is complete and has a height-to-width ratio of over 3:1. If we allowed an extra four inches of width for the Latheron slab then, if reconstructed in the same proportions as the Skinnet slab, it would be of the order of 5' 10" tall. If the 'missing' 2'10* were distributed 10"at the bottom and two feet at the top, there would be
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
362
ample space to complete the designs.
The tapered Ulbster stone is broader than Skinnet, more of the order of 2:1. This ratio would allow Latheron only a further 6". The Class II stone from Edderton in Easter Ross is a more sophisticated piece than Latheron, but has a number of features in common with Latheron and Skinnet. It has a simple band round its outer margin, a relief cross above symbols on one side and on the other, a relief cross above horsemen. Though not quite as slender as Skinnet, Edderton has a height-to-width ratio of 8:3, which if applied to Latheron would yield an extra 1'8\ So, if Latheron had been originally rather tall and slender, there would be no lack of precedent.
Macalister argued that, since the symbols were carved with the pock and smooth technique and the oghams were chisel cut, that the two were not contemporary. Since three different techniques were used to carve cross, symbols and horsemen, this is hardly a compelling argument. Also the lightly incised straight lines of the ogham would be most easily cut in sandstone using a blade, while the more substantial curves of the symbols might be considered more suited to the traditional punched technique. Since we would expect the ogham to be carved last in any case, that the stem appears to bend at two points to avoid the carving is of no significance. The only conclusive proof that all the carving on the slab was contemporary would be if the overall design could be seen to make allowance for the ogham inscription. While the space between the intact border and the symbols appears more generous than necessary if the carver had not had to take the ogham into consideration, it is impossible to get a clear sense of the relative position of the decorative motifs without the right margin. Nonetheless, there is no compelling reason to doubt that the ogham is intrinsic to the original design.
DESCRIPTION OF CARVING - DECORATION The carving occupies the whole of one face of the slab. At the top is a double rectangularfigurein relief which has been interpreted as the lower portion of a cross shaft. The upper rectangle is 'filled with double
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
363
spiral ornament arranged in C-shaped scrolls placed back to back' [RCAHMS 19llb:82]. The lower, narrower, rectangle is filled with an interlace pattern. If this is the remains of a cross shaft then it is of unprecedented shape. The shaft of the crosses at Ulbster, Skinnet, and Monymusk ABD, narrows, though in each case it expands again into a base [ECMS 33, 30, 192].
Below the rectangular figure there is incised a bird of prey, presumably an osprey, with its talons sunk into a fish. It is possible that these are to be interpreted as a unitary symbol rather than the typical pair, and thus in a different category to the goose and fish on the Class I slab from Easterton of Roseisle MOR [ECMS 124]. Charles Thomas believed the bird with the fish on the Class I Gairloch slab to be a goose also [1963:94], though others have interpreted it as an eagle [Stevenson 1952:110-1]. The eagle with its talons in a fish reappears on St. Vigeans 1 ANG [ECMS 238] and in two mid-ninth century Irish manuscripts, the Book of Armagh and the St. Gallen Priscian [cited by Allen ECMS 238]. Allen also mentions an Anglo-Saxon and a Continental example of the motif, which later appeared in the Bestiary tradition representing Christ and the world. Whether the Latheron creatures were intended to convey such a meaning is unclear. They may fit more easily with Pictish images of wild animals such as the Ardross wolf ROS, the Burghead bull MOR, or the Knocknagael boar INV, which surely are not open to Christian exegesis [for the Christian background to their an rather dian meaning see Hicks 1993].
The lowest register of the Latheron slab is occupied by two horsemen in profile, one behind the other, partly broken away. Pairs of horsemen are iiot uncommon in Class II and III hunting scenes, but a closer parallel for this design is found on the lowest panel of die Edderton ROS cross-slab. The Latheron 'cross' is carved in thick relief, the bird and fish deeply incised, while the horsemen are only lightly incised. This distinction in technique, in tandem with the vertical ranking of the three elements, may be indicate their relative importance within the over-all message. There are numerous parallels for a relief cross with incised symbols, for instance, Migvie ABD, Glamis ANG, and Golspie SUT, and the Caithness examples mentioned above, Ulbster and Skinnet. On an historical grounds, Curie dated die Latheron slab to the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
364
seventh century, but this seems rather early. The horsemen sit much more happily in the context of southern Pictish art of the eighth century or even ninth.
DESCRIPTION OF CARVING - OGHAM INSCRIPTION The ogham inscription is arranged in the usual position, vertically up die left-hand side. One might have e x i t e d the stem to have been placed up the middle of the raised border as it is at Brodie, but instead it runs parallel to it on the sunken background to the cross. Contrary to Macalister's sketch, the stem runs the entire length of the extant portion of slab. At the break, the stem has already begua There is ample space before the first stroke to make it clear that the break has not occurred in mid-letter, what is not clear is whether any letters have been lost from the beginning. Although the stem of several Scottish oghams begins at the first stroke (e.g. Brandsbutt, Ackergill, and Cunningsburgh 3), there are a few examples where the beginning is intact and the stem begins a little before the first stroke (e.g. Bressay). Just over 30mm of stem is visible before the first stroke, which is proportionally longer than most other lead-ins, but, on the other hand, it is also longer than the gap between any of the other letters in the Latheron text. The mark to the right of the stem immediately after the break might be a letter-stroke, the end of any letter of the b~aicmey but it is very feint. The carving is rather worn towards the top. The upper fracture occurs part way through a letter and, from the extant text, it seem likely that the inscription continued at least a little longer. The direction of reading is confirmed by the orientation of the hammer-head A (9) and by the slope of b- and h~aicme consonants (7, <5, 75). The angle of the consonants confirms that the inscription is to be read from the bottom up, as follows [see fig. for key to numbering]: 1
Two short strokes to the left of the stem, sloping backwards - D.
2
Three long strokes across the stem, sloping slightly forwards. If the strokes were vertical they would be read - U, but the forward slope would indicate the m-aicme and hence the very rare Gw. Since this, however, would produce a very awkward reading and since the slope is so slight (and appears greater because of the backward slope of the preceding letter), and certainly less than for 77, a vocalic interpretation is to be preferred. Padel thought to do so 'quite acceptable' [112].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
3
L ATHERON
365
Three short strokes to the right of the stem, sloping slightly forward, traces of a founh are just visible, and a fifth possible - N. Padel read only three, followed by a space - V, instead of the N of previous authorities. He interpreted 7-i as a separate word, though spacing does not appear to have been used to denote word division on other epigraphic oghams. Even though the stem appears intact at this point, there is no remnant of a fifth stroke, and the surface is 'not excessively worn1 [Padel], nonetheless it is more likely that the space was originally filled.
4
Five short strokes to the right of, and perpendicular to the stem, each a little longer than the previous - N. The stem appears to bend to accommodate the last two, which cross the stem slightly.
5
Two long strokes perpendicularly across the stem - 0 .
6
Two short strokes to the left of the stem, sloping backward - D.
7
Five short strokes to the right of the stem, sloping forward - N.
8
Five short strokes to the right of, the stem, sloping only very sightlyforward- N. Since no other reading is possible, the likely explanation for the difference in gradient, which is not great. Is that it was intended to make it easier to read two successive five stroke letters. In any case it may not be deliberate.
9
A 'hammer-head A* consisting of a single long stroke perpendicularly across the stem, bound by a single short stroke across its right tip J _ . As Padel identified, a mark to the left of the stem gives the false impression that the letter is shaped T , but it is simply a faint patch of wear [see discussion below for possible phonetic value of this character].
10
Three short strokes to the left of the stem. The first slopes backward markedly, the slope on the latter two is much less. This is probably to be taken as a slightly splayed group of three hstrokes, thus a somewhat careless T. The second stroke, however, is a little shorter than the other two and rather faint, and Padel preferred to interpret it as two single strokes separated by a slip HH. While the variation in stroke gradient - the use of both sloped and perpendicular h- and baicme consonants - is a notable feature of this inscription, within letters the component strokes are
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
366
roughly parallel. It is not clear if the variation is random, the carver's lack of precision, or, deliberate and significant. As explained above, in the case of 5, the difference appears intended to enhance legibility. It is possible that the carver began to carve this T sloped, like the previous h-aicme consonant (6), then changed his or her mind, and completed it perpendicularly, like the following h-aicme consonant (13). 11
A single long, oblique stroke across the stem sloping slightlyforward- M. At this point the stem seems to bend very slightly to accommodate the beak of the bird.
12
A single stroke across the stem, curled like an ' S \ or more accurately an integration sign - /. The lower hook is curved, but the upper seems straighter and more angular. Unfortunately this area has suffered damage.
Given the surrounding letters this character is probably to be
interpreted as A, but see discussion below. 13
Five short strokes to the right of, and perpendicular to, the stem - Q. Macalister thought they crossed the stem slightly.
14
As above - Q. There follows a very worn patch.
15
Five short strokes to the right of the stem, sloping slightly forward - N.
16
Four long strokes across the stem - E. Macalister thought the first two straight and the third angled forward at their mid-points ^ - > \ Padel preferred to read a single vowel-stroke - A and take the subsequent strokes as two sloping h-strokes, spaced to make two separate letters, possibly HH. His explanation is not entirely convincing since he admitted the possibility of a stroke opposite the second of these h-strokes was Very dubious'. This area is worn.
17
Three short strokes to the left of, and perpendicular to, the stem - T. The third of these strokes has the last few letters of the old museum label 'Caithness1 painted over it.
18
Three long strokes across the stem sloping slightly forward, before the fracture - U or any subsequent vowel (E, 1).
Which gives a reading of: DUnNODNNATMAQQNET(u+)-
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
367
FORM OF SCRIPT The drawn-in stem line and two supplementary letters qualify Latheron as a fairly typical Type lib ogham. Aesthetically, it is not a panicularly pleasing example, but, allowing for the wear on the stone, is a clear and legible text. Slight and 'free-hand' as it is, Latheron lacks the design element of more monumental oghams such as Brandsbutt, Brodie, or Scoonie, but it is not in any way deficient as a piece of writing. Nor should it be considered as a piece of graffiti, its caiving is at least as substantial as die two horsemen at the bottom of the slab.
The stem wavers a little to fit round protruding bits of the design. The letters are well-spaced and successive letters of the same aicme are clearly separated. Padel suggested the apparent gap between 3 and 4 was meant to indicate word-division, but it is more likely to have contained strokes now lost. A more definite gap occurs before 7, but the possible b-stroke at the very edge of the fractured lower edge is very faint indeed. Since the text makes sense as it stands we should perhaps follow previous commentators in disregarding this stroke and taking the gap to be no more than a slightly longer than usual 'lead-in' on the stem (cf. the lead-in on Birsay 1 which is considerably longer). There are no bind-strokes. All the letters slope as one would expect; except for 2, the forward slope of which is not great. Though strokes are roughly parallel within letters no great attention has been paid to consistency of gradient between letters of the same aicme. Component strokes are mostly of equal length and are generally evenly spaced, however, between letters there is considerable difference between the more generously spaced (7, 5,6) and the more cramped (7, 75).
Vowels consist of long strokes occupying the whole width of the ogham-band, diere are also two vocalic forfeda. The 'hammer-head' A, which here occupies only the middle half of the ogham band, may be presumed to provide an orthographic contrast to the standard A, and is perhaps to be transliterated ai [see Formaston for fuller discussion of this character]. The supposedly angled strokes of 16 are doubtful due
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
368
to wear but the character cannot be a fully-fledged angled vowel since the first two strokes are straight. Burrian, which is even more cursive in appearance than Latheron, has two instances of the final stroke of a group being angled, but neither is a vowel. Unfortunately 12 is also doubtful due to wear. Macalister depicted it as a smooth S-shape across the stem, Padel drew a distinction between the curved lower hook and the straight-sided, angled, upper hook. There are two other instances of what is presumably the same character at Lunnasting and Lochgoilhead, though in both cases these are ordinary straight vowel-strokes with long straight serifs at their distal tips. Context suggests the Lunnasting example is a vowel, perhaps some kind of A, but the Lochgoilhead example is immediately followed by an A so perhaps some other explanation need be sought. This straight, seriffed, character appears on the first two Birsay inscriptions but in both instances as the middle of a group of five strokes, either R or I. The seriffed letter may have been meant to indicate some nuance of pronunciation, or merely to bring together visually the five strokes of the group and make the letter easier to read. The only other example of a truly curved S-shaped stroke is provided by Bressay which has one group of five, representing a vowel, presumably I. These Bressay strokes are, however, backward-sloping mirror images of the S-shaped stroke here.
Palaeographically, the Latheron ogham is, in general terms, closest, to Buckquoy, which too has a hammerhead A, and Lochgoilhead, which includes a 'seriffed' letter. Birsay 1 is also similar, though it has angled vowels. All of these inscriptions are notable for their cursive, 'free-hand1 appearance. The Buckquoy ogham is probably eighth century in date; another case of the hammer-head A, Formaston is probably later in that century or into the next at the earliest. Padel was uncomfortable with Curie's an historical dating of the cross to the seventh century, preferring, for palaeographical reasons, to place Latheron in the first half of the eight century [112]. I would be prepared to go later - even into the ninth century.
INTERPRETATION OF TEXT If 1 represents the beginning of the text then the ogham would have begun a little above the feet of the horses at the bottom of the slab. It is difficult to judge, given die current state of wear, but it seems as
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
369
if the completed design may have come quite close to the lower end of the border, in which case there may not have been room to start the ogham level with the bottom of the design. Even if a portion has been lost, it could have been little more than a letter or two. But in that case, why leave a gap before 1 ? The possible stray b-stroke is very faint, and on balance it seems far more likely to take 1 as the beginning of the text. At the other end, the surface is fractured in mid-letter and it is clear that a portion has been lost. If the upper rectangular figure is reconstructed as a full cross, then a further distance would have been available to the oghamist, though perhaps not such a great deal since the cross-arms would almost certainly have projected to the outer border.
The transliteration of the hammer-head A is not in much doubt, but the precise phonetic significance of the serifs on 12 is not clear. Other than that the only real difficulty concerns 3. Still we can be reasonably confident of the reading Dunnodnnaitma(i?)qqnet(u+)-.
Straightaway it can be seen diat Latheron
conforms to the formula X MAQI Y familiar from Ireland and seen on half-a-dozen Scottish oghams. The mainland Scottish examples are consistent in spelling mace with QQ, an unhistorical spelling which must reflect the influence of the old ogham formula. In Scotland the letter Q appears only in this formula word. Though standard in the consonants used, there is some variation in the vowel in the various examples of this word. Formaston and Altyre have the expected A, the more northerly examples, St. Ninian's Isle, Bressay, Golspie, have MEQQ (cf. ?MEQ Blackwaterfoot 1). Latheron is unique in using kforfid in this position. It could be simply an elegant variant on the standard A, but the inscription is otherwise lacking in flourishes, and elsewhere the use offorfeda seems to reflect a desire for orthographic precision [SimsWilliams 1992 passim]. If the Northern MEQQ represents a dialect difference, then perhaps theforfid in this position on the northerly Latheron, represents half-way house between die traditional 'A' and the pronunciation spelling ' E \ If a genitive is meant, the contemporary manuscript spelling would be make with the / indicating the palatal quality of the ee. The hammer-head A (representing ai ?) might be thought more appropriate in this position, but the reconstruction of its phonetic value is not beyond doubt. Still,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
370
there can be no doubt that Latheron's 11-14 represents the Irish word 'son1.
Following it we have Netu* presumably the same element seen at Ackergill (NEHT-), Lunnasting (NEHHT-), Formaston (NEAHHT-). and Bressay (NAHHT-) [see Ackergill for a discussion of this name element].
The Lunnasting example, NEHHTONN, is the well-known Pictish male personal name
Nechtan/Neithon. Formaston's NEAHHT may be a shortened form, and Ackergill's NEHTETREB and Bressay's NAHHTVVDDAddS may be compounds using Necht- as afirstelement. Unfortunately we lack the conclusion of the Latheron text, but 18 already has too many strokes for the ending -ONN or -ANN to be reconstructed. If complete as it stands NETU may be a complete hypocoristic form of Nechton or a Necht- compound (cf. Inchyra's (s/n)ETU). One further stroke would give E, and suggest comparison with Ackergill's NEHTETREB. Latheron contrasts the other examples cited in not having an H(H) before the T. If this represents a real difference in pronunciation we may be dealing here not with a variation on the Necht theme, but instead with the Irish male personal name Neit, Néfi)d(e) [see Ackergill]. In discussing the Buckquoy ogham it was suggested that, in certain cases, Scottish oghamists followed the old convention of showing lenition with simplex consonants and non-lenition with double. Whereas T = /6/ would be a satisfactory explanation for 77, cf. Brittonic Neithon, it seems less likely for 10 where T appears to be a cipher for the manuscript t = /d/. Since t is a satisfactory explanation of 77, in line with the Gaelic Nechtan or Pictish Nation/Neiton, and we would expect an inscription to be internally consistent, it appears that Latheron is not following this old convention [Harvey 1987, where the author admits the convention was not universally applied even in classical ogham orthography]. Since Latheron is probably later than Buckquoy it is not surprising that its orthography should be more or less a caique on Olr. manuscript spelling.
Turning now to the beginning of the inscription, Anderson and Macalister read DUNNODNNAT, while Padel read DUV NODNNAT. Either form is acceptably Celtic. If the latter, then DU V may represent
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
371
a pronunciation spelling of Irish Dub /duv/. Dub 'black, dark' is a common element in (close-) compound names, either with its literal meaning, or more loosely as an intensifier [Uhlich 1993:231-4]. Dub was also in widespread use as the first element in loose-compound names in a similar way to Cú 'hound' and Mael 'devotee'. The second element was commonly a provincial or more local place-name (eg. Dub-Connacht)> or a personal name (eg. Dub-Chormacy or a common noun, often with toponymic significance (eg. Dubcluana), a parallel series wasformedon Dub-dá, 'Dub of the two . . / , (eg. Dub-dá-leithe, Dub-dá-thuatlu Dub-dá-Suibne\ cf. Cú-Ulad, Cú-Nuadau MáelSechnaill, Máel-topair [see CGH for further examples]. That such names were popular in Pictland is indicated by one Dub-Thalorg, king of Picts, and Dudabrach (for Dub-dá-brath ?), the father of the scribe of Meigle. If Dub is the preferred interpretation, then the second element NODNNAT, may, as Padel suggests, be compared with the Celtic divine figure Nodons, Nodontos, cf. Ir Nuadu W Nudd [O'Rahilly 1946:495-6], though as he admits, this is 'highly speculative'.
On balance, however, the preferred reading is not DUV NODNNAT, but DUNNODNNA(I)T, i.e. Dunodnait. Irish dun 'high place, fort' was a common element in compound male personal names, eg. Dúncath, Dúngus, Dúnlang [Uhlich 1993:235-6]. Latheron's Dunod- may be compared with Dimadach (Dúnad + adjectival ending -ach\ 'one who leads on campaigns' [Ó Corráin & Maguire 1990:80]. In this case, -NNA(I)T, could be the Irish diminutive suffix -natl-naiu commonly used to form a female version of a male personal name in -án, eg. OdránlOdornat, GobánlGobnat etc. By analogy at Latheron we may have Dúnad-nau That the name is followed by MAQQ implies that it is, however, masculine, but this need not be a problem in a Pictish context since the Pictish Gartnait is certainly male (one wonders if Irish redactors embarrassed by this apparently inappropriate name, might have emended the name of the father of the Pictish Cano, to Gartnán from Gartnait). As Thumeysen explains, though feminine in Old Irish, the suffix also appears with other genders, and was probably originally neuter becoming feminine 'as a result of its frequent use to denote females' [1946:174-5 §273.3]. Another possible example of -nait for a male Pict is the name of the father of the prehistoric king of Picts Drest, given in the P-Version of the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
372
King-List as Munait (though Anderson compares it with a Pictish Moneit AU 728) [Anderson 1973:248]. Also, it is always possible that the parent named was the mother, cf. Nechtán son of Derile. Another possibility for DUNNOD- is a form of Irish DUNAIDONAS (< *Duna + -aidonas 'fiery1) which appears in ogham at Rathglass, Barony of Rathvilly, Co. Carlow (CIIC 16) [McManus 1991:117 §6.26]. In unstressed position the diphthong was reduced to /3/, (cf. manuscript Lugedon, Lugadon, for Lugaidon, and ogham BIVODON (CIIC 285) for BIVAIDONAS (CIIC 504) (all genitive), manuscript Béoáed [McManus 1991:121 §6.28; Uhlich 1993:177], so Dunnodnnait for Dunnáednait is perhaps possible.
The above interpretation is on the basis of Latheron being a Gaelic text. If MAQQ is taken as a purely stereotyped form without bearing on the language of the rest of the text, then it might be possible to explain the remainder as Brittonic. The element duna- is not restricted to Irish, of course. The precise cognate of Irish Dúncath (
DISCUSSION Two features of the classical ogham pillar are reflected in the Latheron inscription, firstly the text formula 'X son of Y', and secondly the position of the ogham vertically up the left edge of the monument. That the flat surface is used rather than the arris is scarcely significant given the form of the monument on which the letters are carved. The Latheron oghamist has not utilized the flat raised border for die inscription, as was done at Brodie, Whiteness, Formaston (outer), and Golspie, but instead carved the letters up the background surface of the carved face, as at Fonnaston (inner). There is no need to assume that the Latheron text is secondary (as it probably is at Formaston). Instead it appears contemporary with the rest
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LATHERON
373
of the carving and thus, we may presume, refers to the patron or honorand of the cross, or less likely the craftsman. The secular equestrian motif may suggest that the person in question was not an ecclesiastic as such. AckergiU is about seventeen miles from Latheron and it is tempting to link the Net(e?)- who was the rather of Dunodnat with the Netoetreb commemorated on the earlier monument at AckergiU.
BIBLIOGRAPHY
OS Record card - ND 13 SE 27
Anderson 1904:534-8 ill.; RCAHMS 1911b:82-3 No.299; Macalister 1940:206-7 pl.IVa; Curie 1940:76, pLxxii; K. H. Jackson 1955:140-1; Padel 1972:112-5; Ritchie & Fraser 1994:14.
PREVIOUS READINGS
Anderson
-DUNNODNATM[-]QQNETO-
Macalister
DUNNODNATMAQQNETU(or)-
Padel
DUVNODNnA(t/hh)MaQQNAH(H/A)T(o>
LATHERON - Ogham inscribed cross-slab [CURLE 1940 pl.xxn]
s.
J » h\W'/'t\ /
//
LATHERON - Ogham-inscribed cross-slab
12 £-L
LATHERON - Ogham inscription (key to numbering)
374
LOCHGOILHEAD
CURRENT LOCATION The stone is currently housed in the parish church of Lochgoilhead and Kilmory, Cowall, Argyll [NGR NN 198 014]. It was discovered inside the church in the early 1990s by Ian Fisher of the RCAHMS during field work tor Volume VII of the Argyll Inventory [pers. cornm*]. Nothing is known of its provenance, neither when it was brought into the church building, which dates from the later Middle Ages* nor where it was previously. The curious patterns of wear on the stone, the scoring and marking, suggest an eventful life, possibly even time in the sea (Ian Scott, pers. coram. ]. Oliver Padel confirmed for the RCAHMS that the ogham was genuine [RCHAMS 1992:194 n. 18 <548)j but no reading was published in the Inventory, though the stone is illustrated. I examined the stone in November 1992 while it was in Edinburgh in the care of the RCAHMS. Shortly after which it was returned to Lochgoilhead church. [I am most grateful to Ian Scott, formerly of the RCAHMS, for facilitating my examination of this stone].
SITE This inscribed block is one of six carved stones inside Lochgoilhead church. Four are medieval and the fifth, a fragment of a millstone, with an incised ringed cross perhaps carved to mark the stone's re-use as a grave-marker, is probably Early Christian in date, though it could be mediaeval [RCAHMS 1992:195 (2)]. The church is first recorded at the end of the fourteenth century, when its dedication is mentioned in a papal mandate of 1392 [RCAHMS 1992:193 n.9 (p-548)]. Later sources specify the dedication as to the Three Holy Brethren. The Statistical Account of 1792 records the corresponding Gaelic name, Kilnam brathairn [OSA Vol. in. 160],
DESCRIPTION OF STONE Described by the RCAHMS as a 'fragment of an inscribed pillar1 this oddly shaped block is something of a puzzle.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHEAD
Stone:
Fine-grained sandstone
Dimensions:
420mm long by 160m wide and tapering from 115mm to 75mm in thickness
Condition:
Fragmentary, damaged, surface severely abraded in places
375
Probably only a small portion has been lost, but it is hard to image the original purpose of the object. Though some of the lettering is ambitious, sections are crude. The stone does not look like a finished product, instead various letters appear to have been carved by different hands over a period of time. Perhaps the block of stone was selected for carving then, for whatever reason, abandoned, ending up as chisel-fodder.
Following the RCAHMS labelling, the edges and faces are as follows [see figs.]: a (edge)
plain except for groove
b (face)
surface shelves away to left, two lines of roman lettering
c (edge)
to left of groove - knife cuts; to right of groove - one line of roman lettering, one line of ogham
d (face)
rough, uncarved surface.
Face (d) is very rough and appears never to have been carved. The remaining surfaces are comparatively smooth. The two edges bear a deep V-section transverse groove, the one across (c) begins 170mm from the left end, that on (a) slightly further to the right. The significance of these lines, which correspond to a marked change of level on face (b), is uncertain. Edge (a) bears no further carving. Face (b) has two lines of irregular incised roman letters beginning to the left of the grooves on (a) and (c). The left end of (b) is damaged, part of its surface apparently cut away. The upper left-hand portion has a number of sharp scores like knife cuts. To the right of the groove on edge (c) are two lines of lettering, the upper in the roman and the lower in the ogham alphabet. The lower left-hand portion of this edge is blank, the upper may contain traces of some roman letters.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHEAD
376
DESCRIPTION OF CARVING Roman Alphabet Inscription Face (b) Top:
After the scores on the left there is a lower-case B with seriffed ascender, it is followed
by a clear, curled lower-case D with horizontal ascender. There is a mark between the two letters which may be the remains of a small C but unless rather small, this would make the sequence cramped. A roundbacked E with protruding horizontal follows, with what may have been an F next. The top line may have concluded with a G but there is no trace of it now. Bottom:
The letters of the lower line begin further to the left of the B, strengthening the
hypothesis that the scores have obliterated an original A. The first two or three letters of die lower line (H I K ?) are much worn and difficult to make out. The extravagantly seriffed L is, however, clear. The RCAHMS suggest that the letter preceding it 'may have been a K defaced by later incisions'. The lower portion of the angular M is defaced but the letter can nonetheless be discerned. The line concludes with what appears to be lower-case N.
Edge (c) Left:
The RCAHMS suggest 'traces of two or three possible capital letters' in the damaged area
to the left of the groove.
Right:
To the right of the groove, above the ogham inscriptions are the small minuscule letters
D and O, followed by 'larger but less regular markings* which may be letters or letter-like doodles. Though there is great variation in size, on average the letters on face (b) are larger than on. edge (c), 4045mm in height, as opposed to 35-55mm. There is variation also in the quality of the lettering, not merely as a result of differential wear patterns. Some of the letters are rather fine, the horizontal-backed D and the round-backed E, for instance, are nicely proportioned. The ambitious use of serifs distinguishes others,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHE AD
377
for example the split haft of the third letter on edge (c) and the triangular, wedge-shaped serifs on the upright of the B and on the L, where they are particularly pronounced.
The extant letters on face (b) are in alphabetical order and the gaps just about sufficient io accommodate the letters missing from the sequence as far as N. The lettering on edge (c) is less easily understood. As suggested by the RCAHMS, the opening sequence DO may represent the Irish do 'for' and indicate the Irish memorial formula oroit do ... (or do ...) 'a prayer for ... \ but there is no trace of a preceding or. If the letters are to be interpreted as miscellaneous essays then do, a frequently recurring combination, would be a logical pair to practice in isolation. Conceivably the carver may just have been practicing curved letters. The nomen sacrum 'dominus' is normally represented by the contraction dns.
DESCRIPTION OF OGHAM INSCRIPTION The ogham stem-line is 195mm long and runs roughly parallel to the edge of the stone. The individual strokes of the letters vary between 20 and 35mm in length, and 3 and 6mm in breadth. The stem appears to have been cut before and again after the cutting of the strokes, which perhaps explains why some of the strokes appear to not quite reach the stem. The ogham carving is course and scrappy, doubtless because of the intractability of the stone and the subsequent abuse the block has suffered. The beginning and end of the text are particularly abraded, rendering any reading of these portions tentative.
The sequence of letters begins a few millimetres in from the end of the stem, as follows [see fig. for key to numbering]: 1
One short stroke above the line sloping diagonally backwards. If the stroke were vertical it could be read as H, however its slope is rather pronounced. If it were a feather-mark indicating the direction of reading, a corresponding stroke below the stem would be expected. The first stroke below the stem does slant in the relevant direction but meets the stem well beyond the tip of stroke 7. It is tempting to interpret this first stroke as a bounding mark indicating the beginning
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHEAD
378
of the inscription, though any other examples of such a stroke are doubtful. 2
One long stroke through the stem sloping diagonally forwards - M. The apparent dislocation of the two halves of this stroke may be illusory, the slight gap between the stem and the left of the upper portion being the result of wear.
3
Three long strokes, perpendicular through the stem - U. The spacing is irregular but a division into 1 + 2 is unwarranted for such a cramped inscription.
4
Two short strokes above the line - D.
5
One stroke, perpendicular through the stem, with what appears to be a diagonal flap attached to each end. The phonetic value of this letter, shaped like an angular, open S, is uncertain, see discussion below.
6
One stroke, perpendicular through the stem, a similar length to the previous stroke minus the flaps -A.
7
Two short strokes below the line - L.
8
Five short strokes perpendicular across the stem, varying slightly in height and spacing but not sufficiently to merit sub-division - 1 .
9
Two very short parallel strokes, sloping diagonally backwards, stopping well short of the stem. The second stroke, if projected would extend beyond the end of the stem. If these two were vertical and reached the stem they would be D. As they are, they can scarcely be interpreted as letters. If not letters, they may perhaps be some form of punctuation or directional indicators, perhaps to be taken together with 1 as bracketing the text in some way. Alternatively, they may relate to the line of roman letters above and be unconnected with the ogham altogether.
10
The stem is terminated by a long perpendicular stroke, thicker than the other strokes. If it were a letter it would be A, but its more substantial aspect suggests it should be interpreted as a termination mark.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHE AD
379
Taking the ogham left-to-right, as outlined above (which corresponds to the direction of the roman lettering), the text reads: ( /h)MUD(?)ALI( /d) The two texts, however, are not necessarily connected and the ogham could as easily be turned over and read in the opposite direction: (/1)ID(?)LUM( Ib)
FORM OF SCRIPT Lochgoilhead is written in a rather scruffy Type Ha or lib form of the script with perpendicular b- and hstrokes and long vowel-stokre occupying between half the ogham band and all of it. The letters are not widely spaced and there is considerable variation in strokes of the same aicme. The status of the marks at the beginning and end of the inscription remains doubtful because of the wear on the stone. The only distinctive letter is 5. Similar flapped cross-strokes appears on several other Scottish oghams. In the Birsay 1 and Birsay 2 inscriptions the character occurs as the middle stroke in a group of five cross-strokes, which are taken together to indicate I. A single flapped cross-stroke occurs on the Latheron stone* where it is interpreted as the A in MAQQ, and on the Lunnasting stone in the cluster TT-NNN, where it is usually read as A or M. A similar character appears on the Altyre slab, but its flaps are much less pronounced, the overall impression is more S-shaped, and the letter may be a simple cross-stroke with terminals clubbed due to weathering. Bressay has a letter consisting of five reversed 'S's, normally interpreted as I, but these are curly in contrast to the angularity of the others and point in the opposite direction. In none of the above examples has the modification been satisfactorily explained, the normal rationalization being 'stylistic variation*. The flapped device does not occur in epigraphic or manuscript ogham elsewhere [See entries for individual stones for further discussion].
INTERPRETATION OF TEXT At first sight, neither proposed reading seems particularly meaningful. If strokes 7, 9 and 10 are not to be taken as proper letters, then the text consists of seven characters. It is unclear exactly how many roman letters there are in the line above, the four distinct ones arefollowedby what might represent one, two or three letters. Any possible correspondence in the number of letters, however, is likely to be coincidental
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHEAD
380
and of no use in deciphering the, probably unrelated, texts.
The principle difficulty, other than the status of strokes 7, 9 and 70, is the phonetic value to be ascribed to letter 5. To interpret the Lochgoilhead example as 'A' presents the problem of two As appearing side by side. Though double consonants are a prevalent feature of Scottish oghams, vowels are rarely doubled. In early Irish manuscript orthography, however, vowels were sometimes doubled to denote length [Thurneysen 1946:20-1 §27]. The reading here may be AA (= á), the flaps having been added to prevent the two being read together as 0 . Another possibility is that the flaps in some way modify the vowel, the two together somehow representing a diphthong. A third option is that M is in fact intended, the flaps emphasising the difference between a long oblique and a shon perpendicular cross-stroke, but 5 is neither as long nor as oblique as 2 which is clearly an M.
I f letter 5 is to be read as M, then the text could be interpreted as a dithematic personal name MUD-MALI. The first element may be compared with *mdd > Olr mitad. The adjective miiad means something like 'noble' or 'good' [DH] and appears as an epithet or independent name, MUAD, on the ogham-inscribed brooch from Ballyspellan, Co. Kilkenny (CIIC 27) [McManus 1991:132 §7.6, n.7onp.l81]. The element more commonly appears with the diminutive suffix -AGNI > -AN(N), i.e. Múadán, or in die feminine form Múadnat, names borne by a number of early Irish saints [Ó Corráin & Maguire 1991.139-140], A very early form, MODDAGN(I), appears on an ogham pillar from Windgap, Barony of Middle Third, Co. Waterford [CIIC 307], which McManus dates to the early fifth century [ 1991:94 § 5.24; 107 §6.12]. The element also occurs in the name of the river Moy in Connaught [ace. Mbdam in VC 1.6].
The putative second element, -MALI, is comparable with the very common personal name element MAGL- ( > Olr. mál 'prince'). Mál occurs uncombined as a personal name [DZL], but is normally found in compounds, in either first or second position [Uhlich 1993:275-6.]. The element is even more common
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHEAD
381
in Welsh than in Irish names, taking the form mael. In Irish it interchanges with máel 'bald, cropped* > 'devotee' [O'Brien 1973:229; O'Rahilly 1946:360 n.2]. That the element was also used by the Picts is shown by the name of the name of the father of hte Pictish king encountered by Columba, Meilochon. There is only one extant ogham example of the element, CUNAMAGLI ( > Olr. Conmál, Conmáet), Arbory, Isle of Man (CIIC 501) [McManus 1991 §§ 6.3-4, 6.10, 6.15, 6.21]. The remaimng examples are found on Roman letter inscriptionsfromWales: BROHOMAGLI ( > W. Brochmaet)fromLlandyssul, Cardigan [CIIC 349] andPentrefoelas, Denbigh (CIIC 401); CATOMAGfLI] ( > Olr. CathmáU Cathmáet) from Brawdy, Montgomery (CIIC 425); SENOMAGLIfromLlanfihangel Ioreth, Carmarthen (CIIC 370) [McManus § 5.23]; SENEMAGLI and VINNEMAGLI from Gwytherin, Denbigh (CIIC 400) [McManus 1991 § 6.26; 103 §6.4 n. 10]. Although there is no extant example of the two elements MOD/MUAD and MAGL-/MAL together, the combination is theoretically possible, perhaps with meaning 'noble prince'.
The problem with the above equation is that there is no G in the Lochgoilhead text. There may, however, be a precedent for such a spelling in the VEDOMALI of the Roman letter inscription from Margam, Glamorgan (CIIC 408) [McManus § 6.8]. The final -i suggests a genitive, but the Lochgoilhead inscription is surely far later than the accepted date of loss of final syllables (end of the sixth century), the roman alphabet inscription certainly cannot be this early. The final -I may represent the anachronistic persistence of the old epigraphic formula. A pseudo-Latin genitival ending is scarcely plausible in an ogham text, but it may be relevant to compare Artmali for Arthmail on the Latin inscription from Llantwit Major, Glamorgan, which dates from the second half of the eighth century (CIIC 1012).
Alternatively, if the fifth letter of the Lochgoilhead ogham is not to be read as M, another possibility is the element -V ALI, from the root *wal 'to be strong' (cf. Gaulish val(l), Olr. flaith 'lordship'), which is common in personal names in all branches of Celtic - Ir. CathaU Old Welsh Catgual, Gaulish Catuwalos; IT. DomnalU Old Welsh Dumngual etc. [McManus 104 §6.7, 178 n.22; Evans 1967.269-271; Pokorny
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHEAD
382
1959.1111-2]. There are several ogham examples of the element in first position - VALAMNI Rooves More, East Muskerry, Co. Cork (CIIC 125); VALUVI Ballyvellon, Middle Third, Co. Waterford (CIIC 302) - and in second position - TOTAVALI ( > Túathal, cf. Welsh Tudwat) from Llaosadyrnin, Carmarthen (CIIC 375) [McManus §§ 6.7, 6.28]; SUVALLOS, Ballintaggart, Corkaguiney, Co. Kerry (CIIC 158); 7SOVALINI, Drumlohan, Decies without Drum, Co. Waterford (CIIC 281); possibly CUNAVA[LI] Ballaqueeny, Isle of Man (CIIC 504) - and in a roman letter inscription, CVONOVALI (> Ir. Conall, W. Cynwal) on die stone known as Men Scryja, 'written stone1, in the parish of Madron, Cornwall (CIIC 468). Against this hypothesis is the fact that the V sound is adequately catered for within the standard ogham alphabet. The third letter of the b-aicme, transliterated V, represents /w/ > Olr. /f/, Welsh /gw/ [McManus §3.15, p.36]. Unless the oghamist was trying to represent some nuance of pronunciation, there seems no reason to invent or employ the flapped letter.
Returning to the hypothesis that letter five may represent a modified A, the reading Mudáli is possible. MU appears as an abbreviation for MUCOI (moccu) on an ogham pillar from Coumeenoole North, Corkaguiney, Co. Kerry (CIIC 178). Might MUDALI be MU(COI) DALI ? Ogham examples of the name Dála are - SIDANI MAQI DALO Rathduff Corkaguiney, Co. Kerry (CIIC 194); DALAGNI MAQI DALI Monataggart, East Muskerry, Co. Cork (CIIC 119) - and it appears as DALLUS DUMELUS on a Roman letter inscriptionfromLlanddewi Brefi, Cardigan (CIIC 35 l.II). The reading MUDALI, can be compared with the MEDALO which appears to occur on the ogham pillar CIIC 279, one of ten from a souterrain in Drumlohan, Decies without Drum, Co. Waterford, DENAVEQA MU]COI MEDALO [McManus § 6.17]. MEDALO represents the sept name Mocu-Dalon, descendent of Dála. The most famous scion of die Dál Mo Dala was St Cainnech of Aghaboe, who visited Columba on Iona [VC III. 17].
DISCUSSION The apparently bewildering number of possible interpretations of die Lochgoilhead ogham (and MOD-
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHEAD
383
MAGLI, MOD-VALI, MEDALI, and MU-DALI doesn't exhaust all the possibilities) should not be misconstrued. The confusion arises because of doubts over the marks at the two ends of the text, over the value of the fifth character, and the difficulty in knowing how much orthographic latitude to allow in determining a reading. The problems of exegesis should not, however, obscure the fact that the ogham is carved with some flair. The ogham letters are certainly no more problematic than the roman ones.
An understanding of how the three texts relate to one another would be of great assistance in interpreting the ogham text. The letters of the ogham inscription are more lightly and finely cut than the roman, but this may be a function of the more laboured technique required for the minuscule curves in comparison with the simple ogham strokes. The juxtaposition of ogham and roman letter inscriptions on the one stone is not unique. To the famous 'Colman bochf stone from Clonmacnoise (CIIC 749) can now be added the Dupplin Cross and, of course, the Newton Stone. Furthermore the sites of Dunadd and St. Ninian's Isle, Shetland have produced separate inscriptions in each script.
If the style of the roman letters could be closely dated it would provide, at the very least, a terminus post quern for the ogham. The pronounced horizontal back of the D on face (b) is distinctive, but the absence of a G, often chronologically diagnostic, makes it difficult to date the lettering more precisely than probably seventh or eighth centuries, or even later [see fig.].
As stated at the beginning, it is difficult to imagine the original purpose of the Lochgoilhead stone. The aforementioned explanation - a stone selected for fine work, abandoned, then found convenient for practicing lettering - is no more implausible than any other. The RCAHMS suggest either a 'didactic or liturgical purpose1 for the alphabetic inscription. Exactly what they mean by 'liturgical1 is not specified and if the carving was didactic, surely it must have been intended for teaching stone-cutters not scholars. If portable wax tablets or sheets of bark were available why wrestle with such a trying medium as this to impart the ABC, and why go only as far as N ? Such alphabets may have been intended to display
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOCHGOILHEAD
384
participation in the prestige technology of literacy. Afine,complete, latin alphabet appears in a secondary position on the possibly sixth century cross-marked pillar at Kilmalkedar, Co. Kerry (CIIC 913) [Cuppage 1986.311, fig. 184 and pl.30]. Perhaps more directly comparable are the incomplete alphabetic sequences which occurs on the stone from Llandecwyn, Merionethshire (CIIC 1029) and possibly on the Dupplin Cross.
Though there is no evidence beyond this stone and, possibly, the millstone, it seems reasonable to deduce some form of ecclesiastical activity at Lochgoilhead in the early medieval period. Though Lochgoilhead was very close indeed to the territory of the Strathclyde British, it was in the territory of the Cenél nGabráin of Dál Riada and ties are to be sought with the Gaelic church. There is nothing about the inscription which would link it directly to Iona. Most of the inscribed slabsfromIona have rather uprights Ds, the closest parallel to the Lochgoilhead horizontal D is No.47, dated by Jackson to late 8th or 9th century [RCAHMS 1982:187-6, ECMS 400]. Its D, however, is not as tightly curled, nor the 'ascender' as long or horizontal. More numerous, and far closer, parallels are to befoundat Clonmacnoise, e.g. the 'or do thuathal saer' slab (CIIC 1087).
BIBLIOGRAPHY OS record card - NN 10 SE RCAHMS 1992:194, No. 87 (1) figs, a-b
PREVIOUS READINGS None
LOCHGOILHEAD Faces * and c fphotograph RCAHMS Argyll. VII.194 B,CJ
f S ^ j V ^ i i JMiJi'i I «** *•»& ****** í^*»*^* 4 **** **«*
;.v •'•«•>' 'Ti •!">**•. *£•». S»**Vo I
»* *^*5* , s * * * i "
«»* ».5,3
::*/AO»-!
Faces a-c [drawing RCAHMS Argyll.VII. 194 A(I)]
%
*
• ^ #•:
*
YlW:
*2b
^
I"" LOCHGOILHEAD - Selection of roman letters (actual size)
LOCHGOILHEAD - Rubbing of edge c (reduced to 60% actual size)
"
^ S ^ ^ ^ ^
5
II
6
7
10
4444
w
*
LOCHGOILHEAD - Schematic representation of ogham inscription (key to numbering)
«lit •
t^ LOCHGOILHEAD - Ogham inscription (actual size)
Vi
385
LOGIE
ELPHINSTONE
DISCOVERY This Class I symbol stone, a scheduled monument, is one of a group of three in the policies of the now ruined Logie Elphinstone House, in die parish of Chapel of Garioch (formerly Logie-Durno), Aberdeenshire (NGR NJ 7034 2588)* According to Stuarttfierewere formerly four stones lying 'a short distance from one another' on die Moor of Carden to the west of dieir current position [quoting die New Statistical Account cf Aberdeenshire, 1843, p.638). About 1821 the moor was planted and the stones built into die enclosing wall of one of die plantations. They were later ereeted in the garden of Logie Elphinstone House. Romilly Allen comments pithily that *[t]he remarks that have been made already as to the unfitness of allowing ancient national monuments to serve the purpose of ornaments to a private gentleman's pleasure grounds apply with equal force in the case of the stones at Logie Elphinstone* [ECMS 177]. While it is of the greatest regret that the stones are no longer in their original location, it could be argued that their role as 'ornaments' protected them from wilful damage: at some point, says Stuart, the founh stone 'which was used by the tenant as a hearthstone in his kiln, was split by the heat and destroyed' [1856:4]. Another stone formerly at Logie House, diough originally from nearby Newton of Lewesk [J. Ritchie 1916], parish of Rayne (NGR NJ 693 279), is now lost [J. N. G. Ritchie & Fraser 1994:10], apparently as recently as post-1972 [Henderson 1972b: 198]. When Padel visited the stones in the early 1970s Logie Elphinstone House, the former seat of Baronet Elphinstone, was a hotel [1972:116]. A few years later a catastrophic fire reduced the building to a shell The stones, however, are still accessible a little back from the estate road to the west of the old house.
SITE The ogham inscribed pillar (Logie Elphinstone 2) is the tallest of the three remaining stones, but all are very similarly proportioned. It is possible that they may haveformedpan of a stone circle; it sounds likely from die description of dieir former position, but it would be difficult to verify since their precise original
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
386
location is unknown. The re-use of pre-historic standing stones for Pictish symbols is common in the North-East, and for ogham is known at nearby Newton, a little further afield at Brandsbutt, and further still, at Auquhollie [see discussion below]. The three Logie Elphinstone pillars are typical Class I symbol stones in being undressed boulders incised with pairs of symbols positioned one above the other. They are all carved from the same kind of blue granite. The two without ogham are as follows [see fig.]: 1
An undressed pillar of irregular five-sided shape (1.06 x 0.73 x 0.30m), incised with a 'crescent and V-rod' above a 'double disc'.
3
A roughly rectangular undressed pillar (1.06 x 0.70 x 0.30m) incised with a 'Pictish beast' above a 'crescent and V-rod'.
The three symbols carved in various combinations on the Logie stones are the three most commonly recurring of all the symbols. The diminutive size of the Logie stones is typical of a North-East sub-group of Class I monuments and is paralleled at, for example, nearby Daviot, Inverurie 4, Kinellar and Kintore.
The place-name Logie is frequent dbroughout Scotland. It comes from Gaelic Logaich (> Lagaigh), dative of logach 'place in the hollow' [Watson 1926:101, 147]. Elphinstone, die family name of the local laird, distinguishes this Logie from nearby Logie Durao, and Logie Coldstone (formerly Logie Ruthven) in Mar. Carden, also, is a common element, deriving from the Pictish word for 'wood', cf. Welsh cardden 'thicket, brake' [Watson 1926:352-3].
HISTORICAL CONTEXT The Logie Elphinstone stone is one of a tight group of three ogham-inscribed stones from the Valley of the Urie, die area known colloquially as 'the back o' Bennachie'. Its current location is no more than 2 V£ miles from Pitmachie (the original location of the Newton Stone) and approximately 4lA miles from Brandsbutt. This area, the lower Garioch, is particularly rich in archaeological remains of all periods, though especially of the Bronze Age and the early medieval period. It boasts the greatest concentration of Pictish symbols stones in Scotland, including some very fine examples, and, despite the fact that no
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
387
modem archaeological excavation of Pictísh period sites has been conducted in the area, an interesting handful of Pictísh artefacts [see catalogue of 'Foul Hordes' exhibition, Ralston & Inglis 1984].
The former imponance of this area results, not only ftom the agricultural wealth of the area (since at least the sixteenth century the Garioch is known as the 'meal-girnal of Aberdeenshire' [Simpson 1944:11; Alexander 1952:285]), but, specifically, ftom its position on the main route north ftom Aberdeen to Inverness, via Inverurie and Huntly (the modern A96) following first the Don and then the Urie. The area's strategic location is underlined by the fighting there of the famous Battle of Harlaw (1411), in Gaelic Bardic tradition Cath Garbhthec, 'the Battle of the Garioch' [Alexander 1952:285]. As Davidson points out, the routes of the old roads were determined by the fordable points of rivers [1878:5]. Logie Woods occupy the confluence of the Urie and the Gadie, opposite the heights of Bennachie.
Dominating the area ftom the south is Bennachie, 'the sphinx of the Garioch' [Simpson 1944:60], atop which is the great hill-fort of Mither Tap [Ralston & Inglis 1984:25-6]]. This, the third highest hill-fort in Scotland (518m OD) is partially vitrified and extremely well-preserved in sections [NJ 682 224]. The fort is undated 'but seems to have more in common with Dark Age than with earlier fortifications' [Shepherd & Ralston 23]. Logie is only about 2'/z miles to the north, Pitmachie about 3Vfc, and Brandsbutt about 5 miles due west. Bennachie has been put forward as a possible location of the battle of Mons GraupiuSyfoughtbetween Caledonians and Romans in A. D. 84 [St. Joseph 1978]. The discovery of the remains of the major Roman fort at Durno, between Logie and Inverurie, emphasises the strategic importance of the area at the beginning of the first millennium [Sawyer 1981:146-8]. The nearby vitrified fort of Dunnideer [NJ 612 281] has a long and complex history of occupation beginning in the early Iron Age, though stray finds ftom the site suggest it may have been used in the Dark Ages [Ralston & Inglis 1984:36].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE E L P H I N S T O N E
388
Aberdeenshire is remarkable for its abundance of Class I Pictish symbol stones. Numbering more than fifty, these account for almost a third of the total for the whole of Scotland. Within the county they are most densely concentrated in the Don-Urie valley (also the richest agricultural land) and, in Henderson's words, 'from Dyce the symbol stones follow one after the other every two or three miles' [1972a: 169]. There is a particular concentration in the area immediately west of Insch, including the remarkable Picardy stone [NJ 609 302]. The numerousness of Aberdeenshire symbol stones prompted Joseph Anderson to posit that the symbol system originated in Aberdeenshire [ECMS.Iicv], On the ground of purity of design and technique Henderson argued instead for the area round the Moray Firth [1958]. More recently, Murray has restated the case for the Aberdeenshire origin of at least some of the inventory of symbols [1986]. Either way, there is no denying that symbol stones are particularly numerous in this region, nor mat, even though some are rather devolved, a number are particularly fine (including Newton, Brandsbutt, and Dyce).
The history of the early church in the Garioch has yet to be written. A few snippets have been gathered by Simpson [1944:84-101], though his theories of early Ninianic activity in the north render his discussion here, and elsewhere, problematic. His suggestion of the extension of the cult of Drostan from its centre at Deer is, however, worth further consideration [1944:86-9]. A few miles to the south, in the valley of the Don, still in the shadow of Bennachie, is the early ecclesiastical foundation of Monymusk. The earliest historical reference is to a Céli Dé community there in 1078 (incidentally the only one between Dee and Spey) [Simpson 1944:87-8], but sculptural evidence points to activity there in the Pictish period. There is a particularly fine Class II cross-slab and four simple cross-marked stones [Simpson]. The Monymusk slab, like the one from Dyce, is untypical of Class II in being sculpted on one face only, and in its lack of 'iconographical padding' [Henderson 1972a: 172]. Both these, and the related stone from Migvie, are irregular slabs with neither margins nor frames [ibid.]. Romilly Allen placed these in a transitional Class intermediate between Classes I and II [ECMS.II:35], Stevenson and Henderson, however, have suggested that the very small number of Aberdeenshire Class II and III crosses are late, well into the ninth century
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE E L P H I N S T O N E
389
[1955:126, 1972a: 173]. The local association of the cumdach or house-shaped shrine known as the Monymusk reliquary (one of the treasures of Pictish metalwork) may not predate the fourteenth cencury [Simpson loc.cit.]. The finest of all the Class II cross-slabs in the North-east, is without a doubt, the 'Maiden Stone* from Chapel of Garioch, no more than a mile south of Logie Elphinstone, across the Urie [NJ 703 247], Described by Henderson as a 'tour de force\
it 'stands quite apart from the rest of the
Aberdeenshire stones' [1972a:173]. It is certainly late in the series, probably ninth century, and may well be the last [Henderson op.cit., Stevenson 1955:126]. Situated on the lower slopes of Bennachie, there is no evidence of any ecclesiastical sites in its immediate vicinity. The presecnce of simple cross-marked stones of early type suggests an early foundation at Inverurie [NMR OS record card].
The only other evidence for early foundations comes in the form of dedications to Celtic saints. Unfortunately for our purposes, the dedications of the ecclesiastical sites closest to the Newton, Logie, and Brandsbutt stones are late - to St. Laurence (Old Rayne), St. Mary (Logie-Durno and Chapel of Garioch) and St. Apollinarius (Inverurie) respectively, and record of any previous dedication to Celtic saints has not survived. The early ecclesiastical foundation at Culsalmond, north-west of Rayne, where there is a crossmarked stone [Ritchie 1916], was dedicated to the Brittonic St. Serf [Simpson 1944:152] (the only such dedication north of the Mounth), and further up the valley the church at Insch is dedicated to the Pictish St. Drostan [Watson 1926:316-8]. The old name for the parish, Inchemacbany suggests a dedication to Mo-Bheathan (Bean) [Watson 311-2]. Daviot was dedicated to Colm, a St. Colm's fair was held there [Redford 1988:158], but, as with so many of the others, the evidence for this association is not early, and any Irish associations discernible in the area need not date to the Pictish period.
The district of the Garioch shares a single river system with the Province of Mar, to the south, with which it has been linked 'by cultural and political contacts since prehistoric times' [Simpson 1944:5]. In 1014, Donald son of Eimhin, mormaer of Mar was killed fighting with the Irish at the battle of Clontarf (according to AU, and to a poem on Mael Shechlainn, King of Ireland [O'Keeffe 1934:32]), but Irish
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
390
connections earlier than this are hard to find. The only other sign of early Irish interest in the area comes from a list of tales in the orgun genre ('ravaging, raid, attack, massacre' [DH]) in the twelfth century Book of Leinster [Best, Bergin, & O'Brien 1.190; and in RIA MS 23 N.10]. These are Organ Bene Ce 'the orgun of the hill of Ce' and Orgain maighi Ce la Galo mac Fephail 'the orgun of the plain of Ce by Gall son of Fephal'. Cé is the Pictish province of the name which is traditionally identified as Mar and Buchan [Watson 1926:108]. In her note on the derivation of the name Cé Margaret Dobbs identifies Ben Cé with Bennachie [1949]. This, presumably, would make Mag Cé the Garioch (< Gaelic Gairbheach 'place of roughness', 'the formation is uncommon' [Watson 1926:118]). Dobbs also mentions a reference in the Metrical Dinshenchas (MD 96, 1.61ff) to Frigriu cerdfromCruith Maighe Cé [Dobbs 1949:138, see also Watson 1926:115]. Frigriu, who is also described ssfoglaid 'plunderer, reaver', is credited with building the hill-fort of Ailech, west of Lough Foyle, in the days of king Fubthair of hi (Iona). Like the nebulous literary evidence for a connection between south-west Munster and Circenn [see Auquhollie], these snippets are in need of further untangling.
DESCRIPTION OF OBJECT A pointed pillar incised on one face with a pair of Pictish symbols, traces of a third, and ogham letters cut on a circular stem [see fig.]. Stone:
Blue granite (waterworn before carving?)
Dimensions:
1.37 x 0.76 x 0.45 m
Condition:
Good. Carving intact and well-preserved.
In every way this is a typical Class I Pictish Symbol Stone and is to be compared with the other oghambearing Class I symbol stones, Brandsbutt, Ackergill, and Inchyra.
CARVING TECHNIQUE Not only is the ogham-inscribed pillar the tallest of the three surviving at Logie, but it is also the most accomplished.
The symbols on each are of average size, but because the stones themselves are
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE E L P H I N S T O N E
391
comparatively small, the designs occupy almost the whole surface. The symbols on Logie Elphinstone 2 occupy the lower two-thirds of the pillar, with the ogham sitting rather pleasingly in the middle of the top third. First glance suggests that the placement of the symbols must have taken the ogham into account, and that it is intrinsic to the original design, but in fact the symbol pair also occupies the lower two-thirds of Logie 1 where there is no ogham to be accommodated. Logie 2 narrows sharply above the crescent which could not have been fitted-in any higher. While it seems likely that the ogham does relate to the same phase as the symbols, it cannot be proved definitively either way.
Both are carved using the standard pock-and-smooth technique used on virtually all Pictish Class I carving, and have lines with a broad and open section. The lines of the symbols are broader, deeper, and smoother than those of the ogham, but this may reflect nothing more than the larger scale of the former. Logie Elphinstone was one of the symbol stones scrutinized by Gordon in his study of Pictish carving technique. He considered it close to the nearby Brandsbutt stone (which he judged particularly finely executed) and also to the even nearer stone from Mounie (a.k.a. Daviot) [Gordon 1956:44].
It is particularly interesting that Logie 2 is a palimpsest. The remains of a partially-erased double-disc and Z-rod are visible under the other symbols. This is brought out particularly clearly in Lines's Impression* [1989]. Whether the earlier symbol was more lightly carved to begin with, never fully completed, or has been deliberately smoothed over, the hardness of the stone doubtless accounts for why more of it was not effaced. Since the underlying symbol is still so visible, and potentially confusing, perhaps it might have been that the stones were painted in some way. This would enable the earlier caiving to be concealed, and if the later pair were picked out in colour the whole would be much clearer. There is no trace whatsoever of a second underlying symbol. I know of no instances of complete symbol stones bearing only one symbol since 'genuine' symbols always occur in pairs, with or without mirror and comb. This leads me to think that the underlying double-disc on Logie 2 reflects an unfinished monument reused. It seems unlikely that so much of the first symbol would have been carved before it was realized there was
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
392
insufficient room for the crescent above, also the second double-disc is of slightly different design to the first. Whatever the explanation, however, it is clear from the balance of the over-all design that the ogham relates to the second phase.
DESCRIPTION OF CARVING - SYMBOLS Isabel Henderson cites Logie 2, along with Newton, as an example of particularly good layout [1972a: 171], and it is true that the carving is very pleasing to the eye. All three of the surviving Logie stones are carved with crescent-and-V-rod symbols, in upper position in 1 and 2, lower in 3. This is the most common symbol on Class I stones, and the second most common on Class II. The combination 'Pictish beast' plus crescent-and-V-rod, as found on Logie 3 occurs also at nearby Crichie, and. inverted, on Kintore 1 ABD and Strathmartine FOR. Logie 1 's crescent-and-V-rod over double-disc is unique. Logie 2's crescent-and-V-rod over double-disc-and-Z rod is found with mirror and comb at nearby Bourtie, and on three Class II cross-slabs, Aberlemno 3 FOR, Cossins FOR, and Elgin MOR (inverted).
The great variety in the decorative detail of crescents and rods has allowed R. B. K. Stevenson and Gordon Murray to devise two, partly complimentary, systems of classification. The main divisions are between crescents decorated with 'dome-and-wing' design, those with spirals and scroll, and those with everted spirals and scrolls (Murray's Class E) [cf. Stevenson 1955:102-3]. All three Logie Elphinstone examples are Class E, a 'relatively infrequent' form, occurring twice on Class II and only six times on Class I [1986:233]. It is striking that three out of these six are from the same site. Amid the great variety of internal decoration on crescents, Logie 1 and 2 have an identical opposed spiral design (Logie 3 has spirals which turn in the other direction), also their rod-terminals are differentiated in exactly the same way. The comparison extends further, both stones swell and taper, though Logie 2 does so to a lesser degree. Both have a crescent-and-V-rod as the upper symbol, and some form of double-disc below (though Logie 1 lacks the typical Z-rod accompaniment). The designs are so close as to suggest that one is the copy of the other. Since Logie 2 is the better and more ambitiously rendered of the two, we might take it as primary.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
393
The double-disc is the second most common symbol on Class I monuments, the most common on Class II, and the most common over-all. The rod-less variant is markedly less common. It occurs on only a few Class I examples (Drumbuie 1 INV, Fy vie 2 and Newton ABD, Inchyra PER - the later two both having notches in one or both discs), though is more prevalent on Class II (Romilly Allen lists ten examples) and features in a number of the Fife caves. There are variations in the internal decoration of the discs and in the precise form of the 'bridge' between the two, but the decoration is too simple to be susceptible to the kind of analysis possible for the crescents [Murray 1986:236], According to Murray, it is relatively uncommon for discs to be completely undecorated, as with the later double-disc on Logie 2. The underlying double-disc is more typical, it has two concentric circles and a central dot, and [pace OS Record Card] it can be seen to have a complete and typical Z-rod The rod on the second double-disc is a mirror image of the first, and it is the earlier which is by far the more common arrangement. Murray argues that the double-disc may have originated and/or developed in Aberdeenshire [1986:243], but this is a contentious view.
DESCRIPTION OF OGHAM INSCRIPTION The carving of the ogham is very clear and shows up particularly well in the natural dye 'impression' (i.e. nibbing) done by Marianna Lines [1989]. The inscription consists of five groups of ogham strokes arranged round a circular stem. The horizontal diameter of the circle (140mm) is slightly wider than the vertical (135mm), but this is scarcely noticeable in the field. The shortest letter strokes (2) are 30mm long, the longest (5) are 70mm, and the remainder vary between 40 and 60mm [see fig. for key to numbering]. The scale of these letters relative to the stem makes the inscription hard to judge: the letters are cramped, and across any individual letter the stem is so curved as to make it difficult to be sure of the slope intended. Macalister attempted to show that the oghamist had been attempting a symmetrical layout, but this is based on an unsympathetic interpretation of the carving and a misleading diagram. His 'ideal* layout does violence to the original and cannot be accepted as it stands [Padel 1972:116]. Nonetheless it is true that the number of strokes per letter is 5, 3, 3, 5, 3.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE E L P H I N S T O N E
394
All previous authorities have agreed in reading clockwise (i.e. looking up and out from the centre). This agrees with the slope of 1 and 4 (pace Padel it cannot be read in either direction). [For further discussion of the reading of circular oghams see entry on Buckquoy]. There does not appear to be any indication of where to start reading. The largest gap is at 10 o'clock (between 5 and 1) but there is also a sizeable gap opposite it between 3 and 4. There is no gap at all between 1 and 2.
I believe that previous authorities have been over-subtle in interpreting the arrangement of strokes, for instance, by reading too much into the differences in gradient within groups. If viewed from a distance of a few feet, rather than from a few centimetres, which is surely what the carver intended of the reader, the overwhelming impression is of five letters. Starting at the top (after the ten o'clock gap), reading clockwise, the inscription is as follows [see fig. for key to numbering]: 1
Five strokes, of differing gradient but roughly equal length, outside the ring. The first two are roughly parallel with one another and slope markedly backwards. The next three, sloping backwards at a less acute angle, are roughly parallel to one another, but not to the preceding pair. Given the substantial gaps between letters elsewhere in the inscription it seems most likely, prima fade, that this is a single group, and the inconsistency in gradient the result of the difficulty of arranging five parallel strokes round a tight bend. Though well-spaced at their outer edge, strokes two and three collide at the stem, and knowing how much to read into this is the key difficulty in interpreting the carving. If deliberate, it might be an attempt to differentiate between two successive letters of the same aicme (in which case DT), but given the considerable space allowed between 3 and 4, this seems less than likely. The awkwardness of strokes 2 and 3 may, however, result simply from a misjudgement of the angle of stroke 2. Stroke 1 is perfectly well-aligned with the others. If 1 is to be read as a single group of five strokes to the left of the stem it is Q.
2
Three short strokes inside the ring. This group consists of the shortest strokes in the inscription, a minimum of 30mm in length. The first stroke joins the stem almost at the same point as the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
395
fifth stroke of group /. Most authorities, including Padel have taken this a single long stroke across die stem - A. Arguments against this interpretation are as follows: the three form a cohesive group of their own, evenly-spaced and parallel. The putative vowel stroke would be almost twice as long beyond die ring as within, and would be strangely angled, its outer portion at almost 45<> and die inner at 90o to the stem. Padel suggests that the lack of available space entailed die need to carve some of the letters over-lapping. This is certainly possible, but a more even distribution of die letters on die available circumference would have been perfectly adequate. If the inscription had been carved without much forethought starting at 2 or 3, then there may have been a need to squeeze / in, but this seems a little far-fetched. It is true that the layout does not suggest careful planning, but equally it is possible that the oghamist began at 1 by keeping the letters close, realized there was, in fact, plenty of room, and spaced the later ones more generously. There are two identical, parallel, very short and slight scores outside the stem beyond the first and third strokes. These are not particularly noticeable with die naked eye, but come out in Lines's impression as clearly artificial. Their length, weight, and gradient (sloping forward, i.e. against the expected direction of a letter-stroke), are all against them being letter strokes. I think, instead, they may be 'punctuation marks', added to differentiate group 2 from the strokes on either side, perhaps after the letters had been carved and die oghamist decided the first two groups were potentially confusing. This appears very necessary in the case of the first stroke, less so for the third, since there is a clear gap between it and the next group. Letter 2 is not as prominent as the other letters, but is clearly visible in Lines's impression and is not in any doubt. 3
Three strokes outside the ring, parallel to one another and roughly perpendicular to the stem - T. These strokes are not much longer than those of 2. They are well-spaced from both the preceding and succeeding groups.
4
Five strokes outside the ring, roughly parallel to one another and sloping slightly forward - Q. The first is the longest, the middle the shortest, but this variation results merely from the curve of the stem. On the inside of die ring opposite the second and third strokes of this letter is a thick
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
396
wavy score shorter than a letter. Most authorities have dismissed it as a mistake or flaw in the stone, and not taken it as a continuation of the second stroke (to make HAT), The group is so clearly separated from its neighbours that this is surely correct. 5
Three long strokes across the ring, each slightly longer than its predecessor, sloping very slightly backwards. None of the three precisely meet, die disparity is greatest with the first and least with the last. Since there are no other letters of either the m- or dHiicmi it is difficult to decide to which of the two 5 should be assigned. Though virtually perpendicular, the slight backwards slope may be there to emphasis that the strokes are not forward-sloping m-aicme. Since the third character of this aicme (G*)is extremely rare in monumental inscriptions the most straightforward explanation may be - U.
Which gives the reading:
QFTQU
FORM OF SCRIPT The inscription is so short that there is very little to go on in discussing the palaeography of die script used. If 5 is a vowel then it is rather long, but not exceptionally so. Strokes are generally evenly spaced and sloped within letters, though both these are open to distortion because of the curving of die stem. Generous gaps are allowed between letters, though again, die difficulty of arranging a text round a closed ring may have affected spacing. In general proportions the Logie ogham is not dissimilar to that from Newton, but the script of die two is not identical. No particular significance can be assigned the lack of forfeda in a text of this length.
The only remarkable feature of die script of Logie is, of course, die circular stem. This is the only monumental example of die device and there is only one other epigraphic example, Buckquoy. As explained in the discussion of this latter inscription, the two are not as closely comparable as they might appear. The ratio in scale between stem and letters at Buckquoy is such that it is really a straight inscription carved on a curved object, just as the 'boustrophedon' sections of Irish oghams which round
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
397
the corner at the top of pillars and continue across the top and down. The fact that, for purely functional reasons, the base-line happens to be curved, makes little appreciable difference to these texts (cf. also the curving stem of Cunningsburgh 3).
The circular layout at Logie Elphinstone, however, is not 'functional', in the sense that the ogham could have gone up the side of the stone or even across the face on a straight stem (though admittedly there would be barely enough room for this latter option). Instead the choice of ring must be deliberate, either for aesthetic reasons (the carving on the stone is very pleasingly laid out) or to signal the out-of-theordinary nature of the text. Previous writers on this stone have mentioned references in Irish saga literature to the carving of oghams on circular withes [for full discussion of the ogham in the sagas, see McManus 1991:156-63 §8.10]. In fact there are only two such references, both involving the great Ulster hero Cd Chulainn, most accounts in the sagas of oghams not connected with memorial inscriptions are to one or more straight lodsflesc(a), of wood (usually yew) [McManus 1991:160]. Thefirstexception is a reference in Cú Chulainn's Macgnímrada, 'boyhood deeds', to an ogham carved on an iron ring or 'withe', id iarnaidi, also described as id niachais,'img of heroic deeds' [O'Rahilly 1967 (Recension II of the Táin) lines 1068ff, O'Rahilly 1976 (Recension I of the Táin) lines 710ff, discussed by McManus 1991:156]. The ring is hung round a pillar-stone in front of an enemy fort and contains a general challenge to single combat. The second ogham, again containing a challenge to enemies, is carved by Cú Chulainn on the peg of a withe which he hangs over the top of a pillar stone [O'Rahilly 1976, lines 220ff; O'Rahilly 1967, lines 456ff, discussed by McManus 1991:157]. The peg, presumably, is straight, so this is not strictly speaking a circular ogham. As McManus points out 'as the peg (menoc) of the withe is the means by which it is locked or released it is appropriate that the inscription, which contains the key to unlocking the geis [taboo], should be written on it' [160]. It is possible that the Logie ogham gives epigraphic expression to the practice of hanging ogham-inscribed rings on pillar-stones - as has been pointed out, its tapering shape makes it ideal to receive such a hoop - but McManus is sceptical that any of the saga references to ogham record actual practice [163].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SC0IXAND
LOGIE E L P H I N S T O N E
398
INTERPRETATION OF TEXT Logie is the shortest complete monumental ogbam text in Scotland. Other short texts appear to consist of single personal names, as with Inchyra C, though this seems an unlikely explanation for QFTQU. Padel offered a number of alternative readings according to the significance accorded spacing and slope. The following read across, taking one option from each column, the topmost in each row being, in his opinion, the most probable:
DDAL T
Q
U
CAL
CH
BHO
DTV
CAHO
DTBL
His preferred reading was DDALTQU, or, since he correctly observed that Pictish oghams tend not to have a double consonant in word-initial position, QUDALT. He did not go on to offer an interpretation of this supposed word. I think that attempts to rend a Celtic personal name out of the letter as they stand are misguided, and, for once, am compelled to agree with Macalister that the Logie Elphinstone ogham has no literary significance' and 'no assignable meaning' [ 1940:189]. The patterning in the number of strokes per letter seems rather more tbancoincidentally regular, five letters: two with five letters (the same aicme) and three with three letters (each from a different aicme). The lack of vowels is also suspicious and strongly suggests that the sequence is either encoded, or has no linguistic significance. The insertion of the clarifying strokes either side on 2 mitigates against the hypothesis that the carving might have been intended merely to give the appearance of ogham, but suggests rather that it was intended to be read Macalister explains a great variety of codes detailed in the Book of Ballymote tract on ogham [ 1937:40-61], but the possibilities of substitution seem almost infinite and there is little hope of deciding which one might have been used here.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
399
DISCUSSION McManus is at pains to stress diat 'the idea diat Ogam is particularly closely associated with magic and secrecy is false' [1991:163], and emphasizes instead the practical, day-to-day purpose of the script. This does not, of course, preclude, die writing of cryptic messages in ogham, but, from the extant evidence, any such messages would have been highly exceptional. A text may be elliptical without being 'cryptic* (which implies die intention to withhold information), but either way a satisfactory explanation is unlikely to be forthcoming since little is known of the context of the Logie ogham, and the lack of parallels is total
The close proximity of die diree oghams of die Lower Garioch suggests a certain currency of the script in the area. Perhaps the innovatory form of Logie may be taken as evidence of familiarity with the script and confidence in experimenting with its conventions. One thing that is striking about the diree monuments is that each appears to be a re-used prehistoric standing stone. The early medieval re-use of such stones is a marked feature of the archaeology of Aberdeenshire. Simpson list more than half-a-dozen Class I Pictish symbol stones which are clearly or probably re-used monoliths, including Nether Corskie, Echt, Kintore, Ardlair (Kennethmont), Stxathbogie, Crichie, and Kinellar, and of course the ogham-inscribed symbol stones of Brandsbutt, and Logie Elphinstone. To this group must be added the Auquhollie ogham pillar from north Kincardineshire [Simpson 1944:97]. The size and shape of both ogham- and symbolinscribed stones at Newton strongly suggests that they too are re-used prehistoric standing stones. This practice, though far from restricted to die North-East (eg. Edderton 1 ROS, and Irish ogham stones such as Faunkill-in-the-Woods CIIC 66), is disproportionate prominent there. Simpson also lists a number of early church sites in die Urie valley which, he claims are built on the sites of stone circles, including Culsalmond, Daviot, and Chapel of Garioch [Simpson 1944:98].
Archaeologists are becoming increasingly aware of die extent to which Neolithic and Bronze Age monnments in Scotland and Ireland were re-used in die early medieval period. On a grand scale this is seen in Ireland at Tara, Lagore, die Boyne necropolises, and other royal centres [Aitchison 1994] and in
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
400
Scotland at, for instance, Forteviot [Alcock & Alcock 1992. See also Bradley 1993]. According to Bradley the selective re-construction of important monuments was a means by which a local elite could 'strengthen its position through reference to the past1 and that such activity 'was really equivalent to the composition of prestigious but fictitious genealogies* [1987:10], The reuse of more minor monuments» such as the stone circles of the Lower Garioch, may reflect the same appeal to the past as a source of legitimate authority, by those further down the political scale.
BIBLIOGRAPHY
OS Record Card - NJ 72 NW 7.2
Stuart 1856:4, pl.III(2); Moore 1865;74ff; Brash 1879:358; Anderson 1881:74-5, 176, 203; Southesk 1884:189; Ferguson 1887:139; Rhjs 1892:279-80; Nicholson 1896:F; ECMS 175-7, figs. 189-90; Browne 1921:145, pis. xlvii-xlviii; Diack 1944:IV; Macalister 1940:189-90, figs. 1-2; Simpson 1943:153 pL55; Henderson 1967:117; Padel 1972:116-9; Padel 1972b: 197-8; Jackson 1984:179, 181-2; Lines 1989.
There is a collection of archive material relating to this stone in the NMR:
MS/600/1-7 (letters and annotated photographs) DC 11991-4 (early drawings) ABD/502/1, AB/5370/po, AB/5852 (miscellaneous drawings and photographs)
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LOGIE ELPHINSTONE
PREVIOUS READINGS All read clockwise, Where known, starting point, indicated by hour of the clock, Moore
OBHEN PETHECH
Brash
(foigery)
Ferguson
TOGTUCH
Rh?sl892
CAHOHTALT (5 o*clock)
Rh?s 1898
CA(l/ba)T HADBHO (11 o'clock)
Southesk
ATHAT BHOTO (2 o'clock)
ECMS
ABATCAHOHT or ALTCAHOHT
Nicholson
OVOBHVAHTA
Diack
CALTIBHO
Padel
DDALTQU (10 o'clock) (other variants given in main text)
or
CAHOHTABAT
(11 o'clock)
401
Class I Pictish Symbol Stones at Logie Elphinstone.
Logie Elphinstone 1 [ECMS fig. 188]
Logie Elphinstone 3 [ECMS fig. 191]
LOGIE ELPHINSTONE Impression by Marianna Lines (plant matter on cotton) [Lines 1989]
IJT^.4'
LOGIE ELPHINSTONE - Detail of ogham inscription (key to numbering)
402
LUNNASTING
DISCOVERY Rev. J. C. Roger came upon this stone in a cottage in the parish of Lunnasting on the north-east mainland of Shetland [NGR HU 46 65], 'about a month' after being present at the discovery of the St. Ninian's Isle ogham slab in July 1876. Brash quotes Mr Rogers's letter to him of 13 November 1876 in which he described how the stone had been unearthed 'last April', five feet below the surface of the peat bog in the parish of Lunnasting [Brash 1879:365]. In 1876 Roger presented the stone to the NMS (cat. no. IB 113).
SITE The precisefind-spotof the stone is unknown, but it was explicitly stated to be 'at a distance of some miles from any known ruins' [Brash 365], The remains of a possible monastery have been identified at 'Chapel Knowe', Lunna (NGR HU 485 691), but, though it may provide a general context for the stone, there is nothing which links the two more directly [MacDonald & Laing 1968:127-8; RCAHMS 1948:77-8 No. 1280; Small et al. 1973:9].
DESCRIPTION OF OBJECT A smooth undecorated flag-stone incised up its broad face with an inscription in bound ogham letters. Stone:
Sandstone.
Dimensions:
1120mm long x 200-330mm wide x 40mm thick
Condition:
Good. The stone appears intact and only a very small potion of the lettering is lost. Ogham clear.
Goudie diought the stone 'very much in its natural state' [1878:.25], but it appears to have been worked at least to some extent. The front is smooth, but the back rough, and the edges have been chipped to varying degrees. The lower two-thirds of die sides are uneven and chipped while the upper third is crisp. Brash described the stone as 'coffin-lid-shaped' [365], the bottom edge is quite curved and worn smooth,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
403
the squared top appears to be original. The thinness of the slab might suggest that it was conceived as a recumbent monument, but it is not so thin that it could not have stood upright. Bressay, which is of very similar proportions, clearly stood upright, and the dimensions of both are far more likely to result from the constraints of local geology. The arrangement of the text, with the lower third of the slab blank, may support this view, since the monument could have stood upright in the ground without any of the text being obscured. The contrast between the crisp worked upper and rougher unworked lower edges also point in this direction. If sunk into the ground and self-supporting, the slab must have stood less than a metre tall.
DESCRIPTION OF CARVING - DECORATION The sole decoration apan from the inscription, is a cruciform mark in the to left-hand corner of the inscribed face. It consists of a vertical line 80mm long bisected slightly below its mid-point by a horizontal stroke of 65mm. At the top are two 20mm lines, like antennae, meeting the stem at an angle of 45°.
The alleged resemblance to a runic letter is slight and probably coincidental, and the figure is
more readily interpreted as aformof the Christian cross. It is impossible to gauge whether the cross and ogham are contemporary, or whether the former was later added to the standing monument. There are numerous Irish ogham stones with crosses of various degrees of sophistication. Some of these may be contemporary with the inscription, some are clearly not. Macalister's theory that such crosses were added to disarm 'pagan* oghams has been utterly refuted by McManus [1991:55-61 §4.9]. If the Lunnasting cross is not original, it might have been added by a later pilgrim to the site. Macalister noted that the cross was fainter than the lettering and for this reason thought them possibly non-contemporary, but as Padel points out, there are no compelling grounds for such doubts.
DESCRIPTION OF OGHAM INSCRIPTION The inscription, placed centrally on the broad face, starts about a third of the way up the stone. The incised stem continues for 680mm before terminating in a patch of wear just 20mm before the edge of the stone. One, or at the most two, strokes may have been lost in this final spall, but no more; the upper edge
CATALOGUE OF THE OGHAM INSCRIPHONS OF SCOTLAND
LUNNASTING
404
of the slab is intact. The letters, which include a number of forfeda are not, as Padel described them, 'scratched* [120], but carefully and clearly carved.
Though the monument, as a whole, is rather
unprepossessing, the lettering has a calligraphic quality which, in its balance and proportions is rather appealing.
All die relevant strokes are bound, greatly enhancing legibility, and furthermore, word-division is indicated, as on the Bressay stone and the fragment Cunningsburgh 3, with pairs of dots. In addition to the three sets of word-dividers, Lunnasting has thirty-eight characters in all, making it by far die longest single stretch of ogham in Scotland. Over-all, the Bressay text is ten characters longer but it is divided between the two edges of the slab. One would imagine that a text of this length would have had to have been carefully laid out on the slab beforehand to ensure that sufficient space was allocated for all the letters. This might have been done in charcoal or with light scratches. It is possible that the inner line of 4 and the overshoot of the first stroke of 5 may be traces of such guide-lines, but the former could as easily be the remains of a proper stroke, and the latter a slip of the chisel. The layout of the inscription on the stone, occupying the whole of the upper two-thirds of the slab, to within millimetres of die top, is of interest. If the stone had originally stood upright, the greater part of the lower third would have been embedded in the ground, thus, when standing, the text would have covered most, if not all, of the visible surface of the slab.
There is some variation in the length of strokes; short strokes are between 7 - 15mm in length, and the long strokes 20-25mm. If die stone did stand upright, the letters could scarcely have been any bigger without having to split the text into two lines. The back is rough and neither the flat edges nor the arris are conducive to bearing text, so the only alternative would be two parallel lines of text, (cf. Formaston. As at Formaston (and Bressay and Burrian), the small scale of the lettering is such that it can be read only at close-hand. The slope of the letters, especially the angled vowels, indicates that the text is to be read, as we would expect, from the bottom up, as follows [see fig. for key to numbering]:
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
1
LUNNASTING
405
Two long oblique strokes meeting on the stem to form a saltire. The stem originates at the point of intersection. RJi$s and Allen take it as a 'feather mark' or flourish, Macalister and Padel read it as E.
2
Three short strokes to the left of the stem, sloping slightly forward, bound - T.
3
As above - T.
4
A single stroke, curved to form a semi-circle to the right of the stem. The top layer of the area of stone bounded by this stroke has been lost through lamination. There is a feint inner line which is unlike any of the other strokes. It could be the trace of a proper line lost in the spall, or an initial guide-stroke lightly incised. There are no manuscript or epigraphic parallels for this character [see below tor discussion of possible phonetic value].
5
Four short strokes to the left of the stem, sloping slightly backward, bound - C. Thefirststroke has a very faint scratch continuing over the stem, this is not part of the letter proper, but may be the remains of a guide-stroke.
6
Three long strokes across the stem, bound. The sections to the right of the stem are roughly parallel to one another and perpendicular to the stem. To the left, die first half-stroke slopes most sharply backward, the second less so and the third is perpendicular to the stem. Both bind-strokes continue beyond the third stroke but there is definitely no trace of a fourth stroke. There appears to be some confusion or hesitation over the carving of this letter which may have been begun in error then corrected. Probably - U. Since the following stroke is a member of the same aicme it may be that the oghamist began to join up 6 and 7 as a single group, realized the error and prolonged the single stroke of 7 to emphasize that it is a separate letter.
7
A single short stroke to the left of the stem, sloping slightly backward - H. This stroke is like a fourth stroke of <5, but it extends further from the stem. The continuation of the bind-stroke of 6 is probably mistaken and 7 should be taken as a separate letter.
8
Two long strokes across die stem curving towards one another but not touching - )(. This character recurs at 75, though there the first stroke is straight and angled rather than curved and
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
406
smooth. This suggests it may be an elegant, more manuscript than epigraphic-derived, version of the epigraphic ^
^ s e e n on several Scottish oghams [see below for discussion of possible
phonetic value]. 9
Three short strokes to the left of the stem, sloping slightly backward, bound - T.
10
As above - T.
11
Four short strokes to the right of the stem, sloping slightly forward, bound - S. This letter is quite curved and cramped out of the way of 12 and 13.
12
A pair of dots arranged vertically, one either side of the stem (:) squeezed in between letters 11 and 13. If the dots were taken away these two letters would be no further apart than any of the others.
13
A single long stroke perpendicular across the stem, with a horizontal stroke across its right distal tip J_, i.e. a hammer-head A (A or AI). The left distal tip curves slightly forward, a feature not found on any other examples of this character.
14
A single stroke to the left of me stem - H. This stroke is rather short, only half the length of 16 which is ostensibly also an H. It extends the same distance beyond the stem as the left of 13.
15
Two long strokes across the stem curving towards one another but not touching, a version of the character at 8. Here the two strokes are more widely spaced than at 8, and the first stroke is slightly angled [see below for discussion of possible phonetic value].
16
Two short strokes to the left of the stem, sloping slightly backward. These are unbound and rather widely spaced, so should probably be taken separately as HH. This impression is strengthened by the presence of what appears to be a deliberate dot between their tips. It is twice as far from the stem as the dots of 72, and is perhaps an addition meant to clarify two strokes carved too close together. The singletons 35 and 36 are similarly spaced but the text is less cramped at that point. To take them together as D makes for a very awkward reading.
17
Three short strokes to the left of the stem, sloping slightly backward, bound - T.
18
As above - T.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
19
LUNNASTING
407
A single long stroke across the stem with short flat strokes meeting the distal tips in a smooth curve, like an open S. This character is similar to ones appearing on the Lochgoilhead and Altyre stones, but in contrast to them is curved rather than angular. It is closer to the version appearing on Latheron which is interpreted as some form of A [see further discussion below].
20
Five short strokes to the right of the stem, bound - N. In this letter, in 27, and in 22, the final stroke of each group is a little shorter than the rest and the bind-stroke curves down to meet it.
21
As above - N.
22
As above - N. The bind stroke of this letter is formed from a continuation of its final stroke.
23
A pair of dots either side of the stem - word division.
24
A single short stroke to the left of the stem, sloping slightly backward - H.
25
Four short strokes to the left of the stem, sloping slightly backward, bound - C. Each stroke is slightly shorter than the last and the bind-stroke slopes towards the stem to meet them.
26
As above, though with less marked diminuendo - C.
27
Three short strokes to the right of the stem, sloping slightly forward, bound - V.
28
As above - V.
29
Four long strokes across the stem, sharply angled at the mid-point > » > >-, bound - E [see further discussion below].
30
Three short strokes to the right of the stem, sloping slightly forward, bound - V.
31
As above-V.
32
A pair of dots either side of the stem fainter than 12 and 23, but still clear - word division.
33
Five short strokes to the right of the stem, sloping slightly forward, bound - N.
34
Four long strokes perpendicularly across the stem, bound - E.
35
A single short stroke to the left of the stem, sloping slightly forward - H.
36
As above - H. Since the inscription is otherwise consistent in having all strokes bound, we can be confident that these are intended as two singletons rather than as a pair (D), which would give an awkward reading.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
408
37
Three short strokes to the left of the stem, sloping slightly forward, bound - T.
38
Two long strokes, perpendicularly across the stem, bound - O.
39
Five short strokes to the right of the stem, sloping slightly forward, bound - N.
40
A number of short strokes to the right of the stem, sloping slightly forward. Three are clear, and a fourth is almost certain, but any trace of a fifth is now lost - S or N. The original surface does not survive beyond this point and any further carving is now lost. Since the upper edge of the stone is intact it can be seen that no length has been lost, thus it seems unlikely that there were any further letters.
Which yields :
xTTuCUH)(TTS : AH)(HHTT/NNN : HCCVVEVV : NEHHTONs*
FORM OF SCRIPT The Lunnasting inscription is written in a controlled and measured form of Type lie ogham, ogham written on a stem-line with a number of forfeda, including a unique forfid, and other 'late' features; the binding of all letters, and the division of words (see introduction for a discussion of this feature). The ordinary vowel-strokes are long, occupying perhaps three-quarters of the ogham-band, and there are no less than five vocalic forfeda. All the letters slope as one would expect. Component strokes are neatly parallel and well-spaced. Most bind-strokes are parallel to the stem, but some letters exhibit a 'diminuendo* effect, with each stroke shorter than the previous and the bind-stroke sloping down towards the stem, others are parallel except for a final shorter stroke. The bind-strokes mean the inscription is very clear, but in addition the letters are all very generously spaced.
The Lunnasting ogham is notable for its use of no less than six supplementary characters; one unique forfeda, four which appear on other Scottish inscriptions, and one which occurs also on Irish inscriptions:
Character 7: X
This saltire occurs in initial position at Cunningsburgh 3, neither there not here is it
certain whether it has a phonetic value, some kind of E, or a purely graphic function in marking the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
409
beginning of the text. Comparative evidence for the latter possibility is lacking (non-phonetic text markers in Irish oghams are always of arrow form' > ' ) , and since both Lunnasting and Cunningsburgh begin with a double consonant, when geminate consonants in word-initial position are generally avoided in ogham orthography, a phonetic value is to be preferred [see Buckquoy for fuller discussion of this character]. When the character appears in non-initial position, as at Burrian and Newton, and on numerous classical ogham pillars in Ireland, it is more clearly the first Jbrfid l%l or later £ ( = epigraphic /ea7 later / s : / , manuscript digraph ea and diphthong éo) [Sims-Williams 1992:62; see Formaston and Newton for fuller discussion].
The difficulty here, as at Burrian, Golspie, and Formaston, is in knowing what relationship, if any, this x-forfid bears to the character > - ^ . Is it a scribal variant, as it appears to be when the two stroke are written close together, or is it a separate character, as seems to be when they are far apart (Burrian and Formaston have examples of both within the one inscription) ? The contrast appears even more marked on Lunnasting because the non-intersecting examples, 8 and 75, consist of curved lines (though the first stroke of 15 is more angled than smooth). Since Lunnasting, unlike, say Burrian, is a very carefully carved text even minor differences in the formation of characters are potentially significant. The difference in spacing between X - and > <-is perhaps less significant than the feet that -*- and >—<-aie formed in completely different ways: the first comprises two straight, equal, diagonal strokes bisecting one another at right-angles; the latter of two separate, opposing mirror-image strokes, whether straight or curved, X - , -){7 meeting or not. That Lunnasting1 s 15 consists of one curved and one angled line suggests the difference between ^ - ^ and*H*is scribal only.
If the distinction between -x-and >~<- existed in one inscription only, it might be explainable merely as imprecise copying, a non-literate craftsman making too much of casual differences in a hand-written model, but that it is consistently maintained across several inscriptions from widely disparate geographical areas implies it has a more solid basis. If the difference is phonetic then the Lunnasting oghamist differentiated
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
410
between four different kinds of E, the standard (?short) Id 34\ the angled (?long) /e:/ 29\ the x-forfid with its older values (/eá/, /e: /, (ae: /), perhaps forgotten and supplanted by the later digraph ea ^indicating nonpalatal quality in the following consonant - not inconsistent with the Lunnasting sequence) or the diphthong éo, 1\ and the variant > < 8, -){• 15. Note, however, the sole appearance of >
Character 4: TZT- This forfid is unique, occurring neither epigraphically nor in manuscripts. It looks like the lower half of the character -0- which appears at Formaston and in Irish manuscripts representing fO:/ [see Formaston for discussion], but the surface opposite is intact and no carving has been lost on the far side. The character bears some resemblance to U-shaped characters on Cunningsburgh 1, but these appear to be merely ordinary letters with curved bind-stroke on one side only, either across the stem (O) or to one side (D). Such an explanation seems unlikely here, since 4 is quite unlike the other b-aicme letters of Lunnasting, all of which are bound with a straight stroke. Character 4 is a broad and open bowl-shape, not the narrow hair-pin of the Cunningsburgh examples. Since it comes betweenconsonants a vocalic value seem likely. Macalister transliterates it as Y, but he cannot have intended this to have had phonetic significance. Rhjs and Allen took it as O, which is plausible since it has two tips touching the stem. If. however, the faint inner curve is the remains of a genuine stroke, then four tips touched the stem, which might suggest E. Padel thought O or U (presumably because of the physical resemblance) [120], but, influenced by the pattern ETTEC on Cunningsburgh 3 posited E for / 3 / . This character constitutes a 'Catch 22' situation, since its value cannot be determined until the word which contains it is satisfactorily explained, yet this is unlikely to happen until the phonetic value of the letter is established.
Character 13: X 'Hammer-head A' (see Formaston for fuller discussion of this character). There are no 'ordinary' As in die text to contrast with this character, so it could be interpreted as a straightforward A, with a bind stroke for emphasis; though the other singletons in this inscription, all Hs, lack bind strokes.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
411
In the Formaston discussion it was suggested that this character might have the value ai (with the / indicating the palatal quality of the following consonant). Unfortunately since no definitive explanation is available for the word in which it occurs here this value cannot be verified. The Lunnasting example is unique among Scottish examples of this character in having the long stroke curve slightly at its left distal tip, but this is probably merely scribal.
Character 19\f
This 'seriff ed' character is discussed in the entry on Latheron, where it appears to
represent some kind of A. It occurs also at Lochgoilhead immediately preceding a standard A, so its correct transliteration is unclear. The other instances of this character, or versions of it, are all within larger groups, Altyre, Birsay 1 and 2, and Bressay (where all the strokes in a group of five are curly backward-facing versions of this character). The Lunnasting example consists of a straight cross-stroke and a straight serif meeting each distal tip in a smooth curve. This seems pan way between the sharp angle of the Lochgoilhead example and the lower loop of the Latheron example.
Character 29i-$b-' Angled-vowel' (see Formastonfor fuller discussion of angled vowels). This is the only example of an E written with angled-strokes. Since 29 contrasts with 34, an E consisting of straight perpendicular strokes, we may assume some difference in pronunciation. In the Formaston discussion it was very tentatively suggested that angled-vowels may be long vowels. Unfortunately the word in which this letter appears here has not been interpreted, so this cannot be verified, though the standard E in NEHHTON is presumably short. Since 6 seems pan-way between a straight and an angled vowel (U) it could be interpreted as an angled one begun in error and 'straightened'.
In its general proportions the over-all impression of the Lunnasting ogham seems very much of a manuscript-derived ogham aesthetic. The curved lines of letters 4, 8, 13,15 and 19, seem, in particular, more suited to writing on wax, or in ink on vellum, than to carving on stone with a chisel. This manuscript-air is something Lunnasting shares with Burrian, Formaston, Cunningsburgh 1 and 3, and
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
4
*2
Bressay. The superficial resemblance between Lunnasting and Bressay is more a question of carving technique than actual palaeography because the scripts of each have little in common other than the use of dot pairs to indicate word-division. Palaeographically, Lunnasting is much closer to Burrian; both have bound letters (though Lunnasting is more consistent in its use of these), hammer-head As, x-forfeda and >—<- forms, and angled-vowels. Lunnasting is neater and more formal, Burrian is very cursive, but the underlying scripts are very close, though not identical. Even the minor feature of having a bind-stroke which is a continuation of a final consonant-stroke is common to both.
Lunnasting shares a number of features with the Cunningsburgh group of oghams; bound letters and angled vowels with Cunningsburgh 2; initial use of the x-forfid (and the sequence xTTe?C-), angled vowels and word-division with Cunningsburgh 3. Were it not for the lack of bind-strokes, the latter would be very close to Lunnasting, though it is markedly stumpy while Lunnasting is only slightly squat. Except for the lack of word-division and bind-stroke, Lunnasting is close to the script of Formaston, the two having a similar range offorfeda, including hammer-head A, > < ; and angled vowels. The use of the seriffed letter (19) marks out Lunnasting from the above examples and links it to inscription with which it otherwise has little in common. The seriffed letter appears at Birsay 1, which also has the hammer-head A and angled vowels, but only as the central member in a group of five straight strokes.
The close similarities between these northern oghams may be as much a function of date as of geographical location, as suggested by the links with the late eighth or ninth century Formaston ogham. A date in the ninth century would be acceptable for the script of Lunnasting, but an eighth century date is also possible, the hammer-head A is found at Buckquoy, which is probably eighth century, and bind-strokes occur at Inchyra for which a date in the ninth century would be uncomfortably late. The word-diving dots may be suggestive of Norse influence but cannot be taken as diagnostic of a ninth century date.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
413
INTERPRETATION OF TEXT The individual letters of the Lunnasting text are clear, the difficulty lies in transliterating them. Doubt arises over the correct transliteration of H and V (see Introduction), and the six forfeda (see above). Since the text is clearly separated by dots, the division into four words must be taken as authoritative, despite the awkwardness of initial H- in the third word and final -S in the first and just possibly the last. The number of characters per word is thus 11, 11, 8, and 8, a disturbingly regular pattern. The symmetrical distribution may be entirely coincidental, however, and we should be slow to attribute a talismaiiic significance, especially since the final word is uncontroversially a common Pictish personal name.
Comparison with related ogham texts may provide some pointers to the meaning of the other words. On analogy with Bressay's CROSSCC : NAHHTVVDDAddS : DATTRR : ANN (i.e. inscribed object : personal name : relationship : ?personal name), ETTECUHETTS might be parallel with CROSSCC and refer to the stone itself, or a grave, and mean something like 'grave-stone*, 'memorial' or 'tomb', but this is just a guess. The -TANN of the second word suggests it too might be a personal name, in which case the third word might indicate a familial relationship (son of, grandson of, wife of etc.), for instance 'Ahehhttann son of Nechtan', again, cf. Bressay. Alternatively the four may represent grammatically unconnected personal names. Really we are just guessing. The following are offered as observations:
xTTECUHeTTS There is doubt over the correct transliteration of at least three of the four vowels in this word, and also over whether the single C represents /k/ or /%/, and if the latter, how then the single H is to be interpreted. Assuming the distinction between T/TT, and C/CC to be linguistically significant, (since the simplex forms do not occur in word-intial position, we cannot invoke Harvey's findings on the Latininspired avoidance of geminitate symbols in word-initial position), one wonders if, as elsewhere, the distinction might be between lenited/non-lenited (spirantized/non-spirantized), thus the single consonants might represent /6/ and l%l. This explanation, however, would appear to be contradicted by the co-
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
414
occurrence in this word of both C and H, the latter haveing already been posited as a spirant (see Introduction). Word-initial H in the third word, appears to preclude interpreting H at Lunnasting simply as an indicator of vowel hiatus. Given these problems, it is perhaps unsurprising, therefore, that little progress has yet been made in interpreting this word. A possible transliteration is Eatt(o?)cuhgatts.
It is striking that Cunningsburgh 3 also begins xTTEC-, and a similar sequence occurs in the Cunningsburgh I fragment -EHTECONMORS, though in the latter case we cannot be sure it is word initial. All three examples are, of course, from Shetland, and the sequence does not occur elsewhere. The letters immediately following the ETTEC- sequence on Cunningsburgh 3 are lost. Cunningsburgh 1 continues in a different vein from Lunnasting, but whether or not this justifies treating the ETTEC as a detachable prefix, is doubtful. The final -s is not easily explained, but cf. two other Shetland oghams BENNISES#, Bressay and -£S#, St. Ninian's Isle. The name Usconbuts occurs in the Pictish king list [Jackson 1955:145], but the text is corrupt and the reading not secure. Any resemblance to the tribal name Attacotti mentioned by Ammianus Marcellinus is doubtless coincidental. At Brandsbutt, final -S was interpreted as a reduced form of the Irish adverb of place s(e)o 'here*. Alternatively, John Koch has suggested to me comparison with the Breton affixed demonstrative -se [pers. comm.]
Leaving -S to one side, for the moment, -HETT may be compared with the Middle Welsh equative adjective, formed by adding -(h)et to the positive [Evans 1964:38 §41]. John Koch has brought to my attention MW. kyftyt 'of the same length, as long', ky+hyt, equative of Mr 'long' [Evans 1964:40 §42], and the fact that it appears in Old Welsh and Old Breton boundary clauses as cihit-on, cohiton meaning 'limits, extents, what is as far as'. He suggests to me that Ette-ciihett-s might mean something like 'this is/these are as long a s \ with Ette perhaps a Pictish copula + pronoun construction cognate with Olr is-e or it-e. A stone inscribed with this phrase may have functioned as a territorial marker [pers. comm.]. Cunningsburgh 3's Ette-conmor-s may be a parallel construction to meaning 'this is/these are as big as',
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
415
again as some kind of territorial marker.
AHEHHTTANNN Dittography may be sufficient to account for the triplication of N, but three Ns occur in sequence in the Newton inscription. Macalister was of the opinion that a vowel in the final syllable had been elided, and took {his as 'further evidence of the strong accentuation of the first syllable of the word' [1940:.211]. So many male personal names, in both Gaelic and Brittonic, end in -an that this seems the most likely explanation for this word, with its final -ann(n). One of the figures in Y Gododdin is named Aeddan, though AHEHHTTANNN seems a rather outlandish spelling for it (and similarly for Irish Aedari). Comparison may be made with the name of a Pict Aeddan father of Gartnait (AU 649, gen. Accidain) and perhaps grandfather of the famous Cano [Anderson 1926.122]. The same name, perhaps the same individual, is noted as the father of one Talorc (ATig 685, corruptly Aithican; AU Acithaen). If the name is correctly identified then it throws into doubt the explanation for the spelling of spirants offered in the Introduction. Irish Echuiden or even a reduplicated form of Achtán are possibilities for a word which could be transliterated A[hsahhttannn (? Aicheachtan).
HCCWEW The transliteration of this word presents a major problem since there is doubt over the correa value of every single letter. Initial H may represent a spirant, CC presumably does not. A single fern would be fairly straightforward as /w/, but double fern /w/ /w/ doesn't make much sense unless it is a cipherforthe Pictish manuscript spelling Uu /wo/, in names like Uurguisu cf. Irish For-, and Welsh Gwor-. The transcription /xkwoe:wo/ seems meaningless. Padel observed perceptively that the Initial group HCCVV looks fairly unpronounceable, but could in fact be a fairly good attempt to render such a sound as Welsh chW [123]. Macalister drew a parallel between H C C W E W and the xEVV (?/x/EW) of Burrian, also the third word of four. If HCC is an idiosyncratic spelling for l%l, ch, then perhaps the incidences of H
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
416
preceding T and C elsewhere at Lunnasting are to be taken as /6/ and l%l (this suggests an ogham orthographical tradition which has come adrift from contemporary manuscript spelling). But why two Hs for one T in the final word, and one H for two Cs in the penultimate word. And what then of the Hs preceding)(?
NEHHTONs+
We are on firm ground, at last, with the final word. The final letter has at least four strokes and may have had five. If the latter, than the interpretation NEHHTONN is unproblematic. If there are only four, then NEHHTONS poses some problems. Final -s appears at Cunningsburgh 1 - CONMORS (Conmor) and, probably, at Brandsbutt ADDOARENS (ÍAdúarárí), the only possible interpretation of the S in these contexts appears to be some reduced form of the Irish s(e)o, or a Pictish equivalent of the Breton demonstrative -se, both of which would make a certain amount of sense in an epigraphic text like this. The other readings are secure, but since there is the strong possibility that a final stroke has been lost here at Lunnasting, we need not worry about trying to explain an S which may, in any case, be a ghost letter.
NEHHTON may be taken as a form of the Celtic male personal name Nechton ( < *Nektonos [O'Rahilly 1946:368]), well-attested in Pictish sources in various spellings. Jackson suggested that the HH of NEHHTON 'may stand for some kind of weakened ch on its way to i\ i.e. {Nechton > Nehton > *Neiton, cf. the Neitano of the cross-slab from Peebles [Steer 1969] and Bede's Naiton [HE v.21]. He argued further that if the sound had been /%/ it would have been spelled with ogham C or CH [Jackson 1955:145]. From the above discussion it should be clear that this is not necessarily the case, and it is quite possible that H represented /%/. Which makes redundant Jackson's attempt to posit a northern dialect difference on the basis of the Lunnasting spelling [Jackson 1955:164 n.4]
Jackson contrasted the Gaelic preservation of the Celtic spirant cht with the late sixth or early seventh century Brittonic innovation -ith [1953:§60]. On the evidence of Bede, who drew on at least one Pictish
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
^
1
text (the letter to Ceolfrid) in his spelling Naiton (elsewhere in HE he uses Naitan), Jackson suggests that 'at least by the early eighth century' in Pictish cht had become -it rather than -ith [1955(1980): 164], He made this assertion before the discovery of the Peebles stone, firmly in British territory, which may throw into question the precise date and geographical distribution of this innovation. The single T of Lunnasting could, as explained above represent /6/, Given the tendency of Irish scribes to substitute Gaelic forms of familiar or apparently familiar names, and given that most of our Pictish sources are transmitted in Gaelic contexts it is not surprising that the form Nechran is frequent, this need not, of course, have any bearing on the historical phonology of Pictish.
The orthographical diversity of Scottish ogham is reflected in the variety of spellings attested for the syllable /next/ (if the following are correctly identified): NEHT- (Ackergill), NAHHT- (Bressay), NEAHHT- (Formaston), NET- (Latheron).
There is no easy explanation for why the H should be
sometimes single, sometimes double* sometimes absent. The name NechtanlNaiton is well attested in the historical sources as the names of various kings [Anderson 1973:s.n.]. It is also the name of a seventh century saint Neehtan/Nathalan/Mo-Necht(óc/án) [see entry on Formaston for further details].
DISCUSSION While all are agreed that the final word is NechtanlNaiton, no satisfactory explanation has been offered for the previous three words. Jackson threw down the gauntlet in declaring that 'it would take superhuman ingenuity to show that that is Celtic or Indo-European at air [Jackson 1955:141]. This may be going a bit far. The primary difficulty with the Lunnasting text is in arriving at the correct transliteration of the ogham characters, and if, as is the case, many or most of the letters in a word are doubtful then the lack of progress is scarcely surprising. Such a proliferation of H(H)s, especially before Cs and Ts, is striking, but, in the light of contemporary experiments in spelling the spirants /6/ and /g/, not as 'strange and inexplicable' as Goudie complained [Goudie 1878:26]. That sequences comparable to the third word and the beginning of the first are found at Burrian and at Cunningsburgh 1 and 3, all of which contain other
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
418
elements plausibly interpretable as Brittonic, suggests we should exhaust all the Brittonic possibilities for the Lunnasting text, before invoking the survival in Shetland of a non-Indo-European language. Nothing is known of the pre-Norse language of Shetland beyond the ogham inscriptions. It may well have been non-Indo-European, but this should not be our explanation of first resort. As discussed above, the use of double-dot word-dividers need not imply a date in the Norse period and, while it is not impossible, a Norse linguistic interpretation of the Lunnasting text seems unlikely.
With the remarkable exception of Bressay, which has been plausibly interpreted as a copy of an earlier monument, and of Cunningsburgh 2, a fragment of an exceedingly crude cross-slab and the Whiteness fragment, the Shetland oghams are notable for the lack of any carving on them other than the inscriptions. This is not surprising for the small fragments from Cunningsburgh, which do not have a monumental air. but is more striking in the case of the two public monuments of St. Ninian's Isle and Lunnasting, both of which have text only. The precise nature of the Lunnasting monument is not clear. It may have been a recumbent, but is perhaps more plausibly interpreted as a rather diminutive upright slab.
Though
monumental, it is far more modest in scale and ambition than grand southern monuments like Brodie, Altyre, and Brandsbutt. This must be due to more than simply geology and lack of resources. The Lunnasting stone must have performed a different social role than the great cross-slabs of the Scottish mainland. In contrast to these other highly decorated monuments, the primary reason for the erection of the Lunnasting slab was the display of the text.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
LUNNASTING
419
BIBLIOGRAPHY
OS Record card - HU 46 NE 11
Goudie 1878:24; Brash 1879:365-6 pl.XLIX; Southesk 1884:202; Ferguson 1887:134; Rhys 1892:293-4; Rhys 1898:373-4; ECMS 17-8 fig. 12; Goudie 1904:35-8; Nicholson 1896 O, 1904:71-3: Macalister 1940:210-212, pl.IIIh; Diack 1944:x 2; RCAHMS 1946:81-82, No. 1301, fig.592; Jackson 1955:164 n.4; Padel 1972:120-4.
Site only:
MacDonald & Laing 1968:127-8; Small et al. 1973:9; RCHAMS 1948:77-8 No. 1280.
PREVIOUS READINGS Goudie
EATTUICHUEATTS : MAHEADTTANNN : HCCFFSTFF : NCDTONS
Ferguson
TTUIC(u/ng)HEATTS : SEAHHTTANNN : HCFF(e/si)FF : NEHHTONN
Southesk
XTTUICUHMTTS : AHMHHTTMNNN : HCVVEVV : NEDT : ONN
Rhys
TT(o/u)CUHETTS : AHEHHTTMNNN : HCCVVEVV : NEHHTONN
ECMS
TT(o/u)CUHETTS : AHEHHTTMNNN : HCVVEVV : NEHHTONN
Macalister
ETTYCUHETTS AHEHHTTMNNN HCCVVEVV NEHHTONN
Padel
ETTECUHeTTS : AHEHHTTANNN : HCCVVEVV : NEHHTON[s+
[GOUDIE 1878 fig.2 (=1904:36)]
[ECMSfig.12]
LUNNASTING - Ogham-inscribed slab
1
2
3
4
5
6 7
8
9 10 11
12 13 14 15 16 17 18 19 20
21
22 23
m v\m ..* N Wn flfl IMUJW W LU • UW ^WIM 23
24
25
26
27
28
29
30
LUNNASTING - Ogham text with key to numbering
31
32
33
34
35 36 37 38
39
40
420
NEWTON
DISCOVERY This ogham pillar, a scheduled monument, currently stands in the ground of Newton House in the parish of Culsalmond in the Garioch district of Aberdeenshire [NJ 6623 2972]. It has been there since at least 1856 [Stuart 1856:1], but there is some confusion as to its earlier whereabouts. Okasha quotes what she claims to be the earliest reference to the stone in the newspaper Tfie Aberdeen Journal for 12 November 1806, where there is a report that 'A stone Obelisk, of rude workmanship, was lately discovered in the park of Newton near Oldrain' [p.4 col.5 quoted in Okasha 1985:54], Stuan reproduced a letter written to him on 10th September 1855 by the recently resigned Prime Minister George Hamilton Gordon, fourth Earl of Aberdeen, describing from memory the discovery of the stone several decades earlier [Stuan 1856:2]. Aberdeen thought it was in 1804 (i.e. when he was twenty) that he had first seen the stone, 'the inscription on which had, I believe, been discovered by some shepherd boys in the preceding year'. By his account the stone was situated 'in a fir plantation, a few paces distant from the high road, and near to the Pitmachie Turnpike'. He conceded that 'it may appear strange that this monument, which had stood exposed for many centuries on an open moor (for the plantation was not more than fifty years old) should not earlier have received more particular attention', but explained that it had been 'encrusted with the hard grey lichen, of precisely the same colour as the stone itself and that the letters 'were, in fact, scarcely distinguishable'. A context for the discovery is provided by Stuan's Spalding Club editor who notes that 'the Great North Road had only recently been opened and the turnpike gate set at Pitmachie ... this circumstance doubtless increased the number of observers, and assisted in bringing the stone into notice'. The old road had been on the other side of the Gadie burn [Stuart 1856:2 note b].
Stuan visited the stone in 1835 and described its location as 'a spot surrounded by a wood, close to the present toUgate of Shevack, about a mile south of the House of Newton'. This cannot be the 'Wood of Pitmachie' featured on the First Edition of the Ordnance Survey map, since Aberdeen, writing before the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
421
publication of the map says the trees 'have since been cut down'. In any case, Wood of Pitmachie, at a third of a mile west of Bridgend of Shevock, the former tollhouse, is considerably more than 'a few paces distant from the high road'. The Earl of Southesk's description, 'on the slope of a hill above Shevack Burn about 3Á of a mile south of Newton House', is too vague to throw further light on the issue [Southesk 1883:6].
Repeating Southesk's description, Romilly Allen claimed that the original site of the stone is marked on a plan of the Newton estate dated 1760 [ECMS 199 n. 1]. He goes on, however, to claim that in 1794 the inscribed stone was placed behind Newton House. Clearly he has confused this stone with the Class I symbol stone described by Stuart as having originally stood half a mile west of the inscribed stone 'on the march between the lands of Rothney and Newton', which was moved to House of Newton 'upwards of sixty years ago' [1856:1]. When he comes to the original location of the symbol stone, Romilly Allen is vague. He cites no authorities for his information and gives, no dates, either for its discovery or its removal to Newton. He says, Vf is said to have been placed ... about xh a mile W. of the tollgate of Shevock and 1 mile S. of Newton House', and was removed to Newton House 'as early as the beginning of the present century' (he must mean the nineteenth century) [ECMS 178, my emphasis]. This is merely a slightly inaccurate paraphrase of Stuart's account.
The inscribed stone may well have been known of in the eighteenth century but its carving was certainly not recognized before c.1803, and such a pillar, one of several monoliths in the vicinity, would certainly not have been moved to the 'Big House' had it not already been known to be carved. In any case, there is ample evidence that it was still in situ in the 1830s and only subsequently moved to Newton. It can only have been the symbol stone which was moved in 1794. Unfortunately I have not been able to track down the estate map referred to by Romilly Allen, which must have been drawn up at around the time the fir trees were planted, and have thus been unable to check which of the two stones is marked. Nor have I been able to verify Stuart's claim that they were about half a mile apart.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
422
By the 1850s, both stones were within the policies of Newton. The First Edition OS map shows them by the estate road immediately behind the House, next to the kennels, which is where they were when illustrated by Moore [1865]. The stones were moved again, according to Romilly Alien in 1873, to 'a circular railed enclosure, [the inscribed stone] side by side with die symbol stone on the lawn of Mr Gordon's house [i.e. Newton House]'. This is where they still stand today, though the railings have since been removed. [I am grateful to Dr Stephen Driscoll for providing me with a copy of the OS First Edition and discussing the problem with me].
This stone has gone by a variety of names. Nicholson and several ctlier nineteenth centuiy writers refer to it as the 'Shevack' or 'Shevock Stone'. Pinkerton and Skene [1832] called it the Pitmachie Stone. As Diack concedes, however, the name Newton is so well entrenched in the literature that any attempt to reinstate the earlier name is futile [1922].
LOCALITY Stuan describes the discoveiy of several dug graves 'in a sandy lidge neai to the [insciibed] stone', when the ground on which it stood 'was in the couise of being trenched' [1856:1]. His descnption is not detailed, but, he says, theie was no trace of flagstones and the only contents was 'a little black mould'. These do not sound like Bronze Age cists, of which theie aie many in the legion. Unfortunately theie is no indication if the graves weie orientated, but the possibilityremainsthat they may date to the Late lion Age oi even the Eaily Chiistian peiiod. Comparison may be diawn with the 'Catstane', Kiikliston, MidLothian, a Bronze Age standing-stonere-usedfoi afifthcentuiy insciiption [Rutheifoid & Ritchie 1974]: IN OC TU | MULO IACIT | VETTA FILI[- | VICTRf'in this tomb lies Vetta daughtei (?) of VictiifThe insciibed stone is part of a presumptively eaily Chiistian long-cist cemeteiy [Cowie 1980], the radiocarbon dates of which have been recently recalibrated to give a floruit of die cemeteiy of approximately 175 yeais from AD 485-660 [Dalland 1992:204]. The re-use of Bronze Age ritual sites in
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
423
Scotland for early Christian cemeteries is a widespread and notable phenomenon [Rutherford & Ritchie 1974:187, Thomas 1971:53-8]. The two Newton megaliths are as alike one another as peas in a pod and it is tempting to think that they came originally from a single stone circle. Stuart's statement that they were originally half a mile apart would seem to rule this out, but Pinkenon had earlier claimed that they they originally stood 'in the same thicket' [1814 vol.1 advenisement p.xiii, quoted by Stuart 1856:1]. The OS map marks the remains of one recumbent stone circle, on Candle Hill, less than half a mile southeast of Old Rayne/Pitmachie, Stuart mentions another 'formerly in the centre of the present chrachyard' [1856:1 note a]. There are, however, no fewer than three other standing stones marked on the First Edition OS map within a mile of the road between Pitmachie and Newton, and the two under consideration might also have been independant standing stones, singled out for carving perhaps because they were closest to the road/river. The two need not, of course, have been prehistoric megaliths, and it is quite possible that they were first erected to bear their carving.
Bridgend of Shevock, the East Lodge of Newton House [Simpson 1944:101], is situated at the confluence of the Shevock and the Urie rivers. Shevock may be a Gaelic diminutive meaning 'little quiet one* [Alexander 1952:372]. Pitmachie (1362 Pethmalchy) has, of course, first element Pet-, a Pictish land assessment term borrowed into Gaelic. The obscure second element is probably a personal name, cf. Malechi in the Book of Deer [Alexander 1952:351]. Immediately adjacent to Pitmachie is the village of Old Rayne ( < rann 'pan', 'division' [Alexander 1952:359] wherein are the medieval remains of the Bishop of Aberdeen's House, including a moat. Behind the village is 'Gallows Hill'. These features combine to suggest an administrative imponance for Pitmachie/Rayne in the immediately post-Pictish period which may well have had older roots. Early ecclesiastical activity in the area is suggests by a crossincised slab from Culsalmond, at the church dedicated to the British saint Serf, the only such dedication nonh of the Mounth [Simpson 1944:152 (not noted by Watson 1926)]. Stuart notes the holding locally of an annual St. Sair's, i.e. St. Serfs, Fair [1856:1 note a] (see entry on Logie Elphinstone ogham for discussion of the archaeological and historical associations of the wider area of the lower Garioch).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
424
DESCRIPTION OF OBJECT An irregular, ridged pillar incised at the top of one face with six lines of (?pseudo-)alphabetic characters, and down the left-hand edge with an ogham inscription [see fig.]. Stone:
Grey gneissose granite
Dimensions:
c.2.09 x 0.70 x 0.40 m (6'8n 2'9" 1'9" [ECMS] = c.2.03 x 0.84 x 0.53m)
Condition:
Complete. Carving very well preserved.
According to the late G. H. Collins of the Geological Survey, the stone 'has probably come from the auriol of one of the major granitic plutons intruded into the Moine' [Okasha 1984:55]. Like it's partner at Newton, it appears to have been waterworn before carving. Gordon describes both stones as 'boulders of a rock called hornfels, found naturally on the southern slopes of the Hill of Foudland not far from these monuments, though they themselves have probably been carried to their present positions as glacial erratics' [1956:45]. The only carving on this stone is the two inscriptions. Henderson records that 'there appears to be a pocked circle' on the natural fracture 'near the base of the stone' [1972b: 199]. This has not been mentioned in any previous descriptions of the stone and I was unable to find it.
CARVING TECHNIQUE The carving technique of the Newton stone is one of the very few which has been subjected to thorough examination and we are fortunate in having Gordon's detailed account of his ingenious study [1956]. He was prompted by Macalister's assertion that the non-ogham carving on the stone was a nineteenth century forgery and set out expecting to confirm this view [40]. In feet he was able to demonstrate beyond any doubt that on technical and petrological grounds both inscriptions are authentically ancient [46]. For the purposes of his investigation into the Newton Stone, Gordon examined the carving technique of 'a large proportion of the existing symbol stones of North-East Scotland'. Hefoundall bar one (Inchyra) to be carved by the 'pock-and-smooth' technique. This entailed 'pocking with a hammer and punch or with a mason's pick to form a groove which, in section, has a curved profile'. Thereafter 'many, perhaps most of them, [were] improved by tidying the edges of the groove and smoothing out its sides and curved
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
425
bottom, sometimes to the extent of producing a surface perfectly even to the touch' [41].
The Newton symbol stone 'has its outlines deeply and roughly pocked1 [43]. The non-ogham inscription, on the other hand, is of a 'far smoother and more careful technique', it bears 'some traces of its pocked origin, [but] has been carefully smoothed off both on the edges of the groove and in its rounded bottom' [44]. The ogham inscription is 'competently pocked, and though there is some attempt at trimming it is of the slightest'. Gordon believed the turned up portion of the ogham to be 'by a different and more careless hand but... an early one' [44]. He concluded that 'both inscriptions on the Newton Stone appear to be typical work of the period of the Class I symbol stones' [44]. Gordon gives the width of the strokes on the Newton stone as 1 l-16mm and the depth 3-4mm. My own measurements of the length of the ogham strokes are 40-60mm for those on the arris, and on the up-turned stem, an average 75mm for long strokes and 50mm for short.
Before interpreting the Newton inscriptions, two key questions must be addressed. Firstly what is the relationship between the ogham and non-ogham texts, and secondly what is the relationship between the longer portion of ogham on the arris, and the shorter portion on the drawn-in stem. Gordon has shown that the ogham and non-ogham inscription have been carved by different hands, though using a similar technique. This might mean that they were carved at different times, or simply that different people were called upon to handle the different scripts. Most authorities have assumed, either that the two are contemporary, or that the non-ogham is the earlier. I consider it impossible to prove or disprove the primacy of one inscription over the other. What does appear clear, however, is that the Newton texts are not transliterations of one another (there is a marked disparity in the number of characters, for a start). Furthermore there are no sections of one which appear to correspond with any section of the other. The texts of each appear quite different and the monument is not closely comparable to the bi-lingual ogham monuments of southern Britain [Diack 1922:18].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
426
As for the two sections of ogham, the longer clearly predates, though not necessarily by much, the shoner which has been added in what Gordon claims is a different hand. The final portion of the longer text has been defaced. If the two were more obviously in the same hand the arrangement might be easily explained by the carver realizing there was insufficient room for the rest of the text and reaching this imaginative solution for the problem. From about 14 onwards the inscription is more cramped, as if the oghamist was aware that space was running out. Insufficient remains of the defaced portion to be certain if it is repeated at the start of the drawn-in stem. It seems more likely that the shoner section is a continuation of the longer, though it could be a correction or alteration. Despite Gordon's opinion, personally I think the general aspect of the two is so close as to be practically identical, and that the apparent difference in technique may be the result of the shoner portion being less fully finished (i.e. more shallowly cut). The two pans may be taken as roughly if not exactly contemporary.
DESCRIPTION OF OGHAM INSCRIPTION The inscription is 145mm long plus a further 50mm on the up-turned portion. The longer section follows the most obvious ridge on the left side of the stone. This has lead it to wander quite considerably. The strokes of the first ten or so letters are well-spaced, from there on strokes become a little more cramped, though not excessively so. The main difficulties in reading the ogham arise from the lack of differentiation between vowels and letters of the m-aicme (the slope is neither pronounced nor regular), and above all, from uncertainty over the course of the stem-line. One is easily misled about the position of the stem if relying solely on a photograph. It is far easier to identify with the stone infrontof one, especially if one feels the carving. Too much has been read into minute differences by some authorities, generally letters are clearly differentiated one from another.
Virtually all vertical oghams read from the bottom up (left-to-right with the head to the left). Skene was the first to suggest that the Newton ogham reads from the top down (left-to-right with the head to the right)
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
427
[1865]. While unusual, such an arrangement is not unprecedented. The slope of the letters 10, 15 and 16 indicate that this is the correct interpretation. Not only does it yield a (at least partially) intelligible reading, but also it explains why the cramping and the amendment/addition should come at the bottom of the stone. If the inscription began here, one would have expected ample space to be available. If the inscription had run vertically upwards it could have continued across the top of the stone and down, if necessary, as is common enough on Irish stones.
Both Newton pillars consists of very hard rock and 'the influence of weathering on these hard stones amounts to very little* [Gordon 1956:44]. Such weathering as there is 'affects the quality and colour of the surface rather than its planes' [45]. The impression one gets is that the inscriptions are not substantially different today from when they were carved, which is true of few enough of the others. The markedly fainter impression of the lowest strokes (19-21) is surely the result of deliberate defacement, and that of the final strokes of the up-turned portion (25-26) perhaps because they were not particularly deeply carved in the first place. The portion up to 19 is very clear, only thereafter is there any room for doubt.
The highest carving on the arris is two short, vertical strokes (almost horizontal to the stem of the arris) which are indistinct and of a different character to the other strokes, very shallow and broad. They are followed by another indistinct mark, markedly long and obliquely across the arris. Some authorities have read this as A. The mark may be natural, or possibly indicate the beginning of the text, but probably all three preliminary marks should be dismissed.
The rest continues clearly [see fig for key to numbering]:
1
Five strokes, perpendicularly across the arris - 1 .
2
Two strokes 'above' the stem (i.e. to the right as one faces the monument, to the left as one follows the direction of reading down) - D.
3
As above - D. These letters are carefully spaced to prevent conflation.
4
A single stroke across the arris - A.
5
Five strokes across the arris, these are sloped forward and therefore are clearly not a vowel - R.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
428
Relative to 'true' vertical each stroke is further over to the left than its predecessor, but since at this point the arris is sloping in that direction they are, in fact, maintaining their position relative to it. I don't at all concur with Padel that 'this may be too subtle, however' [1972:127]. 6
There follows a gap larger than that between the preceding letters. Then five strokes sloping forwards.
To which aicme these strokes belong depends on the relative position of the
hypothetical stem-line at this point. The arris is rather flat in this area, if anything it swings a little to the right and back. The other vowels in this text (7, 5, 77, 14, and 77) are roughly perpendicular to the 'stem', and since 6 has a pronounced slope vocalic value can thus be dismissed. Padel thought it would be 'stretching it a bit' to read this as R, but thought it 'in line with the propensity for double letters. He suggested also AC, but agreed that the spacing 'does not encourage this'.
Visually, this group seems to be Q, but if one feels the carving the
impression is much more of an R, which gives a much more likely reading. The pronounced gap between 5 and 6 would simply be that expected between two successive letters of the same aicme. 7
Five strokes to the left of the arris, i.e. 'below' the stem - N.
8
As above - N. The proximal ends of strokes 3 and 4 come together in a point. This may be stylistic, but is more likely the result of a small knob on the surface of the stone at this point (however, cf. 14).
9
As above - N. As Padel says, 7 , 5 , and 9 are 'perfectly clear, well-spaced and there is no doubt as to their being to the left' of the stem.
10
Three strokes on the b-side of the arris clearly sloping forward - V. The slope on this group is more pronounced than on the other h- and b-consonants. This may be to clearly differentiate 10 from the preceding group, or, as Padel suggests, to indicate the beginning of a new word, though I am not aware of any other examples of what sounds a rather sensible device. The stem is at this point well over to the right, but is just about to bend back.
11
Two strokes perpendicularly across the stem - O. Padel worries that the left tips scarcely overlap with the preceding V and suggests a 'rather unlikely' D. The disjuncture is the result of the stem
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
429
swinging to the left. Careful comparison with the next letter confirms the reading O. 12
Five long oblique strokes sloping forwards - R. The mid-point of the strokes of 11 and 12 is at the same height, but the two groups are clearly differentiated by length. Also 12 is at a different angle. The one confirms the reading of the other. The first stroke of the group appears to be bent at its mid-point, though the arris is such that it is difficult to decide how deliberate this is.
13
As above - R. The strokes of this and the following letter are proportionately longer than the rest, perhaps because the arris is less well-defined at this point. As Padel points out, the position of this group helps define the stem for the preceding three letters.
14
Four strokes across, and roughly perpendicular to, the stem - E. The h-side distal tips of the first two strokes join for no apparent reason, unless to differentiated them from the preceding letter (cf. 5).
15
Five short oblique strokes to the left of the arris, sloping forward - N.
16
As above - N. Letters 15 and 16 are lightly carved and cramped, though not to the degree that the reading is in doubt. Could it be that the oghamist suddenly realized that there was very little room remaining ?
17
Five strokes perpendicularly across the arris - 1 . These two are closely spaced.
18
The next character consist of a large saltire cross on the stem. The left-to-right diagonal is straight, but the other is an elegant double curve. Presumably this is the x-forfid, a 'shapely example1 [Padel 1972:129]. This may be /k/, /%/, or E, see discussion below.
19
The next section is particularly confused because of the insertion of the drawn-in stem-line and the partial erasure of the succeeding letters. It appears that 18 was followed by two strokes perpendicularly across the arris - O. Padel read instead three Very obscure* strokes spaced 'as for one letter'; the first 'mainly on the b-side', the second 'goes right across', the third 'a mere nick', these he read as U.
20
The next portion is very faint. Either it was never cut very deeply, or it has been partially erased. There is a stroke of similar length to those of 19 but apparently further over to the b-side of the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
430
arris. The drawn-in stem has been insened at this point and obliterated whatever immediately followed. There is the trace of stroke and then a third of comparable length and position to the first. If these diree formed a group - T, there may have been room for four strokes - C. 21
There follows a gap then four faint strokes either across (E) or on the h-side of the stem (C). Thereafter the arris is fractured, a break predating its use for ogham, and no further carving is visible.
The drawn-in stem is insened between the distal tips of the first two strokes of 20. It rises straight up from this position then curves round sharply to the left. The next letter is placed on the curve, but thereafter the stem is straight. It runs parallel to the hypothetical stem on the front face of the stone, a few centimetres to the right of another arris almost parallel to the first. 22
Five strokes across the stem. Since the stem at this point is still curving it is difficult to judge if they are meant to be perpendicular to it or oblique (I or R). Padel thought them too short to be m-aicme. They are certainly shorter than 25 and 26, and almost identical in length to 23, though perhaps the length of strokes varies too much within the inscription to put too much weight on this consideration. Thefinalstroke of 22 is definitely oblique in comparison with 23\ but the oghamist may have wished to keep the constituent strokes of 22 roughly parallel, so again, not too much weight can be attached to this.
23
Two strokes perpendicularly across the stem - O.
24
Four strokes to the b-side of, and perpendicular to, the stem - S. This group is clearly separated from its neighbouring letters and there are no grounds for reading it as 2 + 2 .
25
Five strokes perpendicularly across the stem. These strokes are a bit longer than those of 22 and 23y but rather haphazardly placed. Thefirstis almost perpendicular to the stem, suggesting I, but the second and third are clearly oblique. On balance I take these as R.
26
Four strokes perpendicularly across the stem. There is a taint mark which may have been a fifth stroke, but it is doubtful. It appears more prominent in photographs than with the naked eye. The marked rarity of the fourth character of the m-aicme in epigraphic ogham and the well-
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
431
established precedent for doubled letters in this text throws extra weight behind the interpretation R. Beyond this point the stone is pitted. The marks noted by Macalister are indeed visible, but are so insignificant in comparison with the clarity of the rest of the stem-line letters, that surely they are not letters. Which gives the reading: TDDARRNNNVORENNIxO(t/c)(c/e)
(i/r)OSR(r/n)
FORM OF SCRIPT This ogham is unusual in reading from the top down, but other than that, is unremarkable in terms of script. What is of interest is the interpolated/additional section and the fact that only it has a drawn-in stem-line. Palaeographically, there is no other difference between the two sections, beyond the trend for the strokes to get longer and longer throughout the inscription. Newton thus underlines the lack of significance, especially dating significance, to be attached to the presence or absence of a stem. The feature is far more likely to be related to the form of monument: if an arris is available it will be used, if the inscription is across a flat face, then a stem will be drawn in. If the upturned portion of Newton had used the second arris the letters of the two sections would have overlapped, hence, probably, the shift to the side. Thus the stem-line on the upturned section of Newton is more likely the result of purely functional requirements than any marked difference in date.
The component strokes are not rigidly uniform in length, gradient, spacing, or parallelism of strokes, but unremarkably so, and never to the extent that legibility is threatened. The hardness of the stone and the wanderings of the stem are probably contributing factors, but at least as likely is an aesthetic which did not set great store by uniformity and regularity. The lines of the non-ogham text also waver about and its characters vary in size. There are clear gaps between all the ogham letters. These vary in length but some are quite substantial, especially at the beginning and the very end. Successive letters of the same aicme are carefully-spaced. The vowel strokes are rather long, at least as long as, if not longer than, b- and haicme strokes and almost as long as some m-aicme. The only character which is in any way unusual is
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
432
the x-forfid which is known on numerous Irish ogham pillars (see below), and appears in Scotland at Burrian, possibly at Golspie, and, if > < is not a separate character, at Formaston and Lunnasting and again at Burrian and Golspie. If not just a slip, the merging of tips at 8 and 14 is, however, unique.
There are several Scottish oghams in forms of the script broadly similar to that of Newton, i.e. without any special characters. Bac Mhic Connain, Gumess and Auquhollie, are not, however, closely comparable with Newton, either because of their complete lack of slope on all but m-aicme letters, the more or less even spacing of strokes regardless of letter boundaries, and the comparatively shon length of the vowel strokes. Brandsbutt does not share Newton's generous spacing between letters, and its strokes are less sloped, also it is written on the face rather than the arris, and has a drawn-in stem. Similarly with St. Ninian's Isle which, though similar in general proportions, is much more carefully and regularly laid out. Ackergill and Newton share a number of features, but the former has shorter vowel strokes.
INTERPRETATION OF TEXT The transliteration of the third character of the b-aicme is a function of date and of language (see Introduction), so HI and /u/ are equally possible. More open to debate is the value to be assigned to character 18, the x-forfid. This character is the only supplementary letter to appear regularly on Irish ogham pillars. It has either consonantal or vocalic values /k/ (/xO or /e/. The former is found between vowels on seven pillars (CIIC 113, 124, 141, 155, 197, 216, 301), none of which need be linguistically late, and is also used as the first letter of the formula word KOI 'here', in at least seven cases (CIIC 22, 34, 38, 48, 120, 156, 163, also possibly 26 and 98). This word is not found on later stones [McManus 1991:80 §5.4]. The vocalic value, however, appears to be later. It is standard in post-seventh century ogham (both epigraphic and manuscript) and its occurrence on pillars is correlated with late linguistic forms and the later formula word ANM (CIIC 104, 129, 176, 187, ?201, 223, ?230, 235, 239, 256, McManus 1991 iv). As McManus points out, this vocalic value is never found in Britain, by which he means southern Britain, and like the ANM formula, is in fact restricted to Cork and Kerry. He concludes 'it
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
433
seems clear, therefore, that this was a relatively recent and increasingly popular convention and it may be considered probative for dating purposes' [see McManus 1991:79 §5.3 for full discussion of this character].
Since letter 19 is 0 , it is somewhat tantalizing to have the first two letters of the KOI formula here at Newton. What follows at 20 appears too far over to the h-side to be considered a vowel, but it is otherwise tempting to extract I from the meagre remains of the carving here. The standard KOI formula has the word in second position followed by MAQI MUCOI, *X KOI MAQI MUCOI Y \ and there are only two example of it without MUCOI; CIIC 38 CORBI KOI MAQI LABRIDf... and 120 BROINIENAS KOI NETA-TTRENALUGOS [McManus 1991:171 n4.18]. Marstrander [1911] has suggested that KOI is a word defining locality (an old adverbial locative from the pronominal stem *kel*ko cf. Irish cé, Latin huefhie)y 'here', analogous to HIC IACIT in the British inscriptions though, as McManus points out, it is never used in these [see McManus 1991:119 §6.27].
Most authorities have preferred to transliterate Newton's 18 as P. This is on the model of the use of a little cross to represent /p/ on two Welsh stones, CIIC 327 TVRPILLI and 409 PVMPEIVS, and possibly CIIC 231 in Kerry [McManus 1991:62 §4.13]. In these three cases, however, the cross is clearly placed entirely below the stem on the b-side, not medially on the stem as in all other instances. Whether this is a deliberate modification of the firstforfid to fill an orthographic lacuna (perhaps with an awareness of the cognate values of P and K in the two languages) is unclear. That the Newton example is placed on, not under, the stem would imply that, if it is to be transliterated /p/ then this is probably an independent invention rather than a borrowing from Wales. The drive to accept the value /p/ comes from the resemblance of 17, 18, 19 (I + P + vowel) to the link word ipe at St. Vigeans, and possibly at Fordoun IpOIdamoin, which has been taken as a Pictish kinship term, possibly nephew (a choice influenced by theories about Pictish matrilineal succession).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
Newton clearly lacks the formula word MAQQ(I).
NEWTON
434
Clues to segmentation are as follows: all the
consonants in the inscription are doubled except for V and S. It appears to be a general rule-of-thumb that in Scottish oghams having doubled consonants, word-initial consonants are generally simplex. If we are to take V as the beginning of the second word then we have at the beginning IDDARNNN, which, as explained below, appears to be the attested personal name Idarnon. Taking the -ENN to be the end of another personal name VORRENN we are left with IxO(t/c)(c/e) + (i/r)OSR(r/n), assuming, that is, that the first I is not the genitive termination of VORRENN.
Thereafter the segmentation is more problematic. Is one to read the rest of the text as one unit ? What, then, of the defaced letters 79, 20 and 21 ? Are they to be read in sequence between 18 and 22, or are they to be ignored ? Does one segment end with 19 (or 21) and another begin with 22, or are the two to be read continuously ? Is the 0 of IPO an alternative vocalism of the same word as at St Vigeans, ipe, or, are we to take the Fordoun inscription to indicate that the e could be elided before vowels, hence at Newton ip'o[- ? Are we to abandon the rest of the defaced letters (T or C) ? Iphitt and Ipite for Ibaitte appear as the mythical names of Fomorians in the Irish genealogies [CGH 162 e 51, 162 f 36] but are not attested for historical figures. If we jettison the letters after 0 we are left with ROSRR or IOSRR or IP'OROSRR. A number of Irish personal names, both male and female, have first element or 'gold' ( < Lat. aurum) [Uhlich 1993:287-6], but the remnant OSRR makes no sense. If the C were allowed then we could have CROSRR. The CROS is interesting, but what then to do with the RR ? It seems unlikely that all the letters are to be read in sequence, including the defaced ones, but if they were, then the CCROS or CEROS would be possible, calling to mind Bressay and Burrian respectively, though thefinaltwo letters of the up-turned section cannot be CC.
ROSRR /ROS1R The Irish male name Ros(s) was 'a relatively common name for legendary heroes and founders of dynasties1
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
435
and was borne by a saint of Down [Ó Corráin & Maguire 1990:s.v.]. Ros Ir could be 'Ros the long', ir being an adjective meaning 'long1 [DE,]; though fota 'long' is the more common sobriquet, and epithets are not usual on inscriptions of the period. Irish has the adjective ro(i)sir 'angry, wrathful' [D/L], though similarly this seems unlikely for a monumental inscription. The Welsh hir 'long' is attested as an epithet, though not in inscriptions, but ROS seems unlikely for the common Brittonic male personal-name Rhys (OW Res [CIIC 1011], MW Rys). Ros is a common place-name element in Scotland, either Goidelic '(small) promontory, wood', or Brittonic rhos 'moor, heath, mountain meadow' [Watson 1926:116,496-7], but again a place-name is unlikely in a monumental inscription. Cross-slabs from Kilnasaggart, Armagh [CIIC 946], and Merthyr Mawr, Glamorgan [CIIC 1024], record the donation of land but do not name the places in question. If the penultimate letter is R, then ROSRR is a bit of a puzzle, though calls to mind the final -RR of Birsay 1, and perhaps the final R of Pool. The confusion over this final section is frustrating, but the first two words seem mercifully clear - IDDARRNNN and VORRENN(I).
IDDARRNNN Padel was not the first to compare Newton's IDDARRNNN with the EDDARRNONN of Scoonie and Brodie, and the -pidarnoin of the Roman alphabet inscription at Fourdoun, and to identify all as versions of the name Itharnan, Etharnanus. [See entry on Scoonie for a fuller discussion of this name]. Points to note about the Newton example are, initial I, where the other two oghams have E, and the omission of O from the final syllable. The former brings Newton closer to the roman alphabet spellings of this name, both manuscript and epigraphic. The latter may be a slip, though this is unlikely, or may reflect the weakening of the final syllable after a strong penultimate, or anti-penultimate, stress. The second word of the Lunnasting inscription also ends in NNN, and may reflect a similar process.
UORRENNIFORRENN Allen linked this word to his reading of die Burrian stone - VORRANN. I, however, would interpret that
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
436
stone differently and am thus forced to reject the comparison. Forann is a well-attested Irish male personal name which appears to have been particularly popular among the Ciarraige [CGH]. The related name Forannán was even more widespread, Ó Corráin & Maguire describe it as 'a relatively common name all over Ireland in the early period' [1990 s.v.]. There are also the compounds Forandil and Forannail [CGH], The -ENN for expected -ANN, presents no problems, a number of the Pictish inscriptions have E where one might expect A and vice-versa, including in this position (e.g. St. Vigeans Drosten, see also manuscript Talorgen for Talorgan [Jackson 1955:164]). The letter following this group is I, and until the remainder of the text is satisfactorily explained, it cannot be ruled out that this is the genitive singular termination. McManus provides examples of early post-apocope Irish oghams which have the termination on one name and not on another [1991:98 §5.32].
Common Celtic /w/ (ogham V), became Irish
manuscript/early in the seventh century and Old Welsh gu in the eighth (Old Cornish and Old Breton have the spelling uu also) [Jackson 1953:385-94 §49]. In Pictish it appears to have remained as uluu [Jackson 1955:163]. We may, then, be justified in interpreting this character as the Pictish equivalent of the Welsh gu. McManus has reconstructed the value of the third character of the m-aicme as /gw/ but it rarely appears outside the manuscripts. All the Welsh ogham inscriptions are earlier than the sound-change in question. Thus the name at Newton may be a Pictish cognate of Welsh Guoren which appears in the Llandaff Charters [Davies 1979:s.v,]
Until the final section is clarified, it is impossible to be sure of die syntax of the text. The possibilities are:
(a)
three personal names with no formula words: 'ldarnon (, or :) Vorenn (and) [?]'
(b)
three names plus KOI: 'Idarnom Vorenn koi [?]'
(c)
three names plus ipe: either 'X, Y ipe Z \ or 'X, Y, ipe-Z'
If Idarnon is the saint of that name, then this may be a monument dedicated to him, set up by or in honour of Vorenn and '[?]' or Vorenn ipe-[?] ('Vorenn nephew of [?]' ?). Or, mutatis mutandis, the text may record a gift of land (KOI 'here' ?) to Idarnon's church by Vorenn.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
437
THE NON-OGHAM INSCRIPTION In the past, and especially in the nineteenth century, much more attention has been given to this inscription, than to the ogham one, but still it resists explanation. It is apparently complete and well-preserved, and occupies a little over a third of the face of the pillar [see fig.]. It is placed at the top, roughly about eyeheight. The inscription consists of forty-six characters arranged in six horizontal lines of unequal length (authorities differ over the number of characters because of the difficulty of distinguishing single and ligatured letters). There is no frame and the panel is not demarcated, but top to bottom, the inscribed area measures 530mm. Left-to-right, the six lines of text measure 180mm (four characters), 420mm (nine characters), 330mm (ten characters), 350mm (nine characters), 240mm (five characters), and 360mm (nine characters). Individual characters are between c.35-65mm, the Os which occur in the third and sixth lines are about 30mm in diameter, and the swastika in line 4 is 60mm Tall and 50mm across. Neither the beginning or the ends of the lines are even and the text straggles across the undulations of the surface, making it difficult to photograph in its entirety.
The characters are 'boldly and deeply incised' [Brash 1874:134] and. due to the hardness of the stone, have suffered little from the effects of weathering. Their obscurity cannot therefore be dismissed as wear. Macalister's contention that the text is a forgery [1935:389] has been refuted by Gordon's analysis of the carving technique [1956]. The inscription is complete, well-preserved, and genuine. Nonetheless it has completely stumped all previous authorities. Yet as Alexander Thompson of Banchory House complained in an address to the Society of Antiquaries of Scotland, ' [i]t is provoking to have an inscription in our own country of unquestionable genuineness and antiquity which seems to have baffled all attempts to decipher it, and that too in an age when Egyptian hieroglyphics and the cuneatic characters of Persepolis and Babylon and Ninevah have been forced to reveal their secrets to laborious scholars' [ 1865, quoted by Diack 1922:5]. Thompson was moved to circulate details of the stone and its inscription to a number of antiquaries in the British Isles and beyond. The results were varied and often outlandish, the swastika, in particular, seems to have caught the imagination of certain writers and turned their thoughts towards
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
438
Buddhism [see Thompson 1865]. Brash's verdict of Moore's interpretation will serve for all of them, 'I find a difficulty in reconciling to my mind the probability of Boodhist priests coming from the far east to the far west, to the cold and then almost uninhabited wastes of the north of Scotland, and inscribing Hebrew words in the Ogham character of the Gaedhil of Erinn' [1874:136].
In the face of such theories, Ferguson cautioned 'against our allowing the not greater distortions of the Newton text to deter us from seeking their explanation in the alphabet by which we have hitherto found all the British and Irish bi-lateral epigraphs explainable. And if the mind be kept from remote and foreign analogies, and prepared to recognize familiar forms, though in glyptical masquerade, the eye, even cursorily glancing over the principal Newton legend, can hardly fail to take in some tangible Latin sequences.' [ 1887:138]. I find myself in agreement with Brash, who declared Though unable to decipher this inscription, I am of [the] opinion that it is inscribed in debased Roman letters of a type frequently found in ancient inscriptions' [1879:359], a view shared by Romilly Allen, who described it as carved in 'debased Roman cursive or minuscule characters' [ECMS 199].
Okasha's verdict, 'the text is indecipherable' [1985:55], is over-harsh. While no coherent explanation of the entire text can be offered as yet, certain letters forms may be readily identified. Os and Is are easily spotted, also some characters which may be capital Es. minuscule N.
Others look just like half-uncial S, R, and
Less clear are possible Ts and Gs. It is as if the sculptor, quite unaware of formal
epigraphic capital scripts has carved on stone the most cursive of informal scripts.
Ferguson offered the following reading, which may be right on a couple of letters, even if most have to be rejected [1887:138]: 1st line
?cypher for CELLACH, ILTUF
2nd line
FURTRIN REX
3rd line
GITO HO LOCO
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
4-5th line
USSCETLI FILI SITRGSI
6th line
ligatures for SEPULTUS
NEWTON
439
Diack's version may be nearer the mark, though the resulting reading does not inspire confidence, and there is still considerable room for doubt over several of his characters [1922:9, 12]: 1st line
ETTE
2nd line
EVAGAINMAS
3rd line
CIGONOVOCOI
4th line
URAELISI
5th line
MAQQI
6th line
NOVIOGRUTA
His interpretation of this reading may be dismissed entirely, based, as it is, on his theories of Pictish being a Goidelic language, the genuine precursor of Scottish Gaelic [1922:18, 23].
Browne read
?UOLOUCO?/LOGOUP?TR which he tried to link to a St. Volocus/Wallach, the saint commemorated at Logie Mar (Logie Coldstone), nineteen miles from Newton [1921:112].
One thing which is clear is that the horizontal inscription is not a transliteration of the ogham. There is a wide disparity in the number of characters in each, and none of the sequences legible in the ogham can be picked out in the other text.
NEWTON SYMBOL STONE Since the Newton symbol stone is so closely associated with the inscribed stone a brief description is appended. The pillar is of the same stone and very similar in general appearance to the inscribed stone. It measures c. 2.06 x 0.61 x 0.41m (6*9" x 2* x l'4 n [ECMS]), and is incised using the same basic technique as its neighbour. According to Gordon, the lines of the symbol stone are a similar depth to that of the letter, but, on average a couple of millimetres broader. These lines have been 'deeply and roughly pocked* and there is very little secondary working [1956:43]. Thus they are not smooth, like Brandsbutt,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
440
or even the Newton lettering, but instead the pock marks are still clearly visible. Henderson singles out Newton as a 'particularly impressive' Class I stone [1972a: 171], and it is true that the execution of the symbols and the composition of the over-all design are very pleasing.
The two symbols, a double-disc (without Z-rod) and a serpent-and-Z-rod, are medially placed one above the other in the typical Class I symbol statement form. The double-disc consists of a pair of two concentric rings, each with a central dot, linked by a very thin curving bridge, also with inner lines parallel to the outer. It is a fine example of a typical form. The double-disc with Z-rod is the second most common of all Class I symbols, Alcock lists twenty-eight examples [1989:9]. The rod-less variant, however, is much rarer. There are ten Class II examples (Ulbster CAI, Golspie SUT, Shandwick ROS, Monifieth I, St Vigean's 4, Woodwray FOR, Dunfallandy, Meigle 3, 6, and possibly Fowlis Wester PER), but only four Class I (Drumbuie IINV, Fyvie 2 (left half missing), Logie Elphinstone I ABD, and Inchyra PER). A noticeable feature of the Newton example is the bite missing from the bottom of the left-hand disc, the inner line shadows this notch. This is the only example of a single notched double-disc, but Inchyra is double-notched, it has a bite missing from each disc (opposite sides). Presumably these are modified forms of the standard double-disc, but the significance is lost to us. The serpent-and-Z-rod symbol is not common, occurring on only five Class I stones, including Brandsbutt [see entry on Brandsbutt for fuller discussion of Newton's serpent-and-Z-rod]. The combination plain double-disc and serpent-and-Z-rod occurs on one other Class I stone, at Drumbuie 1, Glen Urquhan, INV, though the symbols are inverted relative to their position on Newton.
[See entry on Logie Elphinstone for a discussion of the early medieval re-use of prehistoric standing stones].
DISCUSSION In archaeological terms, were it not for the non-ogham text, Newton would be typical of the classic ogham
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
441
pillars of southern Ireland (and thus dateable to before the seventh century). The linguistic forms of the ogham text, such as they can be determined, imply, however, a date later than the bulk of Irish oghams. Padel suggests seventh or eighth century, which is probably correct, though more likely the former than the latter, since the individual inscribed memorial appears to be dying out as a monument type throughout the British Isles in the seventh century [Thomas 1994]. A date in the sixth century can scarcely be ruled out, however, without an improved understanding of the text.
It is tempting to imagine some kind of connection between the two carved stones at Newton, but, though they are probably roughly contemporary, it cannot be proved. The formal (archaeological) similarity between the two monuments does, however, underline that ogham pillars and Class I symbol stones are essentially cognate, what Charles Thomas would call 'individual inscribed memorials' [1994]. Though it remains unclear which was found in the wood and which down by the river, it seems that one may have been in the focus of pit- estate (if that is how the conglomeration of features at Bridge of Pitmachie is to be interpreted) and the other at a boundary (subsequently 'the march between the lands of Rothney and Newton').
BIBLIOGRAPHY
OS record card - NJ 62 NE 12
Diack [1922] gives a fairly comprehensive bibliography and Okasha's [1985] is virtually exhaustive. Both lists, however, contain material which is of purely antiquarian interest. Below I give only those works which I consider of immediately practical use.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
NEWTON
442
Stuart 1856:1 pl.I; Skene 1865; Brash 1874, 1879:359-61; Southesk 1883, 1884:191, 1893:54-76; Ferguson 1887:137-9; Rhjs 1892:280-7; Nicolson 1896:5; ECMS 198-9 (symbol stone 178-9); Browne 1921:110-2 pi. xl, xlii; Macalister 1935:389-98; 1940:189, 192-4, pl.IIc; Diack 1944:51-3; Simpson 1944:101-4, figs.; Jackson 1955:139; Gordon 1956:40-46, figs.; Simpson 1965:114; Henderson 1972:199; Padel 1972:9, 153-4; 1972b: 196-7; Shepherd & Ralston 1979:28; Okasha 1985:54-56, pl.III.
Nineteenth century casts of the stone are to be found in the stores of NMS and the National Museum of Ireland, Dublin.
PREVIOUS READINGS OF OGHAM
Skene
UD DDAROT NUN NGORRMAONN EAGE JOSAEI
Brash
AIDDARCUN FEAN FORRENNI EA (i/r)S IOSSAR
Rhjs
IDDA(iq/rhc)NNNVORRENNIP(u/o)A(io/i)IOSIR
Diack
IDDAIQNNN VORRENNI CI OSIST
Macalister
IDDARRNoNN VORRENN IPUOR
Padel
IDDA(r/i)(q/r)NNNVORRENN I(k/p)[U- A
IOS(r/i)(z+/e+)
TON - Location Map. OS First Edition» Sheet 44 (detail), showing Newton House, Bridgend of Shevack, Wood of Pitmachie, Bridge of Pitmachie, and Old Rayne. Scale - 6" to the mile.
-*
• s-.si
NEWTON - Class I Pictish Symbol Stone [ECMS fig. 193]
21
^
(key to numbering)
NEWTON - NMS Cast of ogham-inscribed pillar [© Trustees of the National Museums of Scotland, Reg. IB 108]
NEWTON - Detail of roman alphabet inscription [Simpson 1940 pi. 60]
hoWypxjrO NEWTON - Roman alphabet inscription
443
POLTALLOCH
CURRENT LOCATION This ogham fragment was discovered in 1931 at Bmach an Dmimein, 500m south-west of the North Lodge, Poltalloch, in the parish of Kilmartin, Argyll, (NGR NR 820 971). Macalister was invited to comment on the stone by its excavator, and included it (without illustration) in his Corpus Inscriptionum. It was not considered by Padel who's study was restricted to 'Pictish' oghams. The stone is part of the Poltalloch Collection on loan in the National Museums of Scotland and is currently on display in the Queen Street Antiquities Museum (Cat. No. HPO 470).
DISCOVERY Bmach an Dmimein is situated on the southern promontory of a gravel terrace, skined at its base by a small burn. The crest of the promontory has long served as a gravel pit. Thefirstrecorded archaeological exploration of the site occurred in 1910 when a cist was opened by Prof. Bryce. Following the RCAHMS's visit to the site in April 1928, the landowner, Sir Ian Malcolm of Poltalloch, decided to carry out a more extensive investigation. That summer J. Hewat Craw excavated a number of cists and the earth bank which cut off the tip of the promontory. Returning to the site in 1931 he discovered an ogham fragment 'lying on the ground close to the position of these graves'. In his opinion, there could be 'little doubt that it has been broken at some recent time from a slab of one of the graves' [Craw 1932:448]. Craw published his find in PSAS with photographs, and an analysis of the ogham by R. A. S. Macalister. In 1959 gravel-extraction immediately to the north of the cists revealed the remains of a multi-period settlement. Eric Cregeen conducted a rescue excavation there in 1960-62 on behalf of the University of Glasgow and the Natural History and Antiquarian Society of Mid Argyll. Interim repons were published at the time in Discovery and Excavation in Scotland, but the final report on the late Mr Cregeen's excavations is only now in preparation [RCAHMS 1988:204]. Since then all traces of cists, bank, and settlement have been destroyed by gravel quarrying.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLT ALLOCH
444
SITE The complex, multi-period site at Bruach an Druimein was never frilly investigated and is still not fully understood [Nieke 1984.224]. Its destruction by quarrying means it will remain 'somewhat enigmatic' [RCAHMS 1988:33]. The cup-marked stone and short cists indicate ritual activity and burial at least since the Bronze Age, the La Téne Ic fibula perhaps indicating settlement from the later centuries BC. The earliest non-burial phase has been interpreted as an open-air cooking site [RCAHMS 1988:204]. Subsequendy, a low bank (0.15m high and 7.3m wide) and ditch were constructed across the end of the promontory, these could be traced for 64m and delimited an area of 0.8 ha. Within this boundary were remains interpreted as two successive circular buildings of timber, the earlier having a diameter of over 10m and a floor partly and roughly paved with stone slabs. Immediately to the west were the remains of stone foundations, including a small round hut, inside and around which was evidence of iron-working. Two glass beads date this phase of activity to the second half of first millennium AD, a date confirmed by possibly 8th-10th century imported pottery.
The nature and status of the site remain to be established, and indeed may have varied over the considerable period of settlement. The bank and ditch, or ditches, cutting off the natural promontory, indicates the site was defended to some extent. Evidence for high level craft-production on site is provided by a crucible and a fragment of decorated slate, possibly a trial piece. The whetstones may be paralleled at nearby Dunadd and Loch Glashan crannog, the iron knife blades at Dunadd. Contact with Ireland is indicated by a fragment of a fine glass from the eighth or ninth centuries [Nieke 1984:226], not to mention the ogham inscription. The imported pottery, which may have arrived via Dunadd, implies elevated social status. Clearly this was more than just an 'ordinary1 settlement.
In addition to a number of short cists, the 1928 excavation revealed four slab-lined long cists oriented eastwest, which contained the remains of extended articulated inhumations and no grave-goods. The best preserved cists comprised three slabs on each side, two end-slabs, and three cover-slabs [RCAHMS
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLTALLOCH
445
1992:199]. The lack of grave-goods makes these burials difficult to date, they may represent a small early Christian cemetery. That the burials took place over an extended period is suggested by the fact that, though the first three cists are in alignment, the construction of the fourth disturbed the east end of the third. The chronological and social relationship between settlement and cemetery have not yet been established.
Today the site is known as Brough an Dnnmmn/Bruach an Druimein 'brae of the elder bushes', but according to local tradition its old name was Kill y Kiaran, i.e. Cill Chiaráin, 'church of (St) Ciarán' [Craw 1929:156 n.l citing authority of Mr Donald Campbell of Poltalloch]. The name implies early ecclesiastical activity at the site. Cill has been described as 'the classic place-name element of early Irish monasticism, within ... the early Irish settlements in Scotland' [MacDonald 1979:9]. As a generic, Cill was already productive before the end of the seventh century and continued in use into the ninth century. The term could be applied to a range of ecclesiastical sites 'from the largest monastic civitates to the smallest developed cemeteries' [18], but the majority of sites so named were minor monasteries, 'the lesser and least churches of monastic paruchiae* [19]. See Taylor 1996 for the most recent assessement of cill names in Scotland.
The productive life of the generic Cill is compatible with the date of the settlement at Bruach an Druimein and if this was the name given to the place by contemporaries it may help establish the precise nature of the settlement. Though the imported pottery, beads and evidence of craft-production are more appropriate to a secular power centre, it is possible that the site was a small monastery or 'minster-church' [see Blair & Sharpe 1992 for pastoral model of church organisation]. The earth bank could be analogous to a monastic vallum, though the lack of carved stones or other identifiably Christian remnants weakens the ecclesiastical interpretation. In modern colloquial Gaelic the meaning of cill is 'graveyard' [MacDonald 1979:9], and it is possible that Cill Chiaráin referred to the cemetery only. Cill- sites often became
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLT ALLOCH
446
medieval parish churches [19], and that this did not happen here, suggests that Cill Chiaráin declined at an early date.
The dedication to Ciarán is an interesting one. A total of twenty-two saints are recorded bearing this name, of whom the two most famous are Ciarán of Clonmacnoise, who died in 549, and the pre-Patrician Ciarán of Saigir. The Ciarán commemorated in Scotland is probably theformer[Watson 1926:278], Dedications to and other associations with Ciarán include, amongst others, five in Argyll (Poltalloch, Campbeltown, Islay, Lismore and the north-east end of Loch Awe), two in Fortingall, Perthshire, and one each at Barvas, Lewis and Dailly, Ayrshire [Redford 1988:155; Watson 1926:187, 278].
Beyond these dedications, there is no evidence for any Scottish involvement on the pan of the Clonmacnoise pamchia. Aidan MacDonald has made the intriguing suggestion that commemoration of Ciarán, and of others such as Brendan of Birr, in fact reflects the pamchia of Iona [1979:15]. According to Adomnln the festival of Brendan was observed at Iona [VC III.2], and it is possible, though not proven, that his culms may have spread to Iona's daughter houses under official sanction. By analogy, the same may be true of Ciarán, for early links between Iona and Clomnacnoise were strong [VC 1.3]. The connection between nearby Dunadd and Iona has been amply demonstrated [E. Campbell 1987]. The form of the cross on the Barnakill slab, just south of Dunadd, is also Iona-influenced. That a Cill site not far to the north of Dunadd would be similarly influenced by the great monastery is not surprising.
The involvement of Iona in the introduction and spread of ogham is still uncertain. There are numerous inscribed stones on the island, though none are written in the ogham script, nor does oghamfeatureat the great Columban sites in Ireland. It should be noted however that there are two oghams, of greatly different character and date, extant at Clonmacnoise [Manning & Moore 1991]. (For a modern ogham inscription on Iona see Appendix - Dubia).
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLTALLOCH
447
LOCALITY To understand the importance of the Poltalloch site, one must look, not only at its strategic location but also at its proximity to other major sites of the early medieval period. Taken together, the Kilmartin and Dunadd areas comprise the hinterland of Loch Crinan which gives access to the sheltered sound of Jura and thence nonh to the Western Isles and south to the Nonh Channel and Irish Sea. Not surprisingly the region has been an imponanc conduit of external influence from earliest times. Poltalloch itself is less than two miles from the sea.
From the Neolithic to the Bronze Age the Kilmanin Valley was a focus for ritual activity [RCAHMS 1988; Ritchie 1988:48-52]. The impressive remains include an extraordinary linear cemetery, groups of standing stones, and the greatest concentration of rock an in Scotland. In the early Bronze Age there was at Poltalloch a flat cemetery of grooved cists, one containing a necklace of jet beads. A halberd found there points to Irish connections already in the Early Bronze Age [Scott 1966:39].
The Atlantic coast of Mid-Argyll and Knapdale together with the Kilmartin/Dunadd area have produced no less than 83 early Christian carved stones from 26 sites, a remarkable concentration. Kilmanin Church itself (RCAHMS 1992 No.68) has a number of carved stones including a ringless cross. Slightly further afield there is high quality Christian sculpture at Kilmichael Glassary (No.69); and from Baniakill (No. 13), three miles away, an inscribed cross slab. The great citadel of Dunadd is less than three miles from Poltalloch, and over the hill, the recently excavated high status crannog at Loch Glashan, is only 6 miles to the east. The natural strategic importance of the region was heightened during the early medieval period by its proximity to one of the principle routes east into Pictland, and by its forming the shifting boundary between the two great segments of Dál Riada, the Cenél Loairn and the Cenél nGabráin [Bannerman 1974:112].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLTALLOCH
448
DESCRIPTION OF OBJECT A small pillar-like fragment, badly worn in places, sharply carved on one arris with ogham letters. Stone:
Sandstone
Dimensions:
254mm long x 82mm broad x 57mm thick
Condition:
Fragmentary, numerous chips lost, but carving generally clear
In Craw's 1929 illustration the stone is inverted in comparison to how it is displayed in the National Museum and illustrated in the Argyll Inventory. Since both ends of the slab are missing it is impossible to be cenain which is the correct orientation, but the NMS/RC AHMS one is followed here for convenience and because it produces a plausible reading. The RCAHMS Inventory drawing is very accurate.
DESCRIPTION OF OGHAM INSCRIPTION The ogham letters are carved along the arris of the slab in very fine, sharp strokes. Their general proponions are long and slim, and their profile V-shaped. The variation in length of stroke within each group, and the manner in which individual strokes trail away at the ends, gives the impression that they were confidently cut. Staning at the bottom [see fig. for key to numbering]: 1
The surface of the opening section is almost completely worn away. There are possible traces of strokes on, or to the left of, the arris, but these are doubtful. Next there appears to be two or more strokes to the left of, and perpendicular to, the arris. These are now faint and worn, not nearly as distinct as the rest of the inscription, but clear enough to be assigned to the h-aicme. These are followed by a spall. The proportions of the space and the preceding strokes suggest T, C or Q followed perhaps by a vowel. On examining the apparent A between theflakeand the C with a strong magnifying glass, Macalister decided it was not artificial [CIICI p.485]. Since two letters of the h-aicme succeed one another the likely explanation is that sufficient space has been left to differentiate one from the other.
2
Four short strokes to the left of, and perpendicular to, the stem - C.
3
Five long strokes, across the stem, sloping forwards - R.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
4
POLTALLOCH
449
Two notches on arris - 0. There is space for a third notch but nothing is visible nor is there any apparent damage at this point. Macalister rejects U, though there is room for it.
5
Five short strokes to the right of, and perpendicular to, the stem - N.
6
There follows a gap with room for a notch but no evidence of one. Previous writers have unanimously identified it as A. The intention, however, may be merely a blank space to differentiate between two successive letters of the B-aicme. The arris is too worn at this point to be certain.
7
Two short strokes to the right of, and perpendicular to, the stem, with part of a third clearly visible. The rest has flaked off, though slight irregularities in the residual surface may be vestiges of further strokes - V, or a subsequent member of the b-aicme: S or, perhaps most likely, N.
Which gives the reading: -CRONa(n)-
Macalister was of the opinion that the spall after the first, indistinct, h-scores 'detached itself when the stone-cutter began his work, and that he made a fresh start at the present C [p.484-5]. I am inclined to reject this scenario on the grounds that a conscientious oghamist would have easily removed traces of any botched letters. I would prefer to interpret the initial strokes as the remains of a letter, perhaps Q, and assume that a number of letters have been lost prior to it. Jackson's opinion was published only as a personal communication by Margaret Nieke. We may assume it was based on his personal inspection of the fragment in the NMS.
FORM OF SCRIPT The Poltalloch ogham is written in a classic Type la form of the script with no written stem-line and vowel notches occupying no more than an eighth of the ogham band. Strokes of the b- and h-aicmi are perpendicular to the stem, the m-adme strokes slope well forward, but occupy no more than the middle two-thirds of the ogham band. Strokes are evenly spaced within letters, though some groups are more cramped than others. As far as can be judged from the extant section there is extra spacing between letters.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLTALLOCH
450
The strokes are long and chin and closely packed, thus the letter profile is noticably elongated, in contrast to other Type la oghams in Scotland, such as Auquhollie, Newton, and Dunadd. Pool offers a parallel for these slender letters though there it is somewhat mannered. An elongated letter profile is a marked feature of certain oghams from Co. Cork.
INTERPRETATION The letters CRON are not in doubt. Less secure are the A and N. As they stand they could be interpreted as the not uncommon early Irish name Crónán, which derives from the colour term cruris 'dark yellow, reddish brown, swarthy' [DH\, plus the diminutive suffix -agno- > CRONAGNOS, gen. CRONAGNI. The fonn Crón and its derivative Cróine appear as the name of a number offemalesaints [Ó Corráin & Maguire 1990:62].
There is no way of knowing if the doubtful N was the last letter, but it is almost certain that part of the text has been lost from the beginning. The possibility that the letter immediately preceding the C was Q opens up the interesting prospect of the Poltalloch ogham conforming to the typical X MAQ Y formula, either MAQ or MAC [McManus 1991 §5.19]. The extant traces neither fully support nor refute this hypothesis. The principle objection is linguistic, one would expect the name to appear in the genitive case, CRONAIN.
The final letter is definitely not G, so the inscription must post-date the sound change gn > /?, which occurred in Irish by the end of thefifthcentury [Jackson 1953:461]. The 'knife-cut' style of the script is distinctive, but further work is necessary on the development of ogham script in Ireland before any chronological or geographical significance can be attached to it. Since the archaeological context of the slab is not closely dateable, the only other possibility is to identify the individual commemorated. Crónán was, unfortunately, a not uncommon name among early Irishmen and a number had Dál Riadic connections. Without a patronymic qualification there is little chance of narrowing down the list of several
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLT ALLOCH
451
possible Cróoáns. Adomnán's Vita Columbae records three Crónáns -Crónánthe poet, whose untimely end Columba prophesies (1.42); Crónán the Munster bishop who visits lona (1.44); and Crónán son of Báitán who, in the name of Columba, is said to have thrown the spear which killed the villainous Lám Dess (11.24). As contemporaries of Columba, the three can be dated to the final third of the sixth century. The first was apparently unconnected with Scotland. The second visited Columba on lona, and doubtless returned to Munster, his pilgrimage complete. The third, Crónán son of Báitán, is a possible contender for the Poltalloch commemoration, but cannot readily be localised. He slew the malefactor in ilia insula quae latine longa vocitari potest In the island that in Latin may be called Longa', which remains unidentified. Marjorie Anderson has backed away from the earlier tentative suggestion that it should be equated with modern an Innis Fhada, die Outer Hebrides [1991.1xxv]. Even if the island were, however, to be identified, the implication for the localization of Crónán son of Báitán is uncertain. At die time of the engagement he may have been pan of an expeditionary force far from home. Marjorie Anderson has suggested that Crónán son of Báitán may be of the family of Cineal Bhaodain (Cenél Bháitáin) [1991:xxxii], which gave its name to Kinelvadon, now Morvern [Watson 1922:122]. This Báitán, a grandson or great-grandson of Loam Mór, is recorded as having had a brother and a cousin called Crónán [Anderson 1991:xxxii] and may perhaps be identified as the fifth, unnamed son of Eochaid son of Muredach discussed below [see Bannerman 1974:71 for genealogical diagram].
From contemporary sources it is clear that the upper echelons of Dál Riadic society, at least, were highly mobile, travelling with ease by sea among the Scottish islands and back and forth to Ireland [see Bannerman 1974:149-50 for list of sea voyages mentioned in VC and the lona chronicle]. Several of the dramatis personae of VC are located far from their place of origin [Bannerman 150] and it would be unwise, therefore, to be too narrow in attributing a locus of activity to individuals. While a cenotaph may indeed have been erected back home on family land, a person would surely have been buried close to where they died.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLTALLOCH
452
With this in mind, it is intriguing to note the commemoration of CRONUN MAC BA1T[- on an ogham pillar from a farm at Ballyknock, Barony of Kinnatalloon, Co. Cork (CIIC 90). McManus suggests CRONUN is an error for CRONAN [1991:§ 6.12]. On the basis of the post-apocope form of the MAC, this inscription should perhaps be dated 'considerably later than the beginning of the seventh century' [McManus 1991:99 § 5.33]. Could this be the memorial of the heroic defender of the church, the Dál Riadan Crónán mac Báitáin ?
The traditional genealogies preserved in the Senchus Fer nAlban record four Crónáns who flourished in the first half of the seventh century (if such a date is accepted for the original composition of the text, since they are all in the 'current' generation of the text). There is Crónán, son of Tulchán, grandson of Barrfind, great-grandson of Madsluaig and great-great-grandson of Oengus Mór, though it is unlikely that a scion of the Islay-based Cenél nOengusa would be commemorated at Poltalloch. Whether the commemorand is more plausibly sought among the Cenél nGabráin or Cenél Loairn depends on the date of the inscription, because at different periods the Dunadd/Kilmartin area was in the hand of each in turn.
The name seems to have been particularly popular among the Cenél Loairn. Crónán son of Cathdub and his cousin Crónán son of Eochaid, were grandsons of Muiredach, head of one of the three divisions of the Cenél Loairn, and therefore great-grandsons of Loam Mór. The latter, Crónán son of Eochaid, and his brother Bledán are recorded as having twenty houses between them. On analogy with the early Irish law tract Ciith Gablach, this would place them in the rank of aire ardd [Bannerman 1974:136]. As suggested tentatively above, Crónán son of Eochaid and Crónán son of Cathdub may have been respectively, the uncle and first-cousin-once-removed of the Crónán son of Báitán recorded in VC, though this might stretch the chronology of the Senchus Fer nAlban.
Of the Cenél nGabráin, one of the eight sons of Eochaid Find, and therefore a grandson of Aedán mac Gabráin, was called Crónán. The death of his first cousin Conall Crandomna, king of Dál Riada, is noted
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLTALLOCH
453
in AU 660 [Anderson 1922:176-7; Bannerman 1974:103-4]. If, as seems likely, Dunadd was the chief seat of the Cenél nGabráin in this period [Bannerman 112], what better place for the king's cousin, and member of his own derbfine, to live than on a defended promontory at nearby Poltalloch ? This however is mere conjecture, the person commemorated in the Poltalloch ogham could be any of the above Crónáns, or indeed any of the multitude of Ciónáns whose names were never recorded in the historical sources. If the suggested reading of the inscription is accepted, the commemorand is in fact the son of Crónán, but there is no indication at all of his own name. It is hardly likely that Kill y Kiaran is a corruption of Cill Gwónáin (there are numerous saints called Crónán), and the slight correspondence between the two is doubtless coincidental.
DISCUSSION As excavator, Craw's belief that the oghamfragmenthad recently broken off one of the grave slabs, must carry a certain weight, however it is curious that the fragment was not noticed during the excavation. Apparently Craw did not test his hypothesis by comparing thefragmentwith the excavated slabs. By that time they were perhaps long gone. This is unfortunate because an element of doubt remains about the fragment's exact provenance.
Irish ogham pillars are frequently found re-used in souterrains where their bulk made them ideal roof buttresses and lintels. The Poltalloch example is, however, quite different in having been found inside a cist. If the position is primary, this is a unique instance of an ogham commemoration inside a grave. Who was the intended reader of such a text, if any ? Was it meant to be read on Doomsday to identify the person who rose from his or her aligned grave to face Judgement ?
It is possibile that the usage as a side- or cover-slab was secondary.
One of the nearby short-cists
(No. 104.2 grave 4) had a cup-marked stone re-used as a cover-slab [RCAHMS 1988:17], and a number of Pictish Class I symbol stones, e.g. Dunrobin and Golspie [Close-Brooks 1980:343-4, 334], have been
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POLTALLOCH
454
discovered re-used in burial cists. Craw may have been mistaken. Macalister was of the opinion that the fragment had 'split off from the top of a pillar-stone1 [CIIC I p.484], but it is hard to imagine the dimensions of a pillar whence came such a tiny inscription.
The provenance of the Poltalloch ogham is further obscured by a tantalizing statement made in 1875 by Capt. T. P. White. In his Archaeological Sketches in Scotland - Knapdale and Gigha, Capt. White refers to a slab, which he does not seem to have seen in person, 'found in a cist near Kilmartin which has been thought to be possibly Ogham-marked' [28]. No further details are given and the stone is never again mentioned in the literature. The fact that Craw does not mention it suggests it was longforgottenby the time he made his discovery. The reference is unlikely to be to the extant Poltalloch fragment, because, given the circumstances in which it was found, it would be difficult to account for its whereabouts between the early 1870s and 1931, unless, for some reason the stone had been reburied. Capt. White may have been referring to another stone from the site, possibly even the missing portion of the extant slab. But whether to that, or to an ogham-inscribed cist slab from another site in the locality, the coincidence is striking.
BIBLIOGRAPHY
OS record card NR 89 NW 37
White 1875; Macalister 1932 in Craw 1932, 448-50, fig. 2.; CIIC I p.484-5 No.507 not ill.; Nieke 1984:264; RCAHMS 1992:199 No.9I, figs. A and B.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND Site:
POLTALLOCH
455
Craw 1929:156-8; Craw 1932:445-50; Cregeen 1960, 1961, 1962; Campbell & Sandeman 1964:62, No.402; Nieke 1984:224-6; RCAHMS 1988:204 No.350, 80-3 No. 104, 4; RCAHMS 1992:199 No.91 long cists.
PREVIOUS READINGS
Macalister
CRONan
Jackson
A7CRON
(pers. comm. to Margaret Nieke [Nieke 1984:264])
(key to numbering)
^ ^
7
6
POLTALLOCH - Ogham-inscribed fragment [® Trustees of the National Museums of Scotland, Neg. P9555]
456
POOL
DISCOVERY This ogham-inscribed slab was discovered in 1984 during the University of Bradford's excavation of the multi-period settlement site of Pool on the island of Sanday, Orkney [NGR HY 6194 3785]. When found, the slab was in a secondary context, broken and reused as an orthostat, i.e. upright. The final floor level of the structure partly concealed the inscription, so it is clear that, unlike Birsay 3, the Pool ogham cannot have been carved in situ. It is now on display in the Tankerness House Museum, Kirkwall, Orkney. I am most grateful to Anne Brundle, the museum curator, for facilitating my visit to examine the stone.
SITE The fertile, low-lying island of Sanday lies in the north-east of the Orkney archipelago, between Eday and North Ronaldsay. On the west coast of the southern tip of the island, on the shore of Pool Bay by North Mire farm, is the archaeological site known as Pool [see map]. By the time excavation began in 1983 perhaps as much as half the site had been lost to the sea, though Hunter thinks the current coastal location is the result of sand movement and erosion and that originally the settlement was 'almost certainly further inshore1 [1990:177].
The detailed programme of excavation and field survey, conducted until 1988, revealed evidence of occupation stretching back as far as the neolithic of the fourth millennium BC. Of relevance to the oghaminscribed slab are the levels dating to the Iron Age and the Pictish and Viking periods. Though Lamb identified between six and ten potential broch sites on Sanday, Pool is not one of them [Lamb 1980]. The implications of this for understanding the social position of the inhabitants of Pool in the period have yet to be explored. Well preserved stratigraphy, pottery evidence and C-14 analysis have allowed excavators to sketch an outline history of the site [Hunter 1990]. In the fifth century a large circular or sub-circular building of internal diameter 6mformedthe core of the settlement. A secondary phase provided evidence
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
457
of ferrous metal-working. The settlement expanded and at a subsequent stage in the sixth century a rectangular courtyard at least 20m long was formed. Hunter describes how this central courtyard gave separate access to a number of 'cellular units' and concludes that its creation 'was of sufficient impact on the existing organisation of the complex to have sociological implications' [1990:185]. The courtyard is of particular relevance because two carved stones were found incorporated into it as constructional elements - a symbol stone, and the ogham-inscribed slab. Radio-carbon dates in the sixth century were obtained for the building of the courtyard thus providing a terminus ante quern for the carving of the stones.
The symbol stone is a chunky slab, 0.73 x 0.43m, of irregular, roughly rectangular, shape. It has been lightly pecked with an unembellished double-disc symbol and two further indistinct curvilinear shapes. The carving occupies the upper half of the stone and the excavator thought it 'unlikely to have been set in the ground with the symbols still visible' [Archaeology Extra 3:2]. It would be possible to imagine this as a miniature symbol stone, but its scale, and the rudimentary nature of both design and carving technique, combine to suggest a different context. It was found recumbent and inverted, circumstances encouraging Hunter to feel this was a deliberate deposition of a carved stone, one to which he imputed ritual significance. The Pool carving is more akin to the symbol graffiti from caves in Fife and Moray than either the large symbol stones, or smaller plaques such as those as those from Burghead MOR and Dunnicaer KCD. The not infrequent re-use elsewhere in Scotland of symbol stones grander than this makes it clear that individual inscriptions could become obsolete. While Hunter's explanation is not implausible, it may simply be that an old unwanted symbol stone was pressed into service in this major building project in which, presumably every available stone would have been utilized. Whether or not the carving of the symbol stone was related to its incorporation in the courtyard, the use of the ogham-inscribed slab was almost certainly secondary. The function of the courtyard has not been established. The orthostat partitions and presence of hearths suggest it may have been roofed, but since die carving of the ogham inscription dates to an earlier phase it is perhaps irrelevant. Of course, the previous history of the slab is lost to us. The creation of the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
458
courtyard, and the accompanying paving-over of the old neolithic mound, marked, in the excavator's opinion, 'the maximum expansion and development of the community' [Hunter 1990:187]. There was no major change after the early seventh century, or thereabouts, but instead a long period of gradual contraction and decline. The site was still 'fairly substantial' at therimeof the Norse settlement in the late eighth or early ninth century, and the types of artefact recovered indicate a continuing native presence 'in what might be defined as a phase of cultural interface' [191]. In this later period the site flourished, due in part perhaps to the cultivation of flax as a luxury or trade crop [Archaeology Extra 2:3]. Occupation continued well into the medieval period.
In addition to the ogham-inscribed slab, a small bone pin wasfounddecorated with lines perhaps intended as ogham [see Appendix - Dubia].
DESCRIPTION OF OBJECT A broken building slab incised with an incomplete ogham inscription, otherwise undecorated [see fig.]. Stone:
Orkney flag
Dimensions:
c.0.85 x 0.50 x 0.60m
Condition:
The stone is fragmentary. The inscription may be incomplete though the extant portion is well-preserved.
In its current condition the slab is fragmentary, though whether it was broken to fit a particular context in the new structure, or re-used because it had broken is unknown. It is clear, however, that the ogham was carved before the fracture. Since no care had been taken to keep the text visible we can infer that it's message was probably obsolete by then. Certainly it has the informal air of a piece of casual graffiti, and it seems unlikely that the stone was erected primarily to bear the ogham text. If the slab had also been used as a constructional element in a previous building it is possible that, like some of the Birsay oghams, the Pool example was carved in situ in its earlier location. The excavator's preliminary suggestion that this was part of a symbol stone was based on a mistaken assumption about the use of ogham in Scotland,
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
459
and can be discounted. I am grateful to Dr John Hunter, University of Bradford, for kindly supplying me with a photograph of the ogham stone [see fig.].
CARVING TECHNIQUE The ogham is inscribed across the face of the slab, more or less half-way along its length. The inscription begins hard against the fractured edge and runs perpendicular to it, just over a third of the way across the broad surface. The stem stretches right to the edge of the stone which lead the excavator to refer to the 'partial survival of the inscription*. Certainly the positioning of the inscription on the slab is rather puzzling if die extant inscription is complete. It is possible that die ogham may have run across two stones, as perhaps happened with Birsay 3, and that what we have here is the end of a longer text. The stem is visible for about 200mm before petering out just after letter 7. Despite personal examination of the stone it is still unclear to me whether any letters have been lost at the inner end of the inscription. The successive strokes of the R are increasingly worn and die stem is visible for a short distance beyond die last. After this there is a slight change in die surface texture of the slab which is perhaps die result of weathering, though absolutely no trace of any lettering survives in this portioa The inscription may be complete as it stands, or it may have lost sections of indeterminate length from either or both ends. The missing portion or portions might amount to no more than a few letters, already the Pool inscription is a letter longer than the apparently complete graffiti ogham Blackwater 2.
In what follows the inscription will be described as if the fractured edge is horizontal at die base with the stem rising vertically. This is merely for die sake of convenience and is not meant to imply that this was necessarily die original orientation. The letter strokes, which have been lightly incised with a sharp blade, are long, thin, and closely spaced, with a defined v-shape section. The ductus is from lower right to upper left, as shown by the splaying of component strokes, die disparity of length between die left tips of strokes, and die sharpness of their right tip. This is the direction one would expect of a left-handed oghamist using hammer and chisel working horizontally in the direction of reading towards the centre, or a right-hander
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
460
working vertically in the same direction. Of course, the direction of carving need not necessarily reflect the direction of reading. It appears that the stem was (re-)cut on top of the strokes. The fine incised technique is in contrast to the pecking employed on the symbol stone.
DESCRIPTION OF OGHAM INSCRIPTION The inscription could be read in either direction. There may be a very slight forward slope on letters 4 and 6, which would imply reading from the interior to the margin, but the general impression is of perpendicularity of letter-strokes, so it may be wrong to make too much of this. A number of features combine to give the Pool ogham a cursive appearance, for instance the slight nature of the incised lines, the scale of the inscription in relation to stone as a whole, and the irregularity of some of the letters. The overall impression, however, is neat and controlled rather than haphazard. Other factors to note are the slight differences in length among component strokes of individual letters, the fact that strict parallelism is not maintained within letter groups and that some of the strokes are slightly curved or bent. The informality is exemplified in, for instance, die difference between die two Ts (the first very neat and tight, the other more lax) and between the two Rs (the second much more sloped than the first). Also there is the apparently erratic spacing within and between letters. With an inscription as short as this, and which is probably incomplete, it is difficult to establish the significance of the spacing between letters. It is almost as if die letters are written in pairs with 213, 415, and 6/7 hard up against each other but widely spaced from the other pairs. I know of no other examples of spacing as an indication of word division (with the possible exception of Dunadd) and die reading thus engendered at Pool does not inspire confidence that this is the explanation here either. Reading vertically upwards from the margin [see fig. for key to numbering]: 1
Four strokes to dierighto f and perpendicular t o die stem. There appears to b e a substantial gap between die edge and die first stroke implying that this is a complete letter - probably S or, if read in die opposite direction, C. The strokes are more or less parallel but are unevenly spaced. The slightly larger gap between die first and second strokes might tempt one to read 1 + 3, (i.e. BV
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
461
or, in the other direction, HT), but the letters are generally very well spaced so it is more likely that 7 is a single group. 2
Five long strokes obliquely across the stem, sloping slightlyforwards- R. The fourth and fifth strokes are spaced rather more widely than the others, but the inter-letter spacing elsewhere is so clear that it is unlikely that 2 is other than a single letter.
3
Two vowel strokes perpendicularly across the stem - 0. These are not strictly parallel but converge slightly at the bottom. This letter follows hard on the heels of 2 though the marked difference in length means there is no possibility of confusion. There is a much more generous gap before the next letter.
4
Three strokes to the left of and perpendicular to the stem (T or, conversely, V).
5
A single vowel stroke perpendicularly across the stem - A. Again, there is a substantial gap before the next letter.
6
Three strokes to the left of and perpendicular to the stem (T or V). The middle stroke is slightly bent, but this is scarcely significant. The component strokes of this T are more widely spaced, less carefully parallel, and generally less neat than those of 4, The next letter follows very quickly though the marked difference in length and gradient remove any possible confusion.
7
Five long strokes obliquely across the stem, sloping forwards - R. The right ends of the strokes are more or less parallel and rather closely spaced but as they cross the stem they splay out slightly, creating the illusion of a distinction between the first pair and the rest. From the spacing it is clear, however, that these five are a single group. The stem is visible for another centimetre at least whether the inscription ever continued further is unclear.
Which gives the following reading: -sROTATR- or, if read in the opposite direction, -RVAVORc-
FORM OF SCRIPT The form of the script used is a simple type Ila, only the m-aicme letters slope and there are no forfeda. The vowels are shon lines occupying the middle two quaners of the ogham-band, if that. They are shoner
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
462
than the individual strokes of the b- and h-aicmi consonants. Thus the vowels are clearly distinguished from the m-aicme not only by gradient, but also, markedly, by length. The m-aicme strokes are even longer than the b- and h-aicme consonants put together. The spacing of the strokes and letters is worthy of comment since there is an unusually large gap between some of the letters, eg. 314 and 516 where there is sufficient to contain four or five strokes. Even the more closely packed letters, 213, 415,617, have more than a stroke's worth between them. In general terms the spacing of ogham letters appears to increase over time, perhaps under the influence of manuscript writing, as can be seen by comparing an inscription like Burrian or Lunnasting with virtually any of the classic ogham pillars of Ireland or Wales. Unambiguous spacing may have been required when ogham was adapted to a new linguistic context and used by those who had not been widely exposed to it. A similar explanation probably accounts for the pioneering use of word-division by Irish when writing Latin manuscripts.
In general terms the script of the Pool ogham may be compared most closely with that of the two graffiti oghams from Blackwaterfoot. Also similar are those of the stones from Scoonie and Brandsbutt, though neither of these has such marked differentiation in length between the various aicmi. Ackergill offers another parallel but, like St. Ninian's Isle, all aicmi bar the vowels exhibit a forward slope. It is unclear how significant is the presence or absence of a stem line. If it were deemed unimportant then comparison could be made with the stem-less knife-handle oghams from Gurness and Bac Mhic Connain, though the former has shorter vowels and the latter has no non-peipendicular strokes.
The elongated silhouette of the long, thin Pool letters is not merely a function of the carving technique, as a look at Cunningsburgh 3 will confirm. Instead it reflects a distinct ogham aesthetic, one reflected in numerous example from Co. Cork and perhaps originating there. The same stylistic is seen in Scotland at Poltalloch which exhibits similar proportions to Pool, though it has only notches for vowels and is regularly spaced. Whether this style can be attributed to specifically Corkonian influence will not be ascertained till the phenomenon is properly understood in its Irish context.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
463
Thefeaturesoutlined above, the typological simplicity and the shon line vowels, combine to suggest a date early in the sequence of Scottish oghams. The radio-carbon date in the sixth century is perhaps earlier than expected but far from unreasonably so. It is significant that, though Pool and Gurness have certain attributes in common, each of the Orkney oghams is typological distinct from all the others, which suggests a range of dates and circumstances for their carving.
INTERPRETATION OF TEXT Given the form and context of the Pool ogham we might expect is to be written, either in Irish, like Buckquoy, or Picitsh (Brittonic), like Burrian. Both these possibilities must be utterly exhausted before 'non-Indo-European Pictishr is invoked. Further, analogy suggests the text is most likely to consist of a personal name, though Buckquoy demonstrates that other possibilities cannot be ruled out. If the text is complete then it is difficult to suggest a possible interpretation, of either SROTATR or RVAVORC. In the case of the former, Pictish may be ruled out, since IE initial sr~ becameyr- or str- in Brittonic [Jackson 1953:541 §128]. Initial sr- is possible in Goidelic, though the rest of the word would be hard to explain in this case [Thurneysen 1946:131 §216]. The final -r is reminiscent of the final ~r of Norse personal names, but such an explanation may be ruled out on the grounds of date. The initial RV of RVAVORC is inexplicable in any Celtic language, unless the text is fragmentary. If, as appears most likely, this is in fact the case, then we have more flexibility, perhaps too much flexibility given the apparent lack of any indication of word-division. If the spacing between letters is taken as significant, then the readings would be: -SRO TA TR- and -RV AV ORC- .
Word-final -sro is impossible in Celtic because already in Common Celtic internal -jr- had become -rr[Jackson loccit. and Thurneysen loc.cit.]. To explain the sequence as Celtic it would be necessary to insert a word boundary between -s and ro- which rather negates the purpose of using spacing as an indication of word division. If one were to ignore spacing and divide the sequence as -s rotatr-, or rather s ro-t-atr~, one could take it as Old Irish with ro infixing the second person singular Class A infixed
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
464
pronoun to a verb beginning atr-. The Burrian text shows that verbs are not unknown in ogham inscriptions, but, though such an explanation of Pool is possible, it seems unlikely.
Abandoning altogether that line of inquiry, Macalister noted that the sequence '(vowel) + T/D + vowel + T/D' occurs, sometimes with doubling of the consonants, on a number of the Pictish inscriptions, namely Ackergill, Brandsbutt, St. Vigeans, and now Pool. He thought it represented a 'non-Indo-European Pictish' genitival ending, but few would agree with this identification, and his observation may or may not be significant. Alternatively, Pool's -TATR- might be comparable with #DATTRR# on the Bressay stone, Shetland (though the later is probably more correctly interpreted as ON 'daughter'), or the -TETR- in Ackergill's NEH7E7REB, if the current explanation of the latter is rejected. But this and all the other parallels, none of which are compelling, may be purely coincidental.
Returning to -RV AV ORC-, AV is reminiscent of Irish avi 'grandson*, so frequent in Irish ogham inscriptions, but this must be discounted since there is definitely no I following the V. The transliteration of the third character of the b~aimce depends on the date and language of the text (see Introduction). If Irish, HI might be more appropriate. If Pictish, we would expect /w/, but since Pool is such an early example, the orthography may not yet have 'settled down* so even in a Pictish context /f/, though less likely, cannot be ruled out. RF could be the end of a Brittonic word such as sarff 'serpent' ( < L. serpens), but this would leave an inexplicable word qf. If one were to abandon the spacing, then the somewhat Brittonic-looking -RFAF ORC- is possible. Thefirstperson singular ending of indicative verbs is -fl/\ so -frqf could be taken as pan of a verb with stem ending -if. The ending -qf also does service in the first person singular form of conjugated prepositions, and in the superlative form of adjectives. In this connection we should perhaps compare Birsay l's -ONNORRAFFRR-. Such an explanation of the Pool text leaves us with the tantalizing syllable Ore-.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
465
Old Irish ore is cited in DIL as an archaic or poetic word for a 'young pig' which is found occasionally as a personal name or sobriquet [DIL s.v. with examples, also Watson 1922:29 n.2]. The word does not survive in Welsh, which employs a different fpigr word, but that is not to say it did not exist in Pictish. The Ore of the Pool text may be complete, a male personal name, but if the text is fragmentary then a more interesting possibility is opened up. Pool is, of course, one of the Orcades, a name first attested in the first century AD. The geographical name (and that of Orcas, mentioned by Diodonis as one of the three chief capes of Britain, drawing on a source perhaps stretching back to the account 4th century BC journey of Pytheas), is derived from a tribal name Orcoi 'the young boars' [Watson 1926:28-9, 62-3]. Even today the Modern Gaelic Arcaibh < i nOrcaibh 'among the Ores', retains what Watson described as 'a tribal suggestion'. This is apparent in early medieval references to indse Ore 'the isles of the Ores' and, for instance, to fecht Ore 'an expedition against the Ores' [AU 579] and helium for Orcaib 'a battle against the Ores' [AU 708].
The plethora of possible interpretations of the Pool text underlines how little importance can be attached to any one in particular, but the possibility that it may contain the name of the place where, or of the people among whom, it was carved, is intriguing. Certainly, it would be premature to dismiss it as entirely incomprehensible or pre-Indo-European.
DISCUSSION The Pool ogham was carved cursively but not carelessly. Perhapsfromthis we can infer confidence in the use of ogham on the part of the person who carved it. The clear differentiation between different groups contributes to the easy legibility of the text. Though toofragmentaryfor a convincing interpretation there is no prima facie case for rejecting the Celticity of the language used. Whatever the languages, the use of the ogham script at Pool implies influence from Ireland, though not necessarily direct. Hunter makes no mention of any other potentially Irish features at the site. Pool is most naturally classified with the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
POOL
466
'graffiti* oghamsfromBirsayand Blackwaterfoot, with which it shares simplicity, informality, and scale. But its real significance lies in having been found in an incontrovertibly dated context and an early one at that. Single-handedly Pool stretches the sequence of ogham in Scotland back to the period then ogham stones were still being erected in Ireland.
BIBLIOGRAPHY
OS Record card - HY 63 NW 17
n.a. Archaeology Extra 1985:2 ill.; Hunter 1990:185-6.
Site:
Lamb 1980:18 No.84; Hunter 1985; n.a. Archaeology Extra 1-3 (1984-6); Hunter 1990.
Pin:
Holder 1990:72-3 ill. (see Appendix - Dubia)
PREVIOUS READINGS
None
POOL - Location map [after Hunter 1990:176 ill. 10.1]
POOL - Symbol-inscribed stone, J. R. F. Burt [Nicoll 1995:179 fig.31]
POOL - Ogham-inscribed slab (detail) [photograph supplied by J.R, Hunter, Dept of Archaeological Sciences, University of Bradford]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
467
ST. N I N I A N ' S
ISLE
DISCOVERY St. Ninian's Isle is a small tidal islet off the west coast of the southern promontory of the mainland of Shetland, in the parish of Dunrossness (NGR HU 368 208) [see map]. Following a careful search in July 1876 this ogham-inscribed slab was found by the Shetland antiquarian Gilbert Goudie 'imbedded in the sand' [Goudie 1878]. He presented it to the National Antiquities Museum in 1877 (NMS Cat.No. IB 112) [n.a. 1878]. Goudie claimed to have found two other fragments with ogham inscriptions in the same place but failed to recover them at the time and they subsequently disappeared [1878]. There is no reason to doubt Goudie's identification of the two ogham fragments and, as a known and respected local antiquary, his testimony may be deemed reliable (Goudie was a fellow of the Society of Antiquaries of Scotland and subsequently its treasurer, his likeness features in the stained-glass window made to commemorate the opening of the National Museum of Antiquities in Edinburgh in 1891 [Clarke 1990:206 ill.26]). Unfortunately Goudie gives no indication of the size or nature of these two lost fragments. Small reports that 'diligent search in modern times' has failed to recover the missing oghams [1973b:4].
SITE Today, St. Ninian's Isle is linked to the mainland by a massive natural causeway of sand and is accessible at virtually all times. In the Iron Age St. Ninian's Isle may have been more truly an island, accessible, perhaps, only at low tide [Small 1973b: 1-2]. The resemblance between this site and the Brough of Birsay in Orkney is indeed striking, and has been commented on by several writers. The similarity extends even to the extent that the chapel on each is on the very edge of the near side of the island, to the right as one steps off the tombolo. The structure now visible on St. Ninian's Isle is the remains of a twelfth-century church, enlarged in thirteenth century, and abandoned and probably demolished ('owing to the barbarous cupidity of a former owner of the island' [Brash 1879:364]) sometime after 1744 [Goudie 1904:32]. The ruins were completely covered by windblown sand and lost until the 1955-9 archaeological excavations of
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
SJ # N I N I A N ' S ISLE
468
Prof. A. C. O'Dell of the University of Aberdeen [O'Dell et al. 1959, I960]. The burial ground continued in use into the nineteenth century; writing in 1879, Brash noted that it had ceased to be used only within memory of the present generation [loc. cit.J. The site is now in State Guardianship. The island was known locally as Ringan, a form of Ninian [Brash 1879:364]. Watson has shown, however, that the dedication to St. Ninian, here as elsewhere, cannot be ancient because the form of the name reflects a late introduction from Latin rather than a continuous indigenous tradition in the vernacular [1926:296, see also Small 1973b:2].
A full discussion of the St. Ninian's Isle site is outwith the scope of the current essay, but it is clear that there was an important ecclesiastical foundation there in the pre-Norse period [for further details see Wainwright 1962b, Small et al. 1973]. O'Dell's excavations revealed not only the medieval church, but also what he claimed was the remains of an earlier structure underneath, and in alignment with it a number of long cist graves of the seventh or eighth centuries, also an earthen rampan surrounding it all, and an important collection of Early Christian carved stones. Inadequate recording methods and the death of the excavator before compilation of an excavation repon, have seriously hampered interpretation of the site [Small 1973b:4-5; Thomas 1983:285]. The precise nature of the St. Ninian's Isle site in the pre-Norse period has been disputed. There is evidence for a 'multi-phase Iron-Age complex of both domestic and sepulchral remains' [Small 1973b:7], but the nature for the Pictish occupation of the site is disputed. Radford thought the structures monastic, Thomas claimed them as a 'developed cemetery' of the late Pictish period [ 1971, 1973:11-4]. Morris has emphasised the limitations of the traditional 'monastic' model [eg. 1990:17] and suggested instead that St. Ninian's Isle may have been an eremetic site serving the social, i.e. pastoral, needs of the immediate area. Whether a minster or a monastery, the artefactual evidence indicates a wealthy foundation, not the retreat of hermits seeking a desen in the ocean.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ST. NINIAN'S ISLE
469
The most spectacular discovery at the site was made in 1958 by Shedand schoolboy Douglas Coutts on his first day as a volunteer on the dig. In a small pit beneath the nave of the medieval church, under a broken cross-incised slab, he found a larch box containing the jaw of a porpoise and the finest extant hoard of Pictish silver [O'Dell et al. 1959; Small et al. 1973; Close-Brooks 1981; Youngs 1989 Nos.97-107; Laing 1993a Nos. 26-7, 38, 55-62, 85, 223-37, pis. 16-24]. The initial suggestion that the twenty-eight gilded silver objects represented a hoard of ecclesiastical silver [McRoberts 1963, 1965] has been conclusively rejected and it is now convincingly interpreted as a collection of secular valuables deposited for safekeeping and never reclaimed [Small et al. 1973]. On the assumption that the motive for the deposition was the imminent threat of Viking attack, the hoard has, from the outset, been dated to c.800. The work is of high quality and the 'distinctive style of much of the silver1 has been 'central to the identification of a larger corpus of Pictish metalwork' [Youngs 1989:108]. The earliest piece is a hanging bowl of the seventh century, most of the odier pieces have been dated plausibly to the eighdi century, and die deposition of the hoard to the decade or two either side of 800 AD. While the direat which prompted the deposition is perhaps most likely to have come from the Norse, it need not. The Nome's Law hoard of Pictish silver has been recently re-dated to the second half of die seventh century [Graham-Campbell 1991].
The diree cone-shaped objects, the sword pommel and two sword chapes (one of them bearing a roman alphabet inscription), the silver spoon and related implement, the twelve brooches, and eight silver bowls, belonged to someone of considerable wealth, their combined weight is almost 2 kg of silver. The extant sculpture at die site is further indication of the considerable resources of its patrons. There are twenty-two sculptured stones [Thomas 1973:33-9], particularly notable are the remains of composite ('corner-post1) shrines of the late seventh or early eighth century, including one finely-decorated double shrine [Thomas 1983]. Thomas has suggested this kind of monument is of ultimately Northumbrian origin [1971:150-59], an interpretation disputed by Isabel Henderson who, while pointing out links with Iona and Ireland [cf. Herity 1993:195], stresses above all the possibility that the form has indigenous roots [1994:88-90]. She
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ST. NINIAN'S ISLE
470
develops Patrick Ashmore's suggestion [1980:353] that the comer-post shrine may be an native development from the pillars and slabs arranged round certain Pictish low kerb burial cairns [note also Stevenson's later dating for the shrine 1981:287].
The Nonhumbrian affiliations of both the metal-work and sculpture from St. Ninian's Isle were perhaps over-emphasized in the initial analysis. More recent work has stressed instead the extent of Irish cultural influence in pre-Norse Shetland, especially at the major ecclesiastical site of Papil, on Burra [Laing 1993:34]. Little is known of the pre-Norse inhabitants of Shetland. That the incoming Norse thought the locals were Picts is revealed by place-names containing the Norse Peti 'Picts', for instance, Pettaster, Pettadale, Pettifirth [Turner 1991:46]. The Class I Pictish symbol stone from Sandness. and fiinher fragmentsfromelsewhere in Shetland, [ECMS 3-5] show the Shetlanders panicipating in wider Pictish culture. Recent writers have been tending to stress cultural integration between incoming Norse settlers and local Celts throughout Atlantic Scotland [Ritchie 1977; Cox 1991; Cant 1984:174]. Bigelow has pointed out how little is known of the date and nature of the Shetland landnám, or first Norse settlement [1992:9]. There is no reason to believe, however, that the Norse settlement would necessarily have brought an end to the Christian practice of the native inhabitants. The Life of St. Findan shows a Pictish bishop still active in Orkney in the 840s [Thomson 1986]. As Stevenson points out, there is a break in the sculptural tradition of Orkney at the end of the eighth century, but in Shetland, indigenous Christians continued to erect monuments into the ninth and tenth centuries [1981]. It is notable that pre-Norse ecclesiastical sites in Shetland 'featured prominently in the organization of the medieval Norwegian church in the islands' [Cant 1984:175]. That St. Ninian's Isle was favoured for the burial of rich individuals of Norse heritage is shown by the presence there of a 'hogback' monument of eleventh century date [Lang 1974:231].
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
§ 7 NINIAN'S ISLE
471
ROMAN ALPHABET INSCRIPTION Once again, a full treatment of this object is outwith the scope of the present discussion. Briefly, the silver-gilt, U-shaped item, 80mm x 80mm, has been identified as the chape from a sword scabbard [see fig.]. On an historical grounds it has been dated to the late eighth century and its decoration shown to have links with the style of the Pictish brooches in the hoard. Along with the rest of the St. Ninian's Isle hoard it is now on display in the NMS (cat. no. FC 282). A full bibliography is given by Okasha [1985:57-9], the script is discussed in detail by Brown [1959 (1993)], and the text by Jackson [1973b], the chape is well-illustrated in Youngs 1989 pi. 103. On one side, in uncial letters, is the text INNOMINEDS, interpreted as in nomine D[ei] S[ummi]. On the other side, in minuscule, is RESADFILISPUSSCIO Resad fili Spusscio 'Resad son of Spusscio' [Jackson], or res ad Fili Sp[irit]us S[an]c[t]io% 'property of the Holy Spirit* [Youngs 1989:110], The former seems the more likely, though the name Spusscio is baffling. Jackson was equally dismissive of the first name, he did not draw attention to the sequence RASSUDD which appears on the Altyre ogham, nor attempt to link either with the well-known Brittonic name ReslRis (Rhys). Though the chape is demonstrably Pictish, it may not have been made locally. A luxury item such as this may well have come from further afield. Still it suggests an interest in literacy on the pan of the secular nobility appropriate in the milieu of such an important ecclesiastical site as St. Ninian's Isle.
DESCRIPTION OF OGHAM-INSCRIBED OBJECT A carefully-prepared, thin, oblong slab, undecoiated except for an ogham inscription carved up the entire length of one narrow edge [see fig.]. Stone:
Lighdy coloured sandstone
Dimensions:
760mm long x 265mm broad x 45mm thick
Condition:
The two broad faces are severely laminated and large chunks of the surface have been lost. Both the upper and lower edges are fractured. A portion of unspecified length is missingfromthe beginning of the edge, but the inscription
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ST. NINIAN'S ISLE
472
is otherwise well-preserved. If the broad faces of this slab ever bore carving it has been lost. There is absolutely no trace of any decoration, but both surfaces are flaky, and it may be that the entire original surface has laminated away. The ogham is a well-carved, formal, monumental inscription, not a casual graffito. Its position on the narrow edge of the slab suggests the main surface or surfaces were given over to other carving, as at Bressay and Altyre.
If the lettering were the only carving one might have expected, instead, an
arrangement across the broad face, as at Lunnasting.
DESCRIPTION OF OGHAM INSCRIPTION The ogham is arranged along the entire length of the narrow edge of the stone. The inscription appears to begin in mid-letter and it is clear that a portion, possibly substantial, has been lost from the beginning. If the stone had stood upright it may have fractured near the base when it fell over leaving an inscribed stump at the top of the ponion embedded in the ground. It is possible that, one or both of the fragments found by Goudie may have been the missing pan of this slab, but since he does not suggest it himself, this is perhaps unlikely. The upper edge of the slab is also fractured, but there is 30mm of stem clear after the final letter, which suggests that no lettering has been lost. Most commentators have adopted Ferguson's view that the beginning was imperfect but the ending complete [1887:134].
The ogham has been carved using the usual Pictish pock-and-smooth technique. Little attention has been given to smoothing and the pocks are still clearly visible. The strokes are well-defined but not deep, they have rounded ends and a curved cross-section. The stem-line, which runs the entire length of the extant portion of stone, has been re-cut on top of the letter-strokes. The inscription is very neat and regular, the letters are clearly differentiated by size, slope, and spacing, hence the almost total consensus regarding the reading. The direction of reading is indicated by the slope of the b- and h-aicme consonants, which is consistent. The stem is visible at the fracture and continues for about 25mm before the first stroke [see
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ST. NINIAN'S ISLE
473
fig. for key to numbering]: 1
A single oblique stroke to the right of the stem, sloping forward. Macalister claims to have seen an additional stroke before this, which may indeed be present. The wear on the stone is such, however, that it is difficult to make out whether the clear stroke was a singleton, or the final stroke of a longer letter. If there were any preceding strokes they can only have been of the baicme - B, or any subsequent member of the b-aicme (L, V, S, or N).
2
Four long strokes perpendicularly across the stem - E.
3
Four short oblique strokes to the right of the stem, sloping forward - S.
4
A single long oblique stroke across the stem, sloping forward - M.
5
Four long strokes perpendicularly across the stem - E.
6
Five short oblique strokes to the left of the stem, sloping forward - Q.
7
As above - Q.
8
Five short oblique strokes to the right of the stem, sloping forward - N.
9
A single long stroke, perpendicularly across the stem - A.
10
Five short oblique strokes to the right of the stem, sloping forward - N.
11
A single long stroke, perpendicularly across the stem - A.
12
A single long oblique stroke across the stem, sloping forward - M.
13
As above - M. Southesk, Rhjs, and Allen read 12 and 13 together as G, but Padel correctly points out that the careful spacing between the two strokes means they must be read as M. There is a small, natural flaw in the stone to the left of the stem at the point where it joins the M-stroke. Macalister was mislead in thinking it was a corrected superfluous A [Macalister 1940:206].
14
Two long strokes perpendicularly across the stem - O.
15
Three short oblique strokes to the right of the stem, slopipg forward - V.
16
As above - V.
17
Four long strokes perpendicularly across the stem - E.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
18
ST. NINIAN'S ISLE
474
Four long oblique strokes across the stem, sloping forward - S'. See below for discussion of transliteration of this character. The stem-line is visible for a further 30mm before breaking off in thefractureat the end of the stone, but, apart from a few minor, natural cracks, there are no further strokes. I can only concur with Padel that the inscription 'gives the impression of stopping at this point'.
Which gives the reading: -b+]ESMEQQNANNAMMOVVE&
FORM OF SCRIPT The St. Ninian's Isle ogham is written on a stem-line, with long, straight vowel-strokes, and consistentlysloping b- and h-aicme consonants. It entirely lacks 'developed' features such as forfeda, bind-strokes, or word-division. As such it is a simple Type Ha ogham. It differs from other Type Ila oghams primarily in the get xous spacing between its letters, and in having consistently sloping b- and h-aicme consonants. While more like certain He oghams, in spacing, proportions and general aspect, it is entirely lacking in the 'developed' features which are their defining characteristic. One distinctive attribute of the St. Ninian's Isle ogham is that both a- and m-aicmi strokes occupy only the middle three-quarters of the ogham-band. It is unusual to have the m-aicme strokes occupy the same band-width as the vowels, unless they both stretch right across. Brandsbutt is one inscription which does, but it differs from St Ninian's Isle in having only the slightest of slope on its b- and h-aicme letters and in having no clear spacing between letters. Scoonie is perhaps closer, but again, it lacks a consistent slope on its b- and h-aicme consonants. Inchyra B is also similar in general appearance (though again, without the consistent slope), and perhaps significantly, only the final part of the text has bound letters, suggesting that they were not always used even when available. For all its simplicity, however, it would be hard to believe the St. Ninian's Isle ogham is particularly early. It shares with Lunnasting and Cunningsburgh a certain squatness which suggests they are not as far apan as their initial appearance might suggest. Perhaps St. Ninian's lack of forfeda is due simply to the coincidence that the text didn't happen to require any. The complexity of the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ST. NINIAN'S ISLE
475
St. Ninian's Isle ogham's palaeographical affinities underlines the difficulty of attempting to construct a comprehensive script typology and, a fortiori, of trying to use this for purposes of absolute dating.
The inscription is extremely clear, the only doubt over the transcription concerns the correct transliteration of characters 15116 and 18, the third and fourth members of the b-aicme, respectively. The correct transliteration of the third member of the b-aicme is discussed in the Introduciton, where it is suggested that in a Gaelic context it's value was HI after seventh century, and in a Pictish context /w/ regardless of date. The character appears doubled at Lunnasting also. In both cases it may be a cipher for Pictish manuscript spellings uu for /wo/, e.g. Uurguist.
Whatever its original value, the fourth member of the b-aicme had become linguistically redundant by the time of the earliest ogham pillars and does not appear in any Irish inscription. It features in an ogham alphabet carved on a twelfth-century stone from the Isle of Man [Kermode 1907:72-4, 100-02], but no indication is there given of its phonetic value. St. Ninian's Isle is thus the only epigraphic example of the character in practical use. Patrick Sims-Williams has discussed possible values for this letter in almost exhaustive detail, but concluded that the original value, perhaps /'s/, perhaps /sw/, cannot be recovered [1993:151-62]. Since this value was forgotten at such an early stage, however, it is probably irrelevant to the St. Ninian's Isle text. In the bardic ogham tradition, this character was called s(t)raif and its standard value was /st/, 'where s stands before d, it is straiph that is to be written there' [Calder 1917:1.443-4, quoted by Sims-Williams 1993:161]. Sims-Williams explains how this would have arisen through a reinterpretation of the old letter-name, an equation with Latin Z, and finally its (functionally redundant) assignment to the sound /st/ [1993:161, cf. McManus 1991:38; Sims-Williams 1992:35]. The earliest manuscript evidence for this is the twelfth century Vatican manuscript Reg.lat. 1308 [Sims-Williams 1993:151]. Sims-Williams over-looks the St. Ninian's Isle ogham in claiming there are no epigraphic instances of the character apart from the late and linguistically unhelpful example from Man. While the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ST. NINIAN'S ISLE
476
Shetland inscription is impossible to date closely, it can scarcely be as late as the twelfth century, and thus must contain what is the earliest extant example of this character. As discussed below, the value /st/ is the most plausible explanation of the character in context. This is important evidence for continuing contact between oghamists in Ireland and the Northern Isles, and underlines the essential unity of the tradition throughout the period of the script's practical use.
INTERPRETATION OF TEXT The letters 4-7 are immediately recognizable as a form of Irish MAQ(Q)I, 'son', which occurs in mainland Scotland as MAQQ, or, as here, MEQQ [see entry on Golspie for a fuller discussion]. This allows die preceding and succeeding letters to be interpreted as personal names: -b*Jes Meqq NannammowesL. Padel drew a parallel with Altyre, pointing out that there too, the 'son' word came 'very near to the start of the legend, and is followed by a very long series of letters' [1972:140], The Altyre text, however, is intact, whereas the St. Ninian's Isle ogham is clearly missing strokes, if not several letters, from its beginning. The extant segment of the first name is too short for analysis beyond noting that -nes, and -les are perfectly possible final syllables for personal names in both Goidelic and Brittonic languages. To take just one example, the element glas meaning 'blue, green, grey' is found in both Irish and Welsh compound personal names in first and second position, and is sometimes spelled gles in Brittonic [Uhlich 1993:263]. Morcleis for Morglas appears in the Book of Llandaff [Davies 1979:s.n.], and Glessic, equivalent of Irish Glaissiuc is the name of a hound in Culhwch and Olwen [Uhlich loc. cit.], cf. Llandaff rs Conglas. Note also Bressay's Bennises. Of course, Nes was an Irish male personal name in its own right.
The name of the father of '-b*]es\ NannammowesL is probably to be transcribed /nanamowest/ or /nanamogwest/, the double M perhaps intended to indicate lack of lenition. At four syllables it is perhaps rather long for a single name and further segmentation may be required. The final syllable /-west/, /gwest/, is perhaps the easiest to explain, it appears to be a form of Pictish -(g)uist ( < Celtic gustus
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ST. NINIAN'S ISLE
477
'choice'), which appears in such names as Pictish Onuist ( < *Oinogustus), cf. Irish Oengus; Pictish Uurguist ( < Worgustos), cf. Irish Forggus, Welsh Gorwsu Gwrwst (<*Gworwst)
[Jackson 1955:163].
Celtic -st became s in Gaelic, in Brittonic is either remained st or gave s, in Pictish it appears never to have become s [Jackson 1955:165]. The name appears, therefore to be Pictish rather than Gaelic, especially since it appears to lack the expected Irish genitive ending -guso; Pictish -(g)uist appears in both nominative and genitive position [ibid.] and, in any case, Koch has argued that Pictish lacked declension [1983:21420].
Working backwards, if the preceding 0 represents a composition vowel, then it has clear implications for dating. Such syllables were lost in Brittonic in the first half of the sixth century 'and in Gaelic slightly later' [Jackson 1955:166]. Bede'sMeilochon [tf£111.4], (<Magloamos, laterMailgun, Maelgwn), written at the beginning of the eighth century of a Pict living in the third quarter of the sixth, implies that syncope was later in Pictish, perhaps in the second half of the sixth century [Jackson 1955:166]. This still seems rather early for a monument such as the St. Ninian's Isle slab.
What then of /nanamo-/ ? A favourite Gaulish onomastic element was narno-* which Evans interprets as deriving from perhaps more than one root, (1) *nem- 'allot', 'take', (2) *nem~ 'bend', 'glade', 'sacred grove', cf. Ir. neimed 'sanctuary', MW. nyuev, possibly W. nqf 'lord' [Evans 1967:234-5]. Theformmost commonly found in composition has the old Celtic t-suffix, namanto-, Evans, refutes, however, the argument that namo- is an abbreviated form, asserting instead that it is an independent o-stem [ibid.]. Perhaps St. Ninian's Isle's Nanamo- is a reduplicated form of Celtic Namo-. Evans makes the link with WCB nam 'defect, blemish, fault, vice' [Evans 1967:369], but rejects the comparison with MIr. náma(e) 'enemy', preferring O'Brien's explanation of am 'love' + ne-y negative prefix [Evans 1967:234]. The latter word is unlikely to be relevant here since it is a t-stem, and St. Ninian's Isle lacks a dental. If the ogham NA- could be a negative prefix (? e > a, vowel harmony), might the St. Ninian's Isle name mean
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ST. NINIAN'S ISLE
478
'unblemished choice1 ? The name Náeingns appears in a Ciarraige genealogy preserved in the Books of Ballymote and Lecan [CGH 160 a 26], but this appears to be for Nóemgus, 'sacred choice', fromnoib (var. naeb) - 'holy, sacred' [DZL]. On another tack altogether, if the inscription were to be transliterated as if Gaelic, then we would have at the end MoffesL O'Brien lists Mofemis (var. Mqfebis) from the prehistoric section of a Múscraige genealogy [CGH 147 a 17, ab 56; 148 b 38], and Mofeda (Mafeta) from another Múscraige pedigree [CGHLL 324 a 41], but, like Rh?s's suggested Mobiu [1892], these seem implausible on both linguistic and historical grounds, and moreover leave NANA- unexplained.
DISCUSSION Neither palaeography, archaeology, or an history provides any grounds for the dating of the St. Ninian's Isle ogham. The ecclesiastical site there may be as old as the seventh or eighth century and enjoyed considerable prestige throughout the Pictish and into the Norse periods. If, however, the text is correctly interpreted as preserving a composition vowel in the second name, then the ogham can scarcely be later than the seventh century (unless we posit a remarkably strong and long-lived early orthographical tradition). This is remarkably early. The second name appears to be Brittonic (i.e. Pictish) rather than Gaelic, though the formula, 'X MAQQI Y \ has been borrowed from the Irish along with the script. Such early Irish associations in Pictish Shetland are perfectly in keeping with the evidence of Irish influence at the other major early ecclesiastical site in Shetland, at Papil on Burra. The inscription on the silver chape suggests that the roman alphabet was also available to the patrons of St. Ninian's Isle making the choice of ogham for the only monumental inscription(s) at the site all the more interesting.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
ST. NINIAN'S ISLE
479
BIBLIOGRAPHY
Goudie 1878, 1904:34-35 ii andiii; Brash 1879:364-5, pi. XLIX; Southesk 1884:205; Ferguson 1887:134; Rh^s 1892:296; ECMS 18, fig. 13; Macalister 1940:205-6, fig. lid; RCAHMS 1946:44-5 No.1185fig.520; Moar& Stewart 1954:94; Padel 1972:139-140; Thomas 1973:38.
PREVIOUS READINGS
Goudie, Brash, Southesk
-ESMEQQNANAGOFFESI
Ferguson
-ESMEQQNAN AMMOFFESI
Rhjs
bESMEQQNANAmmOVVEF
Nicholson
LESMEQQNANAMMOVVESI
ECMS
BESMEQQNANAMMOVVEF
Macalister
LJESMEQQNANAMMOVVESZ
Padel
-BJESMEQQNANAMMOVVEZ
v
1
^u ^f •r "^
SRUTLAND St Ni 11 ian's Isle -
< Lerwick
f Si Ninian's Isle—»£«
ST N I N I A N S ISLE
$fs ic-l\
K/ II 11
tf>
«~J0+*&
10 km ,
J
^r
s
1
•
II ' ST. NINIAN'S ISLE - Location Map [Close-Brooks 1981:3]
' oT -JS
QW c
to
,
r
~
S=*
|
\
1
\ ^v -c * ^ 1 _ ^ %
i/
I
/^5- *—
•
/
\\
ch u r c h - § 3 ^ s ^ f
^ip 1 km
1 1
}
-^íA
1
W
/I
P)
1
v#:i
./ •'.
(key to numbering) ST. NINIAN'S ISLE - Ogham-inscribed slab [© Trustees of the National Museums of Scotland, Reg. IB 1121
ST. NINIAN'S ISLE - Inscribed sword chape (front and back) [Close-Brooks 1981 fig. 15]
ST. NINIAN'S ISLE - Inscribed sword chape. Detail of inscription [Jackson 1973b fig.29]
ST. NINIAN'S ISLE - Form of script used [Brown 1993:246]
480
SCOONIE
DISCOVERY This Class 0 symbol-incised cross-slab was discovered in the former churchyard of Scoonie» Fife (NGR NO 3840 0167). It was taken first to the church of nearby Leven, before being gifted to the National Antiquities Museum by the Heritors and Kirk Session of the parish of Scoonie in April 1866 (NMS Cat. NaiB110)fn.a. 1866].
SITE The old church at Scoonie [RCHAMS 1933:267 No.491] had been in use for services until 1775, when it was taken down and a replacement built at Leven [Millar 1895:47-8; RCAHMS 267]. The cross-slab is the only archaeological evidence for early medieval activity at the site. The earliest documentary reference, dated 1055x1059, is among the noútía in the Register of the Priory of St. Andrews: the granting of the church of Sconyn to the Culdees of Loch Leven [Lawrie 1905:7]. According to the Register of the Priory of St. Andrews and the Pontifical Offices of St. Andrews, the church of St. Memma the Virgin at Scoonie was re-dedicated by Bishop David of Birnam in May 1243 [Anderson 1922.ii,520). Anderson equated 'Memma' with Monenna, the supposed sister of St. Patrick, who, according to legend, founded seven churches in Scotland [for possible Monenna dedications see Watson 1926:190, 342,366]. Evidence for a Scottish cult of Monenna is suspect and late, however, and, the dedication here may well have supplanted an earlier one (see discussion below). The name Sconyn, like its more famous Perthshire counterpart Scone, is of dubious origin. Popular dictionaries of Scottish place-names link it with Gaelic sgonn, flumpf or 'mass* [e.g. Wood 1989:82]. This word does not appear in Irish and may well be a Pictish loan-word.
DESCRIPTION OF OBJECT A Class II cross-slab, sculptured on both sides. On the front a badly weathered cross in relief, on the
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
481
reverse [see fig.] an incised 'beast' symbol above a hunting scene. An ogham inscription runs up the right hand side of the reverse. Stone:
Sandstone
Dimensions:
1.06m x 0.7m and 0.1 m thick
Condition:
Portions missing from the top and probably also from the bottom. Front severely weathered, reverse better preserved. Ogham inscription clear and possibly complete.
The top of the slab has sheered off and a portion of indeterminate size has been lost. Theformof the cross should provide the best indication of how much has been lost, but unfortunately it is extremely badly weathered. We would normally expect another symbol above the beast to make the customary pair, and this would require more than a few centimetres of additional height. The paired syntax is elsewhere so dominant that we may suspect the loss of a second symbol. Allen lists a handful of Class II examples of lone symbols, but in almost every instance, including Scoonie, the slab is broken, the missing portion very probably having contained the second symbol (an exception is Kirriemuir 2 FOR which, like Scoonie, has a single symbol above a hunt scene). The slab's current mounting obscures the bottom of the carving which suggests that a portion has also been lost from the foot of the slab. If the slab was originally sunk in the ground or a base, a considerable depth would have been left blank. Since the slab is on public display it has been impossible to verify whether or not the bottom is intact.
DESCRIPTION OF CARVING - DECORATION According to Allen the cross on the front has an elongated shaft and round hollow angles, the most common cross-form in Pictish sculpture (his form 102a). In pan III of ECMS he suggests the cross has a ring (though his '?' indicates his doubt), but contradicts himself in pan II where he lists Scoonie along with otherring-lesscross-slabs [51]. Unfortunately the cross-slab's current position against a wall in the NMS makes it difficult to check. Allen says that the details indicated in Stuart's drawing 'have entirely disappeared'. The period between Stuart's drawing and die slab's entry into the protection of the museum
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
482
is too shon for the degree of weathering implied, therefore Stuart's details must be considered suspect.
The reverse face of the slab is well preserved. The carving consists of a beast symbol occupying the whole width of the top of the extant ponion, and below that a hunt scene comprising three horsemen, accompanied by two dogs, pursuing a stag. In later medieval an the hunt was endowed with a Christian symbolism, the stag representing either the persecuted Christ or the Christian soul pursued by devils. The hunting theme was particularly popular in Pictish an where it may well have had a powerful secular significance as well as, or instead, of its canonical Christian symbolism. Alcock has stressed the social and cultural role of hunting as an expression of heroic values (its relation to warfare), wealth (the resources required to train hunting animals and maintain equipment), and status (he quotes Nelson on the hunt in Carolingian royal ritual and ceremonial [Nelson 1987:166-72]). Alcock considers the depiction of the hunt in Pictish an as 'a statement of social superiority, and a reminder of the rights and exactions which accompanied it* [1993:232].
Pictish representations of the hunt an are all multiforms of a single type, the 'stereotyped hunting scene formula', which may have been an indigenous invention or an adaptation of imponed models [Henderson 1967:137]. Henderson nominates Hilton of Cadboll ROS as perhaps the finest example of this scene which involves mounted hunters and a 'leaping hind being attacked simultaneously on the shoulder and flank by two racing hounds' [ibid.]. The Scoonie scene is considerably less ambitious or accomplished. Some detail has been lost through weathering (the spear in the stag's flank is barely visible), but even so the figures are somewhat crude. The smaller of the two dogs, in particular, is rather poor. The diagonal alignment which is habitually used to suggest movement is lost at Scoonie. Instead the figures are arranged horizontally, from left to right, and the effect is rather static. Stevenson suggested the rudimentary character of the carving might reflect an early date (early to mid-eighth century), though conceded inferior workmanship might be the better explanation [1955:114]. The Scoonie hunt scene is paralleled on crossslabs at nearby Largo, in Forfarshire at Inchbrayock 3, St. Vigeans la, Kirriemuir 2, and Aberlemno 3,
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
483
and, in Moray, at Burghead and Elgin. The triple horseman motif, without stag/hind, is found, for instance at Rossieand at Meigle 1, 4, and 11 PER.
There can be no doubt that the ogham is contemporary with the rest of the carving. One of the more pleasingfeaturesof this slab is the way in which the two are integrated, the stag stepping across the stem and interrupting the letters with its fore-leg and muzzle. It is clear that the original design allowed for the ogham: the beast symbol is further to the right, the larger hound though hard up against the preceding horseman stops well short of die line, the lowest horse is well back from the outer edge of the stone compare the tail of the upper horse which is almost touching the edge. At Latheron the letters are carved on the face of the slab, but no attempt is made to integrate them with the neighbouring designs, elsewhere (as at Brodie, Abernethy and Whiteness) the ogham is entirely removedfromthe surrounding sculpture and placed on a separate raised band. The Golspie ogham is on a corner moulding, but the fish's tail and the tip of the crescent impinge slightly on the band lessening the rigidity of the whole. The permeability of borders is a recurring feature of Pictish an. Macalister was too harsh when he said that to the engraver the 'inscription ... was a mere nuisance', fitted in round the designs which were his primary interest [1940:193]. The same incised line has been used for both. Macalister made too much of the alleged squeezing and irregularity in the setting out of letters and in artistic impression, the ogham is on a par with the rest of die scene.
DESCRIPTION OF CARVING - OGHAM INSCRIPTION The ogham inscription is carved on a stem-line which gently undulates up the reverse face of the cross-slab close to and parallel with the right-hand edge. Padel pointed out two incised lines in the form of a cross in the bottom right hand comer of this same face [141]. They are clearly artificial and are incised on the same scale as the ogham strokes and stem, and at a similar distance from the outside edge. As he says, if this is an ogham letter it is a hammer-head A, though only if one reads from left-to-right down, as at Newton. If not ogham, these lines are otherwise hard to explain. The substantial bar, on the left arm
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
484
only, mitigates against a simple incised cross, which in any case, would be rather incongruous in this position in a hunt-scene.
As currently displayed the portion which might contain further letters is
obscured. If as much of the surface is intact on the left as on the right of the face there should be traces of another two or three strokes. Allen's photograph reveals a few centimetres more than is currently visible and seems to include two oblique strokes to the left of the stem, sloping against a top-down reading. These are less clearly anificial and may be simply natural cracks in the surface above the fracture. If they are pan of a letter they would be L or a subsequent member of the b-aicme, but their slope contradicts the left-hand bar on the A. If their slope is given priority, then we might have the end of an ogham reading from the bottom up, in which case '(D + )A\ with the bar on the left rather than the standard right tip of the A. The continuation or otherwise of the ogham could be verified easily when the stone is next moved.
Since it is not known how much, if anything, is lost from the bottom of the slab we cannot begin to assess how much text might be missing from the lower ends of the two lines, though it is significant that the left hand line begins no higher than it does (it could have run the entire length of the left side). The right-hand text makes sense as it stands and there is no need to posit missing text. The lone A on the left, however, is more plausibly the remnant of something longer, and if something has been lost on the left, then perhaps also on the right, for why would the two lines of text be staggered in their layout ?
The right-hand ogham stops level with the beast's tail. The surface is clear thereafter so we can be sure that nothing has been lost from this end. The ogham has been nicely integrated into the design. The forehead of the lowest horse abuts letters 2-4, the stag's leg crosses the stem line between letters 7 and 5, its muzzle between letters 9 and 70. The strokes and the stem have a similar depth and width, the latter underlying the former. That the ogham reads from the bottom up, as expected, is confirmed by the slope of letters 2, J, and 7. All the lettering is clear, and, unusually enough, there is no doubt whatsoever over the reading, which is as follows [see fig. for key to numbering]: 1
Four long strokes perpendicularly across the stem - E. As currently mounted in the NMS only
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
485
two of these are visible, but Padel repons he 'had the wood removed, and this letter is quite certain' [141]. Previous authorities made no mention of the possibility of anything preceding this letter, Padel's verdict was 'unclear but doubtful'. 2
Two strokes to the left of the stem sloping slightly backwards - D. The second stroke, like the following three, abuts the forehead of the lowest horse and is therefore slightly shorter than its equivalents elsewhere in the text. It seems more upright than the first which is clearly sloped.
3
As above, though more markedly sloped and parallel - D. These two pairs are carefully spaced, Rhys's HT has been unduly influenced by the disparity of slope between the two strokes of 2.
4
A single long stroke perpendicularly across the stem - A.
5
Five long oblique strokes sloping forward across the stem - R.
6
As above - R. This group is at slightly less of a slope than 5, but not sufficiently to warrant the reading I. The contrast with the next letter, which is perpendicular is clear. Letters 5 and 6 are distinctly spaced.
7
Five strokes to the right of, and perpendicular to, the stem - N. The first of these protrudes slightly beyond the stem. Presumably this is just a slip. Padel suggests comparison with Golspie 7 7 and 18, but concedes there is less reason for it here. On Allen's photograph [fig.360] there is a dark patch to the left of the stem which looks like a stroke. As Padel pointed out, this is merely a shadow. These five strokes are a little less generously spaced than their predecessors, compressed to fit into the space available before the leg of the stag crosses the stem.
8
Two long strokes perpendicularly across the stem - O. Their left distal tips abut the stag's chest.
9
Five strokes to the right of the stem sloping forward markedly - N.
This area is
weathered and the lines are a little faint though not in any doubt. The strokes are slight and very closely spaced to fit into the room available, in fact the fifth stroke abuts the stag's muzzle rather than the stem.
This is the only N in the inscription to be sloped, both the others are
perpendicular. Presumably the slope enabled the carver to cram the strokes into the available space without sacrificing legibility.
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
10
SCOONIE
486
Five strokes to the right of, and perpendicular to, the stem - N. These strokes are the most generously spaced of all because there is ample room available. The stem stops very shortly after the final stroke. The surface immediately thereafter is unbroken and it is clear the inscription stopped here.
Which gives the reading: EDDARRNONN
FORM OF SCRIPT The Scoonie text is carved in a simple Type Ila ogham, with a drawn-in stem-line and long vowel-strokes occupying virtually the full ogham band. Unless the hammer-head A on the left is accepted, and it remains doubtful, there are no supplementary letters or other developed features. The spacing of strokes is fairly generous, except for 9 which is squeezed into the space before the stag's muzzle, but the letters are mostly rather close. The gap between successive letters of the same aicme is less than a stroke's worth, and between successive letters of different aicmi only marginally more than between component strokes of a single letter. The slope on m-aicme strokes is not great. Some b- and h-aicme letters are perpendicular, others - 2, 5, and 9 - are sloped (in the case of the latter perhaps due to lack of space). In palaeographical terms, the closest parallels may be drawn with the script of Brandsbutt. Ackergill is similar, but has shorter vowel-strokes. St. Ninian's Isle is more generously spaced and has b- and h-aicme letters more consistently sloped The same goes for Buckquoy which also has a hammer-head A.
INTERPRETATION OF TEXT There is no doubt at all over the reading of the text - EDDARRNONN. This has been universally, and surely correctly, interpreted as the male personal name Etharnon. The name appears twice more in the Scottish ogham corpus - in identical form at Brodie (EDDARRNONN) and slightly modified at Newton (IDDARRNNN) - and probably also on the roman alphabet inscription from Fordoun -]Idarnoin[- [Okasha 1985:51-3 and refs.]. In die manuscripts the form is Itarnan [AU 669]. tf these are all correctly
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
487
identified, then they constitute variant early attestations of a name, and thus provide a rare opportunity to explore Pictish orthography.
The triplication of N at Newton is scarcely important [see Newton entry for discussion], of more interest is the alternation of initial E (Brodie and Scoonie), against I (Newton, Fordoun, AU). Since both are attested in ogham the difference cannot be explained away as merely script-based. The difference might be dialectal, though it should be noted that the most northerly and southerly epigraphic examples agree against the ones in the middle. Date may be more relevant, since Newton and Fordoun are almost certainly both earlier monuments than Brodie and Scoonie. In later medieval tradition the name is Ethernanus [Aberdeen Breviary (Elphinstone 1509-10) 2 December].
All witnesses agree on the second vowel -0-, the Chronicon Scottorwn [Hennessy 1866] is unique in having Ituman, which could be a simple scribal error (though u is still a back vowel). For the third vowel, all the epigraphic examples have -on, against the Gaelicized -an of AU and the Aberdeen Breviary. The alternation of final -onl-an is seen elsewhere, for instance in Bede's NaitonlNdtan.
The -i- in the final
syllable at Fordoun may be an Irish genitive, especially since the word is embedded in a longer text - the segment occupies the penultimate of at least three lines of text. If the text is not in Irish, then an alternative explanation must be sought. Jackson raises the possibility that a number of apparently 'intrusive1 -is in Pictish orthography might be following Gaelic practice in using e and / to indicate die palatalization of the following consonant, giving as examples Vipoig and Tarcdn [1955:165].
Asforthe consonants, if the name is correctly analyzed (see below) the first consonant sound was, at least to begin with, /d/ < Lat. /t/. In Old Welsh orthography we would expect this to be represented by -r(cf. Nennius's Guotodiri). In Old Irish -r- could represent III or /d/, thus from AU's harnan we may take it that the name was still pronounced with a stop rather than a spirant. There was a convention available
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
488
in ogham orthography whereby non-lenition of consonants could be indicated by writing them double. This may be reflected in Brodie, Newton, and Scoonie's -DD-. The picture is complicated, however, by the fact that at some point the stop in this word became lenited; the name appears in charters of the 1140-50s as YdernimiSy and in a charter of 1260 Etheminus [Barrow 1973:186 n. 100], thereafter it is regularly some variant of Etheman, or Etharnan (where AU has Itarnan, the Annals of Tigernach has Itharnari).
In Old Welsh orthography -*/- represented a voiced spirant (cf. Guotodin), only later did it come to represent a voiced stop, thus the Fordoun form, Idarnoin, which may be dated to the eighth century on an historical grounds [Okasha 1985:53] is of interest. According to Welsh orthography of the period we would expect Idarn- to represent /ióarn-/, cf. OW. Patent /padern/, only later Padern /padern/. If the orthography of Fordoun is Gaelic, then -d- must represent a voiced spirant [Thurneysen 1946:21-24 §§2932]. The lack of unambiguous symbols for spirants in the Latin alphabet caused problems for those attempting to generate onhographies for the Insular Celtic languages. Harvey has discussed at length an initial period of experimentation and the eventual standardization off/*, dlu ch, etc. [Harvey 1989, 1990a, see also Sims-Williams 1992 ]. The doubling of R and N in the ogham examples of Ethernan is in contrast to the single consonants of die roman script examples, both epigraphic and manuscript. On Irish ogham pillars, -AN ( < -AGNI) is occasionally found, e.g. BRANAN (CIIC 88), GATTEGLAN (CIIC 239), but the most common reflex is -ANN, e.g. RONANN for manuscript Rónán (CIIC 145), SILLANN for manuscript Sílán (McManus vi), CONANN for manuscript Conán (CIIC 74) [McManus 1991:107 §6.12]. NN appears in personal-name-final position at Bressay, Burrian, Golspie, and perhaps twice at Lunnasting [for a discussion of apparently random gemination of some consonants in ogham see McManus 1991:124-6 §6.31].
The name Edern is well attested in Welsh and appears to derive, through regular sound changes, from the late Latin Eternus. Edern occurs in the Welsh genealogies as the name of three early figures all from same
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
489
family and all having some connection with North Britain, they are Edern ap Nudd/Ludd ap Mi, Edern ap Cunedda Wledig and Edern ap Padarn Beisrudd [Baitrum 1966 s.n.]. In discussing the rule that, before rn Brit, e remains [1953:278], Jackson gives W. Edyrn as an instance of the occasional example of e > i [279]. He goes on to add that, in Latin loans only, Brit, em may give WCB. am, e.g. Patemus > W. Padam (but OW. Patem and B. Padem) [280], explaining that er > ar was a Vulgar Latin rather than a British development [281]. If Patemus could become Padam, then there seems no difficulty in explaining a North British name Edam from Etemus. As for the final -an I-on, it is presumably the diminutive, the pair Edarn/Edarnan being paralleled in other Pictish pairs, e.g. Drust/Drostan, Talorg/Talorgan. An alternative derivation is provided by Koch, who suggests that the Pictish Idaman may be the equivalent of the Romano-Celtic hammus (with /5/ < /j7 < intervocal /s/) [Grammar of Old Welsh § 7, in prep.]. Whatever the preferred interpretation it is hard to accept Jackson's dismissal of the name as apparently not Celtic* [1955:140].
The appearance of the name Ethernan on no less than four of the extant Pictish inscriptions is all the more striking since it is not attested in the Pictish King-List, or any other source of Pictish secular names. There was, however, an important Pictish saint called Ethernan and it is tempting to interpret some or all of these inscriptions as honouring the saint rather than a string of local magnates coincidentally all called Ethernan. The Aberdeen Breviary gives the collect for Ethernanus, bishop and confessor, on the second of December, crediting him as the founder of Rathen in Aberdeenshire [Laing 1854]. As Forbes comments, the legend of Ethernan given therein 'contains no historical facts' and does not appear to be ancient [1872:333]. The saint is said to have been of noble binh among the Scots, to have travelled to Ireland for his education, and to have returned to Scotland where he became an exemplary bishop, but these are merely standard topoi in later medieval Scottish vitae. Ethamon is not mentioned in Irish calendars (though Watson points out a Bishop Ethirn son of Laitbe, of Domnach Mór maic Laitbe at May 27 in the Martyrology of Gorman [1926:321], cf. epscop Eithern mac Laithbe in an Airgialla genealogy in Rawl. B.502 [CGH p. 152]).
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
490
There is, however, a reference in AU to the death apud Piaores in 669 of one Itarnan and his companion Corindu. No further information is given, though it seems plausible from the context that they are churchmen.
Whatever the historical origin of the cult of Ethernan, and the nature of the connection, if any, with the figure named in AU, it is clear that Ethernan was an important Pictish saint. There is evidence for his cult over a wide area between the Moray Firth and the Firth of Forth [see map]. His link with Rathen in Buchan, ABD [Alexander 1952:118], is strengthened by the supposed site of the saint's hermitage nearby on the east side of Mormond Hill in 'a large solitary den called St. Eddran's Slack* [Macfarlane 1908:(i)57]. Further south, the church at Madderty in Perthshire was dedicated to Ethernanus (1220) (roughly contemporary instances of the forms hhernanus 1219, and Ydarnasius 1220, show the continuing alternation between I- and E- [Inchaffray charters, quoted by Watson 1926:321]). A fair at Forfar on his day appears to have retained the name also, in the light of a local reference to St. Tuetheren's fair [Forbes 1872:331].
Perhaps the most important site associated with Ethernan is the monastery on the Isle of May at the mouth of the Firth of Forth. In the 1130s a community of Cluniac monks from Reading was established on the Isle of May to service shrines of St. Ethernan, who was later confused with Adrian [Barrow 1973:186]. A charter of the 1140s or 50s provides the earliest documentary link, the patron is refered to as Sancto Ydernino de mai [Duncan 1957:74], but the Ethernan connection must be much older.
Recent
archaeological excavation by Glasgow University Archaeological Research Division has revealed a long cist cemetery of Pictish date under the later medieval monastery [Stephen Driscoll, pers. comm.]. The record of Alexander Cumyn, earl of Buchan granting wax ad luminarium S. Ethernini de insula de May is of particular interest in the light of the saint's association with Rathen [Register of the Priory of St. Andrews, quoted by Forbes 1872:331]. Fordoun is an early foundation associated with the Palladius traditionally the educator of Ternan and other Pictish missionary saints. A link between Ternan and
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
491
Ethernan has been dismissed in the past [Anderson 1922:180 n.3], but is perhaps deserving of further scrutiny, as is the possibility that it is Itharnan who lies behind the Itharnaisc commemorated at Leuchars and at the imponant site of Lathrisk, both in Fife, who is equated with Itharnaisc of Cloenad, 22 December [Watson 1926:324]. It is possible that other Ethernan commemorations lie behind dedications to Ernan etc. The personal name MaelEthny ('devotee of Ethernan'?) may provide further evidence for the cultus of the saint.
Over and above the ogham inscription, Ethernan is commemorated in the parish of Scoonie in the modern place-name Aithernie. The name appears in several twelfth and thirteenth century charters following its gift by Duncan earl of Fife to the nunnery at North Berwick (Adhernin 1178, Athernin 1178, Adernin 1199, /Ethernin 1212x1214, Adherenin 1220 also Cnocdereniri) [I owe these references to Simon Taylor and am most grateful to himfordiscussing Ethernan and his cult with me, see Taylor 1995]. In the first four references Aithernie is paired with Montrave in the same parish (Matheryue, Mathrithe, Mathrive, and Mathriue). Watson interprets this as British (i.e. Pictish) mad-tref'good farmstead* [402-3], but, as Dr. Taylor points out to me, it could equally well be Gaelic.
DISCUSSION Though a substantial Class II monument, the Scoonie slab is not in the first rank of Pictish sculpture, in terms of either ambition or accomplishment. Its inferior workmanship and comparatively small size point to a date in the second half of the eighth century or later. It is entirely typical in its choice and arrangement of elements. Other examples of slabs with a cross in relief on one face and an incised 'secular' scene with symbols on the reverse are too numerous to mention. Amongst the ogham corpus both Golspie and Brodie conform to this pattern. Like Golspie, Scoonie's ogham is on the right edge, not the more usual left. This may reflect a weakening of the old ogham convention established in Ireland, but in Scoonie's case there is still an element of doubt over the possibility of additional lettering on the left.
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
SCOONIE
492
The Scoonie inscription is written in a simple form of the script. The lettering, however, is competently executed and Macalister's assessment is unduly harsh. The spacing of letters and their relation to the stemline is not, as he maintains, careless, but merely constrained by the impingement of the animal carving. He has overstated the irregularity of the letters which is merely in keeping with the 'free-hand' style of the rest of the carving. While Scoonie lacks the neatness of some of the more delicately and finely carved oghams, such as Formaston, this is due in part to the rather rough stone and to the carving technique used. As Gordon has pointed out, the smoothness of the Brandsbutt lines was achieved only with the expenditure of considerable effort [1956].
The Scoonie text is clear and uncontroversially interpretable as an otherwise attested Celtic male personal name. The impon of this name, however, is obscure, especially since there remains the possibility that a small ponion of text is missing. Since Pictish appears to have lacked nominal declension [Koch 1983], there is no way of knowing the grammatical case of Scoonie's EDDARRNONN. Thus there is little to choose between a range of possibilities, e.g. 'Ethernan (made this cross)', '(this cross is dedicated to) Ethernan', '(pray for) Ethernan', '(in memory of) Ethernan'. The lack of a patronymic may be significant. Charles Thomas is of the opinion that 'the most likely explanation is that single-name memorials were for priests, clerics, who proposed to discard all early parentage ("call no man on eanh your father"; Matthew, 23:9)' [1993:22]. This interpretation is strongly supponed by the roman alphabet inscriptions of southern Scotland [Thomas 1992], but is not necessarily universally applicable. In Ireland one of the very few ogham inscriptions to explicitly identify a priest, also gives the name of his father: QRIMITIR RONANN MAQ COMOGANN 'the priest Rónán son of Comgán' (CIIC 145 Anaglen, Corcu Dhuibhne).
There are a small number of single-name only ogham inscriptions in Ireland, McManus gives as examples CIIC 190 and 191 [1991:52 § 4.6], and a few more with ANM + single-name. Amongst roman alphabet inscriptions in Ireland, single-names are most commonly prefaced with a requestfora prayer, though there are a few which appear to stand alone, e.g. on the east and west crosses at Galloon, Co. Fermanagh (CIIC
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
956-7).
SCOONIE
493
Single-name only inscriptions are more common in Wales, for instance, the Eiudon stone,
Llanfynydd, Carmarthenshire (CIIC 997, Nash-Williams 1950 No. 159), the Gwhirt stone, Llanarth Cardiganshire (Nash-Williams 1950 No. 110), and the Briamail Flou stone, Llanfyfeelog-fach, Brecknockshire (CIIC 997, Nash-Williams 1950 No. 49). Whether these men were ecclesiastics or secular magnates, however, is not known.
Since we know Ethernan was a saint, and that he appears, on place-name evidence to have been culted at Scoonie, I am inclined to interpret the ogham text as referring to St. Ethernan. Irish examples of dedications to saints include the cross of Patrick and Columba at Kells, Co. Meath (CIIC 587) Patriái et Cdumbe crux I do rigne Muiredach, and the slab from Kilbrecan, Inis Mór, Aran (CIIC 531) inscribed S[an]c[t]i Brecani.
Such dedications need not be contemporary, indeed this is unlikely, and so the
identification of Ethernan gives no more than a terminus post quern of 669. If Ross Samson is right and Pictish symbol pairs represent personal names, then the name of the patron of the cross may have been contained in the symbol or symbols at the top of the slab. The choice of a hunting scene may indicate that this benefactor was a layman. If Scoonie and Fordoun are to be interpreted in similar terms, then Ethernan appears to have been particularly popular among pious Pictish magnates.
BIBLIOGRAPHY
OS Record card - NO 30 SE 15
n.a 1866; Stuart 1867:6 pi. 12; Brash 1879:353-4; Southesk 1884:183-5, 1886:39; Ferguson 1887:140-2; Rhjs 1892:267, 1893:411, 1898:344-5; Nicholson 1896:A - 23-5, App.62-3, 87; ECMS 347; RCAHMS 1933:268 No.495; Macalister 1940:193 pLIIb; K. H. Jackson 1955:139; Stevenson 1955:114; Padel 1972:141-3; Ritchie & Fraser 1994:7.
CATALOGUE OF OGHAM INSCRIPTIONS OF SCOTLAND
PREVIOUS READINGS
Brash
DOCEIOSOSN
Ferguson
EDDARRNONN
Southesk
[MA]QQI DAHQUQUALLE
Rh?s
EHTARRNONN or EHTARRMSONN
ECMS
EDDARRNONN
Macalister
EDDARRNONN or EHTARRMSONN
Padel
EDDARRNONN
SCOONIE
494
en
cm
2
i 8 | c £ cs 00
O
o o u to
10
9 8
#
SCOONIE - Ogham inscription with key to numbering
*
O
10
*o
liiiiiifnii ; . : T T i
aa
:":i
*>o "i""" _' _"j
SITES ASSOCIATED WITH THE CULT OF ETHERNAN
495
WHITENESS
DISCOVERY This ogham-inscribed fragment was dug up in the kirkyard of St. Ola's, Whiteness, in the parish of Tingwall on the mainland of Shetland (NGR HU 3866 4442). No further details are recorded concerning the date or circumstances of its discovery. It was presented to the National Antiquities Museum in 1946 by Peter Moar (NMS Cat. No. IB 256).
SITE The graveyard at Whiteness has been in use for many centuries but all traces of a former chapel have disappeared. The low platform in the nortli side of the kirkyard identified by MacDonald as the site of the old chapel [1968:131], is probably, in fact, the foundations of a now demolished building shown as still standing on the 1903 OS map [OS record card]. Little is known of the site's early history. Stevenson suggested that the church was probably founded before Earl Sigurd's forced conversion in 995 on the grounds that the ogham was unlikely to be later than the tenth century, though he admitted that a later date was not impossible [1981:286]. While, in all probability, the site does pre-date the conversion of the Vikings of the Northern Isles, our current knowledge of ogham in Scotland does not permit dating on the basis of the inscription alone. The dedication to the eleventh century Norse saint King Olaf Haraldsson does not rule out the possibility of an earlier foundation subsequently re-named.
A total of threefragmentsof early Christian sculpture have been discovered at St. Ola's, all of which are now in the NMS. The first (NMS IB 248) was unearthed during the enlargement of the graveyard in 1933 by workmen rebuilding a wall [Tait 1938:370]. It is pan of a grey sandstone slab [see fig.], carved on two faces with, on one face, interlace forming part of a cross-shaft, and on the other two vertical bands joined at the foot in a triquetia knot [Stevenson 1981:285]. The second (NMS IB 257), also of grey sandstone, was found in 1939 by Mr J. H. Goodlad when digging a drain about 100 yds from the
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
WHITENESS
496
churchyard on the croft known as Kirkhouse. His impression was that the stone could have been re-used in a previous drain [OS record card]. Peter Moar donated it to the NMS in 1946 [Stevenson 1947:193]. It is incised on one face with a meander pattern which appeared to Stevenson [1981:287] to be a simplified form of an all-over key-pattern design which occurs on various stones from eastern Scotland. He compared it with the Fan stone [ECMSfig.51]which may date to the second half of the ninth century. The third fragment, which bears the ogham inscription, is discussed below.
A stone-lined grave was discovered on the Whiteness site in 1938. It contained an iron axe-head of late ninth century Norse type, which is now in Lerwick Museum [Shetelig 1954:69]. The association of axehead and grave is disputed.
The buried axe may represent a residuum of pagan practice among
Christianised Vikings, though Crawford considers it possibly just a stray find [1987:163-4].
DESCRIPTION OF OBJECT A small fragment of sandstone, carved in false-relief on one face with the remains of an interlace knot, bosses, and incised with four or five ogham letters [see fig.]. Stone:
Red sandstone
Dimensions:
180mm wide x 220mm max. height x 73mm max. thickness
Condition:
Severely trimmed down, carving generally clear. One small section of a larger inscription, a portion of the ogham is damaged.
The fragment appears to have been deliberately and carefully trimmed rather dian smashed. It is uncertain whether the bare reverse face of the fragment is original. The edge parallel with the ogham may be intact, but even with this clue, it is difficult to establish either die correct orientation for the fragment or the form or scale of the monument from which it survives. If Stevenson's reconstruction of the interlace design is correct [see fig.], we must posit an original at least four-times the size to complete the design. If that itself is only one arm of a cross, as Stevenson appears to have thought, then a considerably larger original must
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
WHITENESS
497
be envisioned. Whether this original was a free-standing cross slab or a recumbent grave-cover cannot be established. Another possibility is that it is an architectural fragment, perhaps pan of a box shrine or frieze.
DESCRIPTION OF CARVING - DECORATION The surviving decoration consists of the remains of two pairs of small bosses, and two bands, each with central groove, which were reconstructed by Stevenson as one quadrant of a cruciform interlace [seefig.], Romilly Allen's pattern No. 758 [ECMS. 11:294]. Apart from this Shetland example, the motif is restricted to Northumbria where it is found only on later Anglian sculpture, e.g. a cross-shaft at Aycliffe from the tenth or possibly eleventh century [Stevenson 1981:286]. As reconstructed by Stevenson, the Whiteness slab would have two small bosses in each cross-arm. There are no precise parallels for such a formation, though Stevenson noted similar motifs in the north of England [286]. The significance of this Northumbrian connection is not clear, especially since the other Whiteness fragments have no such stylistic affiliations.
The ogham letters on the Whiteness fragment are contained in a 50mm wide border which is very similar to the broad bands carrying the letters of the Abemethy and Brodie oghams. In all three instances it appears that the border was designed to take the ogham and that the text is an integral pan of the composition. It is conceivable, of course, that an opportunistic oghamist made use of a highly suitable surface on a standing monument, but if that were so, one would expect to find monuments with similar bands, and no ogham, which is not the case. The scripts of the Abemethy, Brodie and Whiteness oghatns are in keeping with the formality and monumentality of the rest of the carving and there are no positive reasons for supposing them later additions. In the Whiteness case, the impression of contemporaneity is strengthened by a comparison of the carving technique of the ogham strokes and the central line on the interlace bands. The flat moulding on the edge of the Formaston fragment is much narrower than the Whiteness, Brodie and Abemethy examples and may not have been originally intended to bear ogham. A
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
WHITENESS
498
related example is the Golspie slab which has, however, not a broad flat band, but a curved arris, turned into a moulding by a deeply incised line. Unity of design of the Whiteness carving is further suggested by the alignment of the oblique first stroke of the R with the strong diagonal axis of the interlace crossarm.
DESCRIPTION OF OGHAM INSCRIPTION The individual ogham strokes appear to have been first pocked, then smoothed, and have a broad, open U-section. The forward slope of the ogham consonants indicates the direction of reading, but since the orientation of thefragmentis unknown it could be taken in any of four positions, and therefore read in either of two directions. The most common position for ogham texts onfree-standingstones is on or against the left-hand edge, reading vertically upwards. ExceptionsfromScotland include the Newton stone (left-hand, vertically downwards), Golspie and Scoonie (right-hand, vertically upwards), Bressay and Brodie (right and left, vertically upwards) and Logie Elphinstone (on the face, in a circle). There is reason to believe that the Abernethy ogham, which is closely comparable with the Whiteness fragment, was arrayed horizontally. We have, therefore, no clear indication of how to orientate the present example. For the purposes of the following discussion, thefragmentwill be referred to as if resting on the fractured straight edge with the ogham running vertically on the left [see fig.].
The original extent of the inscription cannot now be determined. There are a few clues to the position of these letters relative to the complete text, but they are inconclusive. There is no linguistic obstacle to letter 5 (R) being the terminus, either initial or final (RAQQ- or -NNAR), but only if 3 is not a mark of worddivision [see discussion below]. If the extant letters came in the middle of the text, the question is, did the ogham continue in a straight line, or turn the corner ? The only lead is the physical appearance of the extant carving. The original surface of the stone is preserved in three areas, along the outer straight edge, as the band on which the ogham is incised, and on the small ponion above the interlace, to the left of the first ogham letter. That is to say, this ponion and the ogham band are one, the extant ponion forming an
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
WHITENESS
499
L-shaped field, Stevenson's diagram [1931 fig. 1] is slightly misleading in this regard. If the interlace had continued beyond this portion one might have expected the surface to be recessed like the rest of the background and the inside line of the band continued. Since it does not, one might deduce, either that this was the edge of the stone and there was no further carving, or that the decoration continued, but with an entirely separate design or motif, A third possibility is that the band turned the corner and continued round the sunken panel of interlace. There may or may not have been a further field of decoration beyond this strip. If the border was indeed L-shaped, it is possible that the ogham may have continued round the comer and along the band. Nothing is visible directly above the top of the interlace. Any letters immediately after the turn would have had to belong to the b- or h-aicmi which would produce an awkward cluster. After about four or five strokes worth of space, there is, however, an indentation which, though probably just a chip, could be the remains of a letter. If such a reconstruction were accepted one might wonder why go to the bother of trimming the stone only an inch or so from the edge. One possibility is that the stone, when struck, fractured along the stem-line. The evidence is insufficient to distinguish between the several hypotheses [see figs, for possible reconstructions]. What remains is as follows, reading from the bottom up [see fig. for key to numbering]: 1
Three short strokes to the right of the stem, sloping forward, bound - V. The fracture occurs at the point of intersection of letter-stroke and bind-stroke so it is impossible to establish if there were further strokes to this letter. It could have been S or N.
2
Five shon strokes to the right of the stem, sloping forward, bound - N.
3
There follow some obscure marks, which, though clearly visible, are not as sharp as the rest of the carving. From a distance they look like a large, back-to-front C across the stem and consist as follows: one shon stroke to each side of the stem parallel to and equidistant from the it, but closer to it than are the bind strokes. The one on the left is joined to the stem by a shon, backwards-sloping stroke, which is itself followed by a similar stroke sloping slightly backwards. Stevenson described these two as 'much less definite than the others' [ 1947:193]. Previous writers have identified these as a possible D, but there are three difficulties with this interpretation -
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
WHITENESS
500
firstly, they slope in the wrong direction in comparison with the other coasonants; secondly, unlike the other consonants they lack bind-strokes (the horizontals are in the wrong position to be bind-strokes, too close to the stem, and placed before the first stroke, not between it and the second - 'their ends are certainly not joined* [Stevenson ibid.]); thirdly, such a letter would be very cramped in comparison with the generous spacing of the others. The remaining letters are so clearly and carefully carved that these appear more like accidental scuffs and scores. The horizontal marks in particular look like spalls. The areas of stone to either side of these scores have suffered damage, perhaps when the piece was so savagely trimmed. The question is of what did this section originally consist ? If it had been devoid of letters there would be an awkward gap, out of keeping with the spacing of the other letters. While it would be unwise to be dogmatic about spacing with so little remaining, another interpretation is possible. There is a possible dot to the right of the stem opposite the second of these oblique strokes. If it is original, it could be the remaias of a pair of the word-division dots which are such a feature of Shetland oghams. The other dot might be obscured by the tip of the second short stroke. A pair of dots as word-dividers at this point would be nicely in proportion with the rest of the letters, Stevenson, however, dismissed 'what might be interpreted as dots' as 'probably accidental bruises1 [1947:193]. 4
One long stroke, perpendicular across the stem - A
5
Five long strokes, obliquely across the stem, sloping forwards, bound at both ends - R. The lower portion of the final stroke is lost, but not in doubt.
This gives : -](v/s/n)N(:/d)AR[?- or, reading in the opposite direction: -?]RA(;/l)Q(t/c/q)[-
FORM OF SCRIPT Though only a fragment remains, it is clear that Whiteness was written in a stately Type lib form of the script, with sloping b-Ih-aicme strokes and vowel-stroke occupying the whole ogham band. The component strokes are generously and evenly spaced with ample room, almost a stroke's worth, between successive
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
WHITENESS
501
letters of the same aicme. In keeping with the inscription's late date, all appropriate letters are bound. The letter strokes are rather short which, combined with the measured spacing, gives a rather squat profile to the letters, a feature noticed elsewhere among later oghams, Cunningsburgh 3, for instance. Abernethy provides an exact parallel in script form, though if Whiteness had continued further we might have been provided with someforfeda or other supplementary features. Cunningsburgh 2 has similarly bound letters, but vowel-strokes a little shorter than m-aicme strokes, and angled vowels.
INTERPRETATION OF TEXT Too few letters have survived to allow identification of the language used, let alone the nature of the text. There is nothing about the sequence either to prohibit or recommend a Goidelic or Brittonic interpretation. If 5 were thefinalletter, the sequence -NNAR could represent a Norse personal name. RAQQ- would be more difficult to interpret. If J is to be interpreted as marking a word-boundary this further enlarges the number of possible explanations of the sequence. Double N is a common enough ending for Celtic personal names, e.g. in ogham onhography EDDARNONN, and AR might represent, for instance, the beginning of a Celtic name in An- or a Norse one in Am-. But really, these four letters could be just about anything.
DISCUSSION Though virtually nothing linguistic can be made of the lettering on this small fragment of a high-status Christian monument, Whiteness has a number of implicatioas for the palaeography of Scottish ogham. If Stevenson's dating of the interlace pattern is accepted, then Whiteness represents an important anchor point in the typology of Scottish ogham, one of the few examples which can be dated independently. The significantfeaturesof the ogham script used are; the bind strokes, though it is interesting that the 'hammerheaded1 bound form of A is not employed; the forward slope of the Xhaicme letters and the fairly wide spacing of the letters, more than one stroke's worth of space is allowed between them. The appearance of word-dividing dots would imply a ninth century or later date.
CATALOGUE OF THE OGHAM INSCRIPTIONS OF SCOTLAND
WHITENESS
502
BIBLIOGRAPHY
OS record card - HU 34 SE 3
Stevenson 1947:191-3, pl.22 (4); Padel 1972:148-50; Stevenson 1981:285-7, pl.22 (4).
Site only:
RCAHMS 1946:125 No. 1527; MacDonald & Laing 1968:131; Shetelig 1954:69 (axe); Tait 1938 (other sculpture).
PREVIOUS READINGS
Stevenson
-N(?D)AR-
Padel
-(v/n)NdAR-
WHITENESS - Ogham-inscribed fragment
0
s
to
iff
t o c*.
WHITENESS - Stevenson's reconstruction of interlace pattern [1981 fig.l]
V v\
/ C
o
1
S
m
s
I2
'S
8.
O
Í
60
a
z
S
WHITENESS Schematic representation of ogham inscription (key to numbering)
WHITENESS Ogham-inscribed fragment [STEVENSON 1947 pL22 (4)]
0
i?
t. A X t ,1 L
ID
-I
"Front"
"Back"
WHITENESS - Fragment of cross-slab NMS IB 248 |TAIT 1938 fig.5]
O
5
10
if
ao en
503 APPENDIX
DUBIA
The following have, at one time or another, been put forward as possible ogham inscriptions [see map]. Unless otherwise stated, I have examined them all personally. Most are clearly not ogham-inscribed. The remainder are too doubtful to merit inclusion in the main catalogue.
Abernethy 2 Brough of Birsay Burrian Drumoyne Foshigarry Gigha 2 Gurness Pool Iona Lochnaw Tollard Vaivoe Others
THE OGHAM INSCRIPTIONS OF SCOTLAND: DUBM
504
ABERNETHY
2
DISCOVERY This stone was discovered on Tuesday 29 January 1895 by the sexton of Abemethy church while he and two companions were digging a grave in the churchyard just north of the ruins of the ancient church (NGR NO 190 164). Full details are provided by Butler [1895] and Southesk [1895]. The stone was sent to the NMS for examination but returned to Abemethy by April 1895 [Southesk 1895:246].
DESCRIPTION OF OBJECT A small stone incised on one face with a crown, a bird, the letter Wand a horizontal 'ogham' inscription. Stone:
'Sharp silicious grit' [Southesk 246]
Dimensions:
approx. 0.48 x 0.33 x 0.10m
DESCRIPTION OF CARVING Southesk describes the lines as 'mere scratches, such as a common iron nail might produce' though some appear to have been chiselled [246]. The stem-line is drawn horizontally 'exactly one foot in length' and is 'straight as if run on a ruler'. From Southesk's diagram [249] the strokes are as follows, reading from left to right as indicated by the slope of the angled vowels [see fig.]: 1
The opening marks appear to consist of a semi-circle below the stem containing three dots interpretation unclear.
2
The same above the stem, slightly overlapping with L
3
One long stroke perpendicularly across the stem - A or M.
4
Three long strokes across the stem, slightly angled at their mid-point, pointing forwards ( > > >) -U.
5
Three long oblique strokes across the stem sloping forwards - Gw.
6
One single, faint, long stroke perpendicularly across the stem - A or M.
APPENDIX: DUBIA 7
ABERNETHY 2
505
A semi-circle above the stem containing a single dot - interpretation unclear, but cf. Lunnasting. Underneath this character, to the right of and parallel with the stem is a short line which meets the first stroke of 8 to the right of the stem. This horizontal line itself has three strokes across its left end, each angled forward at its mid-point ( > > > ) - U?.
8
Four short strokes to the left of the stem sloping slightly forward, above three short strokes to the right of the stem sloping slightly forward. The first and last strokes of each group meet at the stem to form a vowel stroke (->}. The third of the right hand strokes meets the stem equidistant between the middle pair of the left hand strokes. If a fourth stroke to the right has not been omitted, these could be read as AHBHA, otherwise E.
9
There follows a gap containing only a single dot to the left of the stem. Thereafter a single short stroke to the right of the stem and, very slightly further on, a single short stroke to the left. If these were intended to meet - A, if not BH.
10
A single stroke to the left of the stem, sloping slightly forwards - H, or if taken with the previous stroke, 9 and 10 could read BD.
11
Two long oblique strokes across the stem sloping slightly backwards - 0 or G.
Which would give a reading as follows: ??(a/m)UGw(a/m)(/u)u(ahbha/e)(ah/bhh/bd)(o/g)
DISCUSSION Southesk was doubtful about the authenticity of this stone and rehearsed in detail the arguments for and against. Subsequent writers have condemned it as 'obviously a fabrication of quite recent date' [Allen], 'evidently a forgery' [Macalister] and 'probably a forgery' [Padel]. Without seeing the stone it is impossible to be certain, but Southesk's arguments in favour of authenticity do not seem sufficiently strong to counter the doubts raised by the presence of moss along one edge of the stone, and 'the fresh whiteness of the lines' [Southesk 246]. The text seems highly improbable, at the very least the stone has 'been tampered with in modern times' [244]. Almost certainly this stone is a creation of the 1890s.
APPENDIX:DUBIA
ABERNETHY 2
BIBLIOGRAPHY OS Record card - NO 11 NE 85 Butler 1895; Southesk 1895:244 ill.; ECMS 341; Macalister 1940:215 n.15; Padel 1972:43.
PREVIOUS READINGS Southesk
AUUAaoOHAAHO
506
t
4, x
.
í
.
7 * " j f n
11 i t 11
^ •1
ABERNETHY 2 - Sketch by the Earl of Southesk [1895:249]
507
BROUGH
OF
BIRSAY
DISCOVERY In addition to the three genuinely ogham-inscribed building slabs from the Brough of Birsay, a small pebble recovered from the Lower Norse horizon has been alleged to be carved with ogham [see fig.]. Curie describes the pebble, ref.no.595, as coming from the paved passage I in Area II. It is now in the Tankerness House Museum, Kirkwall.
DESCRIPTION OF OBJECT A rounded pebble, incised on its flat base with 'lines resembling ogham characters' [Curie 1982:120]. Material:
Stone of unspecified type
Dimensions:
42mm diameter
DESCRIPTION OF CARVING I have not examined the item personally, thefollowingdescription is based on Curie's illustration. The 'stem' is carved straight across the diameter of the base, from side to side. Three strokes depend on it. Two, coming approximately one third and two thirds of the way along the stem respectively, are to the right of and perpendicular to the stem and could be interpreted as two Bs. The third stroke crosses the stem between them, much closer to the second, with which it intersects, rather than equidistantly between the two. The cross-stroke occupies the middle two quarters of the ogham band and slopes slightly backwards. Since the gradient is not marked it is difficult to decide whether or not it is deliberate. If not, then the letter could be A. If the slope is intentional it is not explicitly so and, more seriously, leans in the wrong direction. The interpretation M should not, however, be entirely ruled out since the m-aicme strokes of Ackergill tilt slightly backwards. So, the three strokes could be interpreted as an ogham reading BAB or BMB, or, if the inscription were inverted, HAH or HMH (the m-stroke would still be 'backwards').
APPENDIX: DUBIA
BROUGH OF BIRSAY
508
DISCUSSION This is certainly one of the most ogham-like of all the allegedly ogham-inscribed diibia recovered in Orkney. The text as it stands, however, makes no obvious sense. This in itself is not sufficient to rule out the possibility that the lines are ogham since there are comparable examples of genuine inscriptions consisting of apparently meaningless letters carved, for whatever reason, on personal belongings. An Orkney example is the runic 'AAA' lightly incised on the bone pin from Westness, Rousay [Liestal 1984:232]. Whether such texts are empty doodles or have semantic meaning, are prosaic or symbolic, is not clear. The rune-inscribed seal's tooth [see fig. following main entry on Birsay] found in the same area of the Brough as this pebble illustrates the amuletic use of inscribed objects at Birsay in this period. There is nothing about the markings on the pebble which preclude interpreting them as ogham, however F am not prepared to accept if as genuinely oghamic until I have had the opportunity to examine it in person.
BIBLIOGRAPHY Pebble: Curie 1982:120, 70 ill.45 See entry on Birsay for complete site bibliography.
BIRSAY - Allegedly ogham-inscribed pebble [Curie 1982:ill.45 detail]
BURRIAN - Spindle-whorl inscribed with allegedly ogham-like markings [MacGregor 1974:91 fig 18, detail]
509
B U R R I AN
DISCOVERY This item is one of several dozen spindle-whorls uncovered in the nineteenth century excavation of the Broch of Burrian in North Ronaldsay. There is evidence of early Christian activity at the site including an ogham-inscribed cross-slab, but it is unclear if this spindle-whorl relates to this or an earlier phase [see entry on Burrian for details of site]. Along with the other finds from the site it was donated to the NMS in 1872 [Cat. No. GB 58]. It was described and illustrated by MacGregor in his account of the Burrian finds [MacGregor 1974:92 No.252].
DESCRIPTION OF OBJECT A small pink sandstone spindle-whorl incised on one surface with ogham-like decoration [see fig.]. Stone:
Red sandstone
Dimensions:
Diameter: 18-20mm; Thickness: 15mm-12mm; Diameter of perforation: 7mm.
Condition:
Intact and well preserved.
DESCRIPTION OF CARVING The whorl is simply incised with a circular 'base-line', the two ends of which overlap in a spiral fashion, which is over-lain with about ten short perpendicular strokes unevenly spaced along its whole length. The strong superficial resemblance to the genuinely ogham-inscribed whorl from Buckquoy has lead some to interpret these lines as ogham. They could, indeed, be read as such, though the result would be a string of inexplicably spaced vowels: AAAAAAAAJ
or UAAUO. Holder suggested 'the rather unlikely
AAAAAAAB' with 'some son of magical or talismanic function* [1990:71], On balance, however, the carving cannot be accepted as genuine ogham. This does not preclude the possibility that the lines are 'an illiterate attempt at an inscription' [MacGregor 92].
APPENDIX: DUBIA
BURRIAN
510
Without becoming embroiled in the vexed question of the talismanic use of pseudo-writing, my only comment would be that the carving on the Burrian whorl is actually quite a long way from ogham and, if it is an attempt to represent an ogham inscription could only have been produced by someone who was rather ignorant of the real thing. One would have thought that anyone with enough exposure to the script to want an inscription in it would have been able to produce a more convincing simulacrum. The crossslab demonstrates advanced knowledge of the script at the site, though, of course, the dating is not sufficiently precise to evaluate the contemporaneity of the two items. If the carving on the whorl is an attempt at ogham it cannot have been provoked by the inscription on the cross-slab since the form of the script used on the latter is very different; virtually all the letters are either bound or foffeda.
On balance I am inclined to agree with MacGregor that '[t]he ogham-like appearance of the decoration ... seems to be purely coincidental'. A number of sites have produced spindle-whorls incised with radial decoration completely unrelated to ogham. MacGregor cites two parallels (a tenth century one from Jarlshof in Shetland, and a Roman-British one from Coygan Camp, Carmarthenshire), Padel mentions a radially decorated whorl 'in Perth Museum1 (I have been unable to ascertain precisely which one). There are numerous other parallels, including one from the same site (No. 249).
BIBLIOGRAPHY Padel 1972:15; MacGregor 1974:92 No.252; Holder 1990:71.
511
DRUMOYNE
In the store of Kelvingrove Museum, Glasgow is a large stone allegedly carved with ogham [Ace. No. '3325]. Up until the 1930s this stone was in use as a gatepost at West Drumoyne farm [NS 548657], just south of Fairfield, Govan, very close to the river Clyde and near Govan Old Church. About 1933 the stone was moved by the Health Department of Glasgow Corporation and the area is now a housing estate. I have not seen the stone in person, but from MacLennan's photograph of the stone, taken about 1931 when the stone was still in situ, it is apparent that the 'writing' is not ogham. The marks appear artificial but quite unlike any recorded prehistoric rock an elsewhere in Scotland. In the note accompanying his photograph MacLennan says that 'The writing seen on stone is not unlike that seen on the (mystery stone) of Langbank...'. The Langbank stone is discussed by Andrew Lang in his The Clyde Mystery [1905:40], but he throws no light on the Drumoyne carvings.
BIBLIOGRAPHY Unpublished:
Archive of Kelvingrove Museum.
A. MacLennan, Monochrome photograph and
accompanying note, 7.7.31.
I am most grateful to Dr Colleen Batey, Curator of Archaeology, Kelvingrove Museum, for bringing this stone to my attention and providing me with a copy of MacLennan's photograph and note.
DRUMOYNE - Pillar with incised linear markings [Photo. A. MacLennan, Kelvingrove Art Gallery & Museum]
\
\r%-
/
DRUMOYNE - Linear carvings
512
FOSHIGARRY
DISCOVERY A disc with allegedly ogham-like markings was discovered during Erskine Beveridge' s pre-First World War excavation of the wheelhouse site of Foshigarry in the Vallay area of North Uist, beside the similar site of Bac Mhic Connain [Beveridge & Callander 1931, 1932; see entry on Bac Mhic Connain for further details of the area]. It is now in the NMS, catalogue no. GNA 11.
DESCRIPTION OF OBJECT A flat bone disc with a large central perforation and smaller radial one. The ogham-like decoration was evidently carved before pan of the disc was broken off. In addition there is an interlace pattern and some other doodle-like scratch* [see fig.]. Material:
Unfused cetacean vertebral epiphysis.
Dimensions:
Diameter 99mm.
Condition:
A ponion of the rim has broken off.
This is one of six such discs recovered from Foshigarry, though the only one to bear carvings. Comparison with closely similar artefacts recovered from the Broch of Gurness [Hedges 1987 fig.2.32 no. 174; see below] lead Hallén to interpret these discs as lids of vessels made of hollowed-out cetacean venebrae.
DESCRIPTION OF CARVING Hallén describes the ogham-like carving as 'a thin circumferential line crossed by 74 thin shon lines roughly at right angles to it' [217]. She quotes Dr. R. M. Spearman of the NMS as saying the carving is not real ogham. Having examined the disc myself I concur, though not with his suggestion that the carving 'was probably intended to create the impression of ogam' [218]. The interlace pattern indicates an early medieval date, and the Bac Mhic Connain knife shows that ogham was known in the immediate
APPENDIX: DUBIA
FOSHIG ARRY
vicinity, but still the resemblance to real ogham is too slight to imagine it is deliberate.
BIBLIOGRAPHY Hallén 1994:218-9 ill. 11.5
513
F •A —i
:\
\
ir-<-
r, .V
/<ss'
M
"3. «
O*
.^:
Q
vC"-
FOSHIGARRY - Bone disc inscribed with allegedly ogham-like markings [Hallén 1994:218 Illus.S]
514
GIGHA
2
CURRENT LOCATION In June 1988 Kalene Douglas noticed a possible second ogham stone on Gigha in use as a fence post at NGR NR 659 538, on the NW edge of the main road facing N [Douglas 1988]. Gigha native Betty McNeill told the party that her brother had found the stone 'in a nearby field' and used it to repair the fence [Frances Hood pers. comm.]. The stone in question may be the one mentioned by a local man to one of the party of Irish Antiquaries that visited the island in 1899 [Rhys 1899.349]. Rhys evidently tried to investigate, but his enquiries to the local minister produced nothing [1901:19-20].
DESCRIPTION OF STONE An oblong shaped stone. Stone:
Dark grey in colour
Dimensions:
Approx. 0.76m high by 0.25m wide
Condition:
Heavily worn and covered in lichen
NB. Frances Hood's measurements differ from those given in Douglas 1988. She states the height as 1.10m and the width as 0.17-0.27m, and gives the location as NR 661 539.
DESCRIPTION OF CARVING There are ogham-like notches on its SW arris.
DISCUSSION Frances Hood, who was present when Ms. Douglas spotted the stone, measured it and, having compared it with the other ogham stone, suggested that the two may originally have been part of the one pillar, i.e. Gigha 2 is the portion broken off the top of Gigha 1 in 1843-4. Though the girth of the two is very similar, doubt is cast on this interesting possibility by Rhjs's statement that the broken portion was
APPENDIX: DUBIA
GIGHA 2
515
incorporated into new houses nearby [see Gigha 1 entry], and by the fact that the two stones are now four and a half miles apart. Careful examination of both is necessary to test the hypothesis, and indeed to establish that the marks on the arris are indeed ogham letters.
BIBLIOGRAPHY OS Record card - NR65SE29 Published:
K. Douglas 1988
Unpublished:
Frances Hood, Letters to the author. 23 June and July 1993.
GIGHA 2 - Lower portion of pillar bearing ogham-like markings [Photo, Fiances Hood]
516
GURNESS
DISCOVERY In addition to the genuinely ogham-inscribed knife handle from Gumess, Hedges put forward two other objects as bearing 'ogam-like' marking. Both were found in unstratified contexts and cannot be closely assigned to a particular phase of occupation [see entry on Guraess for details of site].
DESCRIPTION OF OBJECTS
1:
CASKET LID
A 'casket lid' made from a cetacean vertebra incised with numerous lines
described by Hedges as 'ogam-like'. Material:
Whale vertebral epiphyses
Dimensions:
202mm x 115mm x 22mm thick
Condition:
Fragmentary
I have examined the item, No. 174 in Hedges's catalogue, and am in no doubt that the scratches, though artificial, are not ogham. The identification as ogham is alsorejectedby Holder [1990:56]. The object is currently in the NMS [Cat.No. GA 77.188].
2:
PEBBLE
A 'naturally occurring sandstone pebble shaped like a whetstone' incised along
one side with numerous lines described as 'ogam-like' by Hedges. Material:
Sandstone
Dimensions:
Max. Length 83mm, Max. Width 23mm, Max. Thickness 14mm.
I have not seen this item, which is not illustrated in Hedges's catalogue (No.802). I am doubtful, however, that it is ogham. Hedges states that at the time of publication, ie. 1987, the pebble was at the Historic Buildings & Monuments Directorate of the Scottish Development Department (the predecessor of Historic Scotland). It's excavation reference number is GA 77.426.
APPENDIX: DUBIA
GURNESS
BIBLIOGRAPHY See entry on Gumess for complete bibliography. Lid:
Hedges 1987:207 No. 174, ill. 106 fig.2.32
Pebble:
Hedges 1987:245 No.802, not ill.
517
GURNESS - Bone casket-lid allegedly inscribed with ogham [Hedges 1987.11:106 fig.2.32, detail]
518
ION
A
An ogham-inscribed slab in the garden of 'Duncraig', lona, allegedly believed by the present occupants to be ancient, is in fact the work of lona bee-keeper and handyman Mr Bruce Wall. In addition to the ogham, the slab is carved with a sword typical of Medieval West Highland sculpture [see Bannennan & Steer 1977]. The story of its creation is recounted in the following extract from Mr Wall's letter to me: 'I first came to lona 25 years ago & after seeing the carved burial stones in the grave yard thought, if the old craftsmen could carve them so could I so I had a go, and the stones in Duncraig are some of them. The stone you are alluding to was a lump of rock I found and I decided to carve the sward [sic] & put an Ogham script just for fun. The inscription I put on it means Olaf Olaffson which I thought sounded Scandinavian. I carved it with a tungsten tipped chisel as it was a particularly hard rock. The old craftsmen would have highly tempered iron chisels but would have had to be sharpened quite often by a black smith. I marked it out with a pencil before carving it/
BIBLIOGRAPHY Unpublished:
Bruce L. Wall, Letter to author, August 1992.
I am most grateful to Ms Marianna Lines, Collessie, Fife, for bringing this stone to my attention and putting me in touch with Mr Wall.
519
LOCHNAW
DISCOVERY This fragment of a possible ogham pillar was found in the late 1960s or 1970s during ploughing at Lochnaw in the parish of Leswalt, Wigtownshire (NGR NW 990620). It was brought to the Observatory Museum, Dumfries during the latter years of Alfred Truckell's curatorship [Truckell pers. comm.] and is now on display in die 'Sacred Stones' room.
DESCRIPTION 3F OBJECT The stone is afragmentof an irregular five-sided pillar [see fig.]. Stone:
Local
Dimensions:
Height: 420-540mm; Width: 390mm; Thickness: 410mm
Condition:
Poor. Fragmentary, surface severely abraded.
It currently stands to a maximum height of 0.54 m but could originally have been considerably taller. The surface of the stone is much decayed and portions have broken off in several places. Incised carving survives on two faces: 'ogham' on the front face, four small crosses on the left. The right and back faces are so damaged that it is impossible to tell if they too bore carving. The surfaces of the incised lines are themselves worn which suggests that the carved stone was exposed to the elements for a considerable time before being buried.
DESCRIPTION OF CARVING - CROSSES The left-hand face of the pillar is very rough and badly pitted, the surface having broken away in a number of places. In addition to diagonal cracks of weathering there are a number of incised lines which, running counter to these natural cracks, appear to be artificial. They comprise at least four small incised crosses which are lightly carved and indistinct. It is difficult to ascertain exactly where some strokes begin and end, the following dimensions are approximations only:
APPENDIX: DUBÍA Cross 1
LOCHNAW
520
Vertical stroke 120mm, the cross stoke 70mm angled left to right downwards. There appear to be small circular hollows under each arm.
Cross 2
Of a similar form to Cross 1, though lacking the hollows, vertical 80-110mm, cross stroke 60mm.
Cross 3
Vertical 40mm, horizontal 40mm
Cross 4
Vertical 25mm, horizontal 10mm
These crosses could belong to any period and are not necessarily contemporaneous with each other or with the 'ogham1. Such simple designs are virtually undatable. That they are early medieval votive crosses is far from proven, they could be mere doodles, or date from any period. Other 'roadside stones with crosses' have been ploughed up in the same area but apparently 'not of the same form' as these [Truckell pers. comm.].
DESCRIPTION OF CARVING - 'OGHAM1 The lines alleged to be ogham letters are carved on the flat face of the stone. They consists of a long vertical line (the 'stem') and eight or nine shorter strokes which cross or abut it at various angles [see fig. for key to numbering]. Thefragmentationof the stone appears to have occurred after these lines were carved. The 'stem' meets the final stroke at the break in the surface and there is no way of knowing how much further carving might have continued. If the lines are indeed to be taken as ogham letters, the proportions and spacing of the strokes suggest that what survives is only a small part of a longer text. If thefragmentis part of a full-size pillar then a considerable portion below the carving would have been embedded in the ground.
Regarding the incision profile of the strokes, the overall impression of shallow, straight-sided, flatbottomed strokes (especially the stem and stroke 3\ is probably misleading, the result of weathering. In most cases only the bottom of each stroke has survived, but in the less badly decayed sections the incisions are deeper. The more sharply preserved stroke 6 suggests that, when first cut, the lines had a broad, V-
APPENDIX: DUBIA
LOCHNAW
521
shaped section. Both ends of the stem-line are lost in patches of wear, what remains is 390mm long. Seven strokes of varying length and direction depend on the stem-line. From the bottom they are as follows: 1
Long stroke (80mm), perpendicularly across, the stem.
Gap of 80mm.
2
Short stroke (45mm), to the left of and perpendicular to the stem. Gap of 80mm.
3
Long stroke (110mm), obliquely across the stem-line, sloping backwards. Gap of 60mm.
4
Long stroke (110mm) parallel with J. Gap of 25mm.
5
Short stoke (90mm), to the left of the stem, parallel with 3 and 4. Gap of 50mm.
6
Short stroke (80mm), to the left of and perpendicular to the stem. Gap of 50mm.
7
Short stroke (70mm), to the left of and perpendicular to the stem.
The relationship of stroke 1 to the rest of the carving is uncertain. The line is clearly visible, but may be the result of a fracture in the surface of the stone at this point. Though on balance it is probably the case that this stroke originally joined the stem, it cannot be proved irrefutably because there is a patch of wear in the relevant area. There is no evidence of either stem-line or further strokes below 1 and, though it could be intended as an ogham letter A, given die large gap (80mm) between it and the next stroke it is perhaps more likely to be an indication of the beginning or end of die ogham inscription.
In the interval between strokes 1 and 2 much of the area to the left of die stem is obscure, but there is no indication that any strokes have been lost. Stroke 2 crosses the stem very slightly (5mm), but such a small over-shoot is unlikely to be significant if the carving is ogham. As a single stroke to the left ('above') the stem, it could be read as H.
Strokes 5 - 5 are parallel with one another. To the left their distal tips are lost in a patch of wear. To the right, 3 appears to stop just before the break in the stone and is fairly well balanced on the stem; 4 terminates before the break but is much longer to the left of the stem (90mm) than to the right (30mm);
APPENDIX: DUBIA
LOCHNAW
522
stroke 5 also has 90mm to the left of the stem but does not cross it. The interpretation of these strokes as ogham presents a problem. If 5 crossed the line they could be taken as Gw, but it clearly does not. If it is a singleton, then why is it parallel to 3 and 4, not perpendicular like 2 ? If 3 and 4 are to be taken as two strokes of the same letter, then why are they spaced as they are, with an interval of 60mm between them, as opposed to only 25mm between 4 and 5 ? Could it be that 4 and 5 are to be taken as one letter, with 4 over-shooting the stem ? But 30mm is a substantial over-shoot and this still would not account for their oblique angle. If 3 and 4 are taken together as one letter the reading could be G H; if 4 and 5 are read as one letter then, perhaps, M G; if each stroke represents a separate letter then M M H.
Strokes 6 and 7 are parallel and of similar length. Since the stone breaks off immediately following the intersection of 7 with the stem, it is impossible to tell if the letter is complete (in which case D), or incomplete (T, C or Q). Given that the interval between strokes 6 and 7 is the same as the gap between 6 and 5, it is possible that the last two are to be taken as separate letters, the first being H and the last being either another H or the beginning of D, T, C or Q. The inscription, however, is so short and fragmented that it would be unwise to make too much of the spacing between strokes.
If indeed this is an ogham text, it is to be read from the bottom up. As it survives today, the pillar sits comfortably in its present position, but it is just possible that when in was fully intact it stood the other way up. This would entail reading the ogham in the 'opposite' direction, but this does not improved the text's legibility. As they stand, read in either direction, the 'letters' make no sense.
Leaving aside the phonetic interpretation of the possible letters, their outward form presents a number of barriers to their identification as ogham. One problem is the backwards slope of the long oblique strokes. The m-aicme strokes in oghams in Scotland normally slope forwards (ie. left-to-right upwards). Aberaethy is a very dubious exception. The second problem is that Lochnaw exhibits appears to differentiate between
APPENDIX: DUBIA
LOCHNAW
523
three kinds of stroke (assuming 1 is not a letter), yet none of them are vowels. The long obliqe stroke has already been mentioned, there is also two kinds of short stroke - short perpendicular or short oblique. Ogham inscriptions tend to be consistent in having only one or the other. Those texts which have both rarely differentiate so sharply between the two kinds, but rather have strokes which slope either a little or not at all. The Lochnaw oghamist has been inconsistent with the slope of the strokes, or else, is using a non-standard form of the script. The third problem is the apparent lack of vowels in the Lochnaw inscription. In standard ogham vowels are represented either by short notches on the stem-line (primarily when the stem is the arris of a stone, rather than drawn in), or by long strokes perpendicularly through the stem (primarily as in this case, with a drawn-in stem). Taking 7 to be a non-phonetic marker, Lochnaw exhibits neither kind of vowel. It is quite possible to have three consonants in a row and in a text fragment as short as this an absence of vowels need not be suspicious, however this particular sequence of consonants is unpronounceable in any Celtic language. Even if 1 were accepted as a vowel the situation would not be much alleviated.
DISCUSSION The closest parallel for the Lochnaw stone is the four-sided granite ogham pillar from Gigha. It is more slender (250 x 310mm at the base), but its height (1.7m) is an indication of how tall Lochnaw may originally have been. The difference between these two stones is that the stem of the Gigha ogham runs along the arris, whereas on the Lochnaw example it is drawn in across the face. The two other undecorated ogham pillars from Scotland, Auquhollie and Newton, also utilize the arris for their stems, though the latter has a continuation along a drawn stem. In terms of the morphology of the ogham, parallels are to be sought in the undecorated slab from Pool and the two oghams on the wall of the King's Cave, Blackwaterfoot. These simple oghams have drawn-in stem-lines and lack modifications such as bind-strokes or supplementary letters. It is heartening that Lochnaw's closest comparanda are all to be found on the western and northern seaboard, but the comparisons are not close.
APPENDIX: DUBIA
LOCHNAW
524
The lack of a defined edge on the Lochnaw pillar may have dictated the choice of face rather than arris for lettering. That this stone was chosen, rather than one of more 'suitable1 form, suggests that special significance was attached to it before the 'ogham1 inscription was added. If so, Lochnaw would fit the prevailing pattern of the re-use of pre-historic stones as ogham pillars as exemplified in Scotland by the megaliths from Auqhollie, Logie Elphinstone, Brandsbutt, and Newton. If the little crosses are to be interpreted as votive, this may be taken in further support of the ritual significance of the pillar.
Doubts have been expressed over the authenticity of the Lochnaw ogham and these have still to be entirely confounded. Though it gives an initial impression of being ogham, closer inspection reveals irregularities. The lack of vowel strokes and the direction of some of the oblique strokes are more serious objections than the real difficulty in determining a reading. Though no other oghams are known from south-west Scotland the location of this stone need not count against it. Lochnaw is well placed to participate cultural dissemination along the sea routes between Ireland, Man and the west coast of Scotland. Place-name evidence indicates that the Rhinns of Galloway was an area of pre-Viking early settlement from Ireland [Nicolaisen 1970]. In fact, Galloway is the only area of known Irish colonization in western Britain to lack ogham inscriptions.
It is possible that the resemblance to ogham of the Lochnaw carving is coincidental and that the marks are those of a plough or the result of some natural process. From the appearance of the strokes, however, neither explanation seems particularly likely. If the lines are artificial, they might be no more than aimless doodles, though the probability of arriving at a design so like ogham simply by chance is surely remote. The eccentricity of the letters suggests that if the Lochnaw stone does bear ogham the carver was either innovative or inept.
The final possibility is that the lines were carved to look like ogham in more recent times, though not necessarily with the intention to deceive. Though ogham 'forgeries' are not uncommon in Ireland, where
APPENDIX: DUBIA
LOCHNAW
525
such monuments are much better known, the relative obscurity of the Scottish examples has meant that modem ogham inscriptions are rare, in contrast, for example, to modern runes [Barnes 1992]. One would imagine, however, that if the intention was to pass the stone off as the real thing, something closer to the ideal would have been produced. In its use of a drawn in stem Lochnaw is quite unlike the typical Irish ogham stone and though it is closer to a handful of Scottish examples all of these have been discovered since Lochnaw came to light. Further information regarding the circumstances of the discovery of the Lochnaw stone is needed before the various possibilities can be fully evaluated. Till then the carving resists convincing explanation.
BIBLIOGRAPHY
OS record card - NW 96 SE 26
Unpublished:
NMR MS A 62484/pe Derek Craig 1986 (Letter from Derek J. Craig, University of
Durham to Dr Anna Ritchie, 7 January 1986).
Alfred E. Truckell (1992) Letter to Jane Brann, Dumfries and Galloway Regional Archaeologist, 29.10.92 (in author's possession); (1993) Letters to author, 21.9.93, 28.9.93.
LOCHNAW Pillar fragment with incised carving
7 * •# *•
6 * #
# I
5 4 . . . . . u •* • **
3
I I
2
• **
* ' #
• *! * %
LOCHNAW Detail of carving with key to numbering
526
POOL
DISCOVERY The pin was discovered during the University of Bradford's excavation of the multi-period settlement at Pool, Sanday [see entry for Pool for details of site]. The pin is currently in the possession of the department of Archaeological Sciences, University of Bradford.
DESCRIPTION OF OBJECT Part of a bone pin incised with a circle-and-dot motif and lines which have been interpreted as ogham- or rune-like. Material:
Bone (species not identified)
Dimensions:
Length: 66mm, Width: 12mm (at head), 6mm (shaft)
Condition:
Fragmentary, top and bottom missing
DESCRIPTION OF CARVING I have not seen the pin personally. The following is based on the discussion and drawing of the pin in Holder, and thus remains tentative. In addition to the circle and dot there are two incised figures. The first consists of a short oblique line, about 7mm in length, intersected near its lower end by two short oblique lines to the right and one, joining just slightly higher, to the left. With a will, the whole could be interpreted as ogham LH, or BA, but the two 'letters' occupy less than the first third of the 'stem*. Holder likened the figure to the Scandinavian rune R, but this requires that the first line on the right be ignored. Nonetheless this suggestion should not be dismissed out of hand, though further information regarding the stratigraphic context of the pin is required.
The second figure, above the dot and circle, is incomplete, part being lost at the fracture at the top of the pin. The extant portion consists of a more horizontal line with two short perpendicular cross-strokes. This
APPENDIX: DUBIA
POOL
527
arrangement bears a greater resemblance to ogham and could be interpreted as a vowel or possibly as a member of the m-aicme. On balance however the resemblance of either figure to ogham is fairly slight and it is almost inevitable that an arrangement of overlapping straight lines will turn out as 'ogham-like' to those with sufficient imagination.
Holder described his initial reaction as being 'rather sceptical that these marks are ogam' [72]. I would have to agree. Holder went on to suggest that they might be an illiterate's attempt to produce something that looked like ogham. As I have discussed, I am doubtful of such explanations and, in the absence of more positive evidence to the contrary, take the lines as doodles.
BIBLIOGRAPHY Holder 1990:72-3 ill.
cm
POOL - Bone pin with ogham-like markings [after Holder 1990:72]
\
528
TOLLARD
HOUSE
Carvings alleged to be ogham were found in a small rock-shelter 60m south-east of Tollard House, Dunoon and Kilmun Parish, Argyll [NS 1341 6791]. There are two areas of carving on the sandstone walls of the shelter, the marks consisting of numerous, regularly-spaced incised lines. The carving is clearly artificial but its date and significance is unknown. Though smaller groups of roughly parallel lines are found elsewhere, e.g. Jonathan's Cave, Wemyss, and Eggeraess near Whithorn, the Tollard House examples are unprecedented in the whole of the 'extensive repenoire' of pre-historic rock an in Argyll [RCAHMS 1992:529]. It is clear, however, that they are not ogham.
BIBLIOGRAPHY OS record card - NS 16 NW 11 Morris 1981:160 ARG 84; RCAHMS 1992: 529 No.288.
529
V AIVOE
In a letter preserved in the archive of the NMR Mr John Stewan of Aberdeen refers to his discovery of a possible ogham inscription on the island of Whalsay, Shetland. The discovery was made during a visit to Vaivoe in May 1951 (NGR HU 575 665), but no details regarding the precise location or context of the find are given. Stewan refers to the item as an 'ogam amulet', which implies that it was small, but gives no measurements, nor does he specify from what material it was made. He read the text as THADHM with two crosses and 'what may be M' 'in another corner'. At the time of writing the letter, the item was still in Stewart's possession. It appears that he had sent photographs of it and the Papil rune stone to Stuart Cruden. The current location of both amulet and photographs is unknown. Stewan refers to informing The Scotsman newspaper of his discovery, but the paper does not appear to have picked up on the story. No-one seems to havefollowedup Stewart's information at the time, though it is unclear whether this can be taken as an indication that his 'ogham' identification was thought to be unfounded. Apparently, Mr Stewan was a knowledgeable amateur archaeologist. Elsewhere in the letter he mentions discovering a new broch site north of Lerwick and a rune-inscribed stone from Papil, now in Shetland Museum (Barnes SH 4). His testimony should not be dismissed out of hand, particularly since examples such as the Buckquoy spindle-whorl, itself possibly with amuletic function, show that small portable objects were inscribed with ogham. Nonetheless, accidental scores or abstract linear decoration are occasionally mistaken for possible oghams, and without further information it would be unwise to admit the Vaivoe 'ogham' to the corpus.
BIBLIOGRAPHY OS Record Card - HU 56 NE 20 Unpublished:
NMR MS/462 - LetterfromJohn Stewart to Stuart Cruden, 11 January 1952, Aberdeen.
530
OTHERS
In an article in Scottish Gaelic Studies in 1931, Francis Diack sets out his theory that a prehistoric writing system was the model from which the ogham alphabet was created, under the influence of Latin. Diack claims to have found a 'great quantity' of examples of this writing, which he labels 'proto-ogam'. He lists, and in some cases, illustrates, carvings from: Kirkhill, Inverness; Findon, and the parishes of Strachan, Banchory-Devenick, and Maryculter, Kincardineshire; Croy, Nairn; and Drumfours (Cushnie), Peterculter, Drumoak, Echt, and Irelandbrae (Rayne) [see also 1925:269], all Aberdeenshire; Yarrow, Selkirkshire; Island of Gigha, Argyll, and Milingavie. He mentions also examples from Ireland and Wales. From his illustrations it appear that these carvings are sharpening marks or similar, and certainly they are nothing to do with ogham. His claims regarding parallel lines carved in the Fife Caves 'bearing an essential resemblence' to ogham can also be dismissed [1925:269]. From personal inspection it is clear to me that, though artificial these are not ogham.
531
BIBLIOGRAPHY OF WORKS CITED n.a. (1855a) [notice of proceedings at the meetings of the Archaeological Institute, 4 May 1855], Archaeological Journal 12, 275. , (1855b) 'Antiquarian researches [notes on meeting of the Archaeological Institute]', Gentleman*s Magazine, Vol. 44, July 1855, 80-1. — , (1865) 'Donations to and purchases for the Museum [Bressay stone]' PSAS 5 (1862-3, 1863-4), 23940 pl.xi-xii — , (1866) 'Donations to and purchases for the Museum [Scoonie]', PSAS6 (1864-66, publ. 1868), 401. — , (1878) 'Donations to and purchases for the Museum [St Ninian's Isle]', PSAS 12 (1876-8), 11. , (1922) 'Donations to the Museum [Bac Mhic Connain]', PSAS 55 (1921-2), 12-13. , (1923) 'Donations to the Museum [Bac Mhic Connain]', PSAS 57 (1922-3), 110. — , (1929) 'Donations to the Museum [Bac Mhic Connain]', PSAS 63 (1928-9), 363. , (1946) [note of communications read at meetings of the society], PSAS 80 (1945-6), 164. — , (1974) [note on removal of Aboyne ogham stone], Discovery and Excavation in Scotland 1974, p.5, The Scottish Regional Group, Council for British Archaeology, Edinburgh. , (1984) Archaeology Extra 1, Spring 1984, Bradford: Undergraduate School of Archaeology, University of Bradford. — , (1985) 'Ogam inscription', Archaeology Extra 2, Spring 1985, Bradford: Undergraduate School of Archaeology, University of Bradford, 2. — , (1986) 'Symbol stone', Archaeology Extra 3, Spring 1986, Bradford: Undergraduate School of Archaeology, University of Bradford, 2. Aberdeen Breviary = see Laing 1854 AHLQVIST, Anders (1992) The Early Irish Linguist, Commentationes Humanamm Literaranim 73, Finnish Society of Sciences and Letters. AITCHISON, N. B. (1994) Armagh and the Royal Centres in Early Medieval Ireland: Monuments, Cosmology, and the Past, Woodbridge, Suffolk: Cruithne Press/Boydell & Brewer. ALCOCK, Elizabeth A. (1991) 'Pictish Stones Class I: Where and How?', Glasgow Archaeological Journal 15 (1988-9), 1-21. , (1988) 'Appendix: Enclosed places AD 500*800', in DRISCOLL & NIEKE, 40-46. ALCOCK, Leslie (1970) 'Was there an Irish Sea culture-province in the Dark Ages?', in The Irish Sea Province in Archaeology and History, ed. D. Moore, Cardiff. — , (1981) 'Early historic Fortifications in Scotland', in G. GUILBERT (1981) Hill-Fort Studies:Essays for A. H. A. Hogg, Leicester, 150-80. — , (1982) 'Forteviot: a Pictish and Scottish royal church and palace', in The Early Church in Western Britain and Ireland, ed. Susan M. Pearce, Oxford, 211-39. — , (1988) 'The Activities of Potentates in Celtic Britain, AD 500-800: A Positivist Approach', ('Appendix: Enclosed places AD 500-800', by Elizabeth A Alcock 40-46), in DRISCOLL & NIEKE, 2246. — , & Elizabeth E. (1992) 'Reconnaissance excavations on Early Historic fortifications and other royal sites in Scotland, 1974-84; 5: A, Excavations and other fieldwork at Forteviot, Perthsire, 1981; B, Excavations at Urquhart Castle, Inverness-shire, 1983; C, Excavations at Dunottar, Kincardineshire, 1984', PSAS 122, 215-287. , (1993) 'Image and icon in Pictish sculpture', in SPEARMAN & HIGGITT, 230-6.
BIBLIOGRAPHY
532
ALEXANDER, W. (1952) Place-Names of Aberdeenshire, Spalding Club, Aberdeen. ALLEN, J. Romilly (1892) 'Report on the Sculptured Stones older than A.D. 1100, with Symbols and Celtic Ornament, in the District of Scotland south of the River Dee; obtained under the Victoria Jubilee Gift of His Excellency Dr. R. H. Gunning, F.S.A.Scot.', PSAS 26 (1891-2), 251-59. , & Joseph Anderson (1903) The Early Christian Monuments of Scotland (repr. with introduction by Isabel Henderson, 1994 Forfar: Pinkfoot Press). ANDERSON, Alan Orr (1922) Early Sources of Scottish History A.D. 500-1286, 2 Vols., Stamford: Paul Watkins (revised edition by Marjorie Anderson 1980). , & Marjorie O. ANDERSON (1961) (ed. & transi.) Adomnán 's Life ofColumba, Edinburgh: Thomas Nelson (Revised edition, Oxford Medieval Texts, Oxford: Clarendon = M. Anderson 1991) f= VQ. ANDERSON, Joseph (1881) Scotland in Early Christian Times (2nd ser.) (The Rhind Lectures for 1880), Edinburgh iDavk1 Douglas. — , (1897) 'Notices of some recently-discovered Inscribed and Sculptured Stones', PSAS 31 (1896-7), 293-308. , ( 1904) 'Notices of a sculptured stone with ogam inscription from Latheron, presented to the National Museum by Sir Francis Tress Barry, Ban. M.P. Hon. F.S.A.Scot., Keiss Castle, Caithness; and of two Sculptured stones recently discovered by Rev. D. Macrae, B.D. at Edderton, Ross-shire', PSAS 38 (19034), 534-41. ANDERSON, Marjorie O. (1973) Kings and Kingship in Early Scotland, Edinburgh: Scottish Academic Press (revised edition 1980). , (1991) see A. O. & M. O. Anderson 1961. ARMIT, Ian (1990a) 'Later Prehistoric Settlement in the Western Isles of Scotland', unpublished PhD dissertation, University of Edinburgh, ( = British Archaeological Répons, British Series 221). , (1990b) (ed.) Beyond the Brochs: the Later Iron Age in Atlantic Scotland Edinburgh U.P. , (1990c) 'Brochs and Beyond in the Western Isles', in ARMIT (1990b) 41-70. , (1990d) 'Epilogue: The Atlantic Scottish Iron Age', in ARMIT (1990b) 194-210. — , (1991) 'The Atlantic Scottish Iron Age: five levels of chronology', PSAS 121, 181-214. ASHMORE, Patrick (1980) 'Low cairns, long cists and symbol stones', PSAS 110 (1978-80), 346-355. BALFOUR, J. A. (1910) 'The King's Cave', in The Book of Arran, Vol.1, ed. J. A. Balfour, Glasgow: The Arran Society of Glasgow, 213-18, pl.33-34. BANNERMAN, John (1974) Studies in the History ofDalriada, Edinburgh: Scottish Academic Press. BANNERMAN, William (1908) 'On the ogham inscription of the Lunnasting and Golspie Stones', PSAS 42 (1907-8), 342-52. BARNES, Michael (1992) 'Towards an edition of the Scandinavian Runic Inscriptions of the British Isles: Some Thoughts', Northern Studies: The Journal of the Scottish Society for Northern Studies 29, 32-42. BARROW, Geoffrey W. S. (1973) The Kingdom of the Scots: Government, Church and Society from the eleventh to the fourteenth century, London: Edward Arnold. — , (1983) 'The childhood of Scottish Christianity: a note on some place-name evidence', Scottish Studies 27, 1-15.
BIBLIOGRAPHY
533
BARTRUM, Peter C. (1966) Early Welsh Genealogical Tracts, Cardiff: Univ. Wales Press. BATEY, Colleen E. (1991) 'Picts And Vikings In Caithness And Sutherland: A Resume', in Sadies in Insular Art and Archaeology (American Early Medieval Studies 1), ed. Catherine Karkov & Robert Fanell, Miami, Ohio, 49-58. — , Judith Jesch, & Christopher D. Morris (1993) (ed.) The Viking Age in Caithness, Orkney and the North Atlantic: Select Papers from the Proceedings of the Eleventh Viking Congress, Thurso and Kirkwall, 22 August - I September 1989, Edinburgh U.P. BEST, Richard I., Osborn Bergin, & M. A. O'Brien (1954-67) (ed.) The Book ofLeinster, Dublin, [refs. to line number] BEVERIDGE, Erskine (1911) North Uist: Its Archaeology and Topography, with Notes upon the Early History of the Hebrides, Edinburgh. , & J. G. Callander (1931) 'Excavations of an Earth House at Foshigarry and a fort, Dun Thomaidh, in North Uist', PSAS 65 (1930-31), 299-357. , (1932) 'Earth-houses at Garry lochdrach and Bac Mhic Connain in North Uist', PSAS 66 (1931-32), 32-66. BIGELOW, Gerald F. (1992) 'Issues and Prospects in Shetland Norse Archaeology', in Norse and Later Settlement and Subsistence in the North Atlantic, ed. Christopher D. Morris & D. James Rackham, University of Glasgow, Department of Archaeology, 9-32. BLACK, George Fraser (1946) Tlie Surnames of Scotland: 7\\eir Origin, Meaning, and Histoty, New York Public Library (reprint 1993, Edinburgh: Birlinn). BLAIR, John & Richard Sharpe (1992a) (ed.) Pastoral Care Before the Parish, Studies in the Early History of Britain, Leicester: Leicester U.P. , (1992b) Introduction, in BLAIR & SHARPE (1992a), 1-10. BOURKE, Cormac (1983) 'Hand-bells of the Early Scottish Church', PSAS 113, 464-8. BRADLEY, Richard (1987) 'Time regained: the creation of continuity', Journal of tlie British Archaeological Society 140, 1-17. , (1993) Altering the Earth: The Origins of Monuments in Britain and Continental Europe (The 1992 Rhind Lectures), Society of Antiquaries of Scotland Monograph Series 8, Edinburgh. BRASH, Richard Rolt (1874) 'Remarks on the ogham inscription of the Newton Pillar-stone', PSAS 10 (1872-4), 134-41. , (1879) The Ogam Inscribed Monuments of the Gaedhil in the British Islands (ed. George M. Atkinson), London. BREATNACH, Liam (1984) 'Canon law and secular law in early Ireland: the significance of Bretha Nemed', Peritia 3, 439-59. BROUN, Dauvit (1995) The Charters of Gaelic Scotland and Ireland in the Early and Central Middle Ages, (Quiggin Pamphlets on the Sources of Mediaeval Gaelic History 2), Cambridge: Dept. of AnglgoSaxon, Norse, and Celtic, University of Cambridge.
BIBLIOGRAPHY
534
BROWN, T. Julian (1993) 'St Ninian's Isle Silver Hoard: the Inscriptions', in A Palaeographer's view: the Selected writings of Julian Brown, ed Janet Bately, Michelle Brown & Jane Roberts, London: Harv • Miller Publishers, 245-51. [Originally published as pan of a joint paper with A. C. O'Dell, R. B. K. Stevenson, H. J. Penderleith, and R. L. S. Bnice-Mitford, m Antiquity 33 (1959), 250-55.] BROWNE, George F. (1891) 'On a Sculptured Stone with an Ogam Inscription', Proceedings of the Cambridge Antiquarian Society 1, 88. , (1921) On some antiquities in the Näghbourhood ofDunecht House, Aberdeenshire, Cambridge U.P. BURNS, Agnes Joan (1981) [= Siobhán de hoir] Scholastic Oghams, National University of Ireland (University College, Dublin), unpublished MPhil. thesis. , (1983) [as Siobhán de hoir] 'The Mount Callan Ogham stone and its context' North Munster Antiquarian Journal 25, 43-57. BUTLER, Dugald (1895) Letter to the editor The Scotsman Wednesday 30 January 1895. , (1897) 77?e Ancient Church and Parish of Abernethy: An Historical Smdy, Edinburgh: William Blackwood. BYRNE, Francis John (1973) Irish Kings and High Kings, London: Batsford. CALDER, George (1917) (ed.) Auraicept na n-Éces, Edinburgh: John Grant. CALDER, William M. & Kenneth H. Jackson (1957) *An inscription from Altyre', PSAS 90, 246-50. CAMPBELL, Ewan (1987) 'A cross-marked quern from Dunadd and other evidence for relations between Dunadd and Iona', PSAS 117, 105-17. , & Alan Lane (1993) 'Celtic and Germanic Interaction in Dalriada: the 7th-century Metalworking Site at Dunadd', in SPEARMAN & HIGGIT (1993), 52-63. CAMPBELL, Marian (1976) 'Rock-carvings at Dunadd', The Kist: The Magazine of the Natural History and Antiquarian Society of Mid-Argyll 12, 1-7. , and M. Sandeman (1964) 'Mid-Argyll: An Archaeological Survey', PSAS 95 (1961-2), 1-125. CANT, Ronald (1984) 'Settlement, Society and Church Organisation in the Northern Isles', in FENTON & PÁLSSON, 169-79. CARRINGTON, Ann (1995) 'The equestrian motif in the early medieval Pictish sculpture at Meigle, Perthshire', Journal of the Pictish Arts Society 8 (Autumn 1995), 28-43. CHARLES-EDWARDS, Thomas (1971) 'Some Celtic kinship terms', Bulletin of the Board of Celtic Studies 24, 105-22. , (1993) Early Irish and Welsh Kinship, Oxford: Clarendon. CHARLTON, Edward (1855a) 'On the ogham inscription from the Island of Bressay, Shetland, 1853', Archaeologia Aeliana, first series 4 (1854), 150-156. , (1855b) ' "Papae" of Iceland and Orkney', Notesand Queries 11 (14 April 1855), 181-2 [No. 285 p.285]. CHILDE, V. Gordon (1961) Guide to the Ancient Monuments 6: Shetland, Edinburgh: HMSO.
BIBLIOGRAPHY
535
CHRISTISON, David & Joseph Anderson (1905) 'Report on the Society's excavations of forts on the Poltalloch Estate, Argyll 1904-5', PSAS 104-5 (1904-5), 259-322. CLANCY, Thomas Owen (1993) 'The Drosten Stone: A New Reading', PSAS 123, 345-53. , & Gilbert Markus (1995) lona: The Earliest Poetry of a Celtic Monastery, Edinburgh U. P. CLARKE, David V. (1970) 'Bone Dice in the Scottish Iron Age', Proceedings of the Prehistoric Society 36 (n.s), Cambridge, 214-32. , (1990) 'The National Museums' stained-glass window', PSAS 120, 201-224. CLARKE, Roy Rainbird (1952) 'A Knife handle from Weeting, Norfolk', Antiquaries Journal 32, 71 pl.23a. CLOSE-BROOKS, Johanna (1975) 'A Pictish pin from Golspie, Sutherland', PSAS 106, 208-10. , (1980) 'Excavations in the Dairy Park, Dunrobin, Sutherland, 1977', PSAS 110 (1978-80), 328-45. , (1981) St. Ninian's Isle Treasure (National Museum of Antiquities), Edinburgh: HMSO. , (1984) 'Pictish and other burials', in FRIELL & WATSON (1984), 87-114. , & R. B. K. Stevenson (1982) Dark Age Sculpture, Edinburgh: HMSO. , (1986) Exploring Scotland's Heritage: The Highlands (RCAHMS), Edinburgh: HMSO. , (1989) Pictish Stones in Dunrobin Castle Museum, (The Sutherland Trust), Derby: Pilgrim Press. CLOUSTON, J. R. (1937) 'The Aikemess Stone1, Proceedings of the Orkney Archaeological Societv 14 (1936-7), 9-19. COLES, Fred R. (1901) 'Report on the stone circles of the North-East of Scotland, Inverurie District, obtained under the Gunning Fellowship, with measured Plans and drawings', PSAS35 (1900-1), 187-248. COLLINGWOOD, R. G. & R. P. WRIGHT (1965) The Roman Inscriptions of Britain, Vol J: Inscriptions on Stone, Oxford: Clarendon. COLLINS, G.H. (1977) 'Chalk spindle-whorls from Buckquoy, Orkney', PSAS 108 (1976-7), 222-3. [= Appendix 7 of Ritchie 1977] CORDINER, Rev. Charles (1780) Antiquities and Scenery of the North cf Scotland, in a series of letters to Thomas Pennant, London. , (1788) Remarkable Ruins of North Britain, London. COWAN, Edward J. (1993) 'The Historical MacBeth', in Moray: Province and People, ed. W. D. H. Sellar, Edinburgh: Scottish Society for Northern Research, 117-41. COWIE, Trevor (1980) 'Excavations at the Catstane, Midlothian 1977' PSAS 109, (1977-8), 166-201. COX, Richard A. V. (1991) 'Norse-Gaelic contact in the west of Lewis: the place-name evidence', in Language Contact in the British Isles: Proceedings of the Eighth International Symposium on Language Contact in Europe, Douglas, Isle of Man, 1988, (Linguistische Arbeiten 238), ed. Sture P. Ureland & George Broderick, Tübingen: Max Niemeyer Verlag, 479-494. CRAMP, Rosemary ( 1984) 'General Introduction to the Series', in Corpus cfAnglo-Saxon Stone Sculpture, Volume I Part I, County Durham and Northumbria, British Academy, Oxford U,P., ix-li.
BIBLIOGRAPHY
536
CRAW, J. Hewat (1929) 'On a Jet Necklace from a Cist at Poltalloch, Argyll', PSAS 63 (1928-9), 154189. — , (1930) 'Excavations at Dunadd and at other sites on the Poltalloch Estates, Argyll*, PSAS 64 (192930), 111-146. — , (1932) Two Long Caiins (One Homed) and an Ogham Inscription, near Poltalloch, Argyll', PSAS 66 (1931-2), 445-450. CRAWFORD, Barbara E. (1987) Scandinavian Scotland (Studies in the Early History of Britain, Scotland in the Early Middle Ages 2), Leicester U.P. , (1995) Earl and Mormaer: Norse-Pictish relationships in Northern Scotland (Groam House lecture 6), Rosemarkie: Groam House Museum. CRAWFORD, Iain A. (1974) 'Scot (?), Norseman and Gael', Scottish Archaeological Forum 6, 1-16. — , & R. Switsur (1977) 'Sandscaping and C14: the Udal, N. Uist', Antiquity 51, 124-36. CREGEEN, Eric (1960) [note on Poltalloch], Discovery and Excavation in Scotland 1960, Edinburgh, 10-
11. — , (1961) [note on Poltalloch], Discovery and Excavation in Scotland 1961, Edinburgh, 10-12. — , (1961) [note on Poltalloch], Discovery and Excavation in Scotland 1962, Edinburgh. CRUDEN, Stewart H. (1965) 'Excavations at Birsay, Orkney', The Fourth Viking Congress, York August 1961, ed. Alan Small, (Aberdeen University Studies 149), Edinburgh, 22-31. CUPPAGE, Judith et al. (1986) Archaeological Survey of the Dingle Peninsula, Ballyferriter: Oidhreacht Chorea Dhuibhne. CURLE, Cecil L. (1940) 'The Chronology of the Early Christian Monuments of Scotland', PSAS 74 (1939-40), 60-116. — , (1982) Pictish and Norsefindsfromthe Brough of Birsay 1934-74 (Society of Antiquaries of Scotland Monograph Series 1), Edinburgh. , (1983) 'The Finds from the Brough of Birsay 1934-1974', in THOMSON, 67-81. — , & Françoise Henry (1943) 'Early Christian An in Scotland*, Gazette des Beaux-Arts 24n.s., 257-72. CURLE, J. (1932) 'An Inventory of Objects of Roman and Provincial Roman Origin found on Sites in Scotland not definately associated with Roman Constructions', PSAS 66 (1931-2), 277-397. DALLAND, Magnar (1992) 'Long cist burials at Four Winds, Longniddry, East Lothian', PSAS 122, 197206. DAVIDSON, J. L. (1986) 'The collection of antiquarian information for the early Ordnance Survey maps of Scotland', PSAS 116 (1986), 11-16. DAVIDSON, John (1878) Inverurie and the Earldom of the Garioch, Edinburgh and Aberdeen. DA VIES, Wendy (1979) The Uandajf Charters, Aberystwyth: National Library of Wales. — , (1982) Wales in the Early Middle Ages (Studies in the Early History of Britain), Leicester U.P. de hÓIR, Siobhán see BURNS
BIBLIOGRAPHY
537
DIACK, Francis C. (1922) The Newton Stone and other Pictish Inscriptions, Paisley: Alexander Gardner. — , (1925) The Old-Celtic inscribed and sculptured stone at Auquhollie, Kincardineshire, and ogam in Scotland', PSAS 59 (1924-5), 257-69. , (1926) 'A Pre-Dalriadic Inscription of Argyll: The Gigha Stone', Scottish Gaelic Studies 1, 3-16. , (1931) The origin of the ogam alphabet', Scottish Gaelic Studies 3, 86-91. , (1944) The Inscriptions of Pialand: an essay on the sculptured and inscribed stones of the north-east and north cf Scotland: with other writings and collections, ed. W. M Alexander & J. MacDonald, Aberdeen: Third Spalding Club. DOBBS, Margaret E. (1949) 'Cé: The Pictish Name of a District in Eastern Scotland', Scottish Gaelic Studies 6, 137-8. DORIAN, Nancy C. (1981) Language Death: The Life Cycle of a Scottish Gaelic Dialect, Philadelphia: Univ. Pennsylvania Press. DOR WARD, David (1995) Scottish Surnames, Glasgow: HarperCollins. DOUGLAS, Kalene (1988) [note on Gigha 2], Discovery and Excavation in Scotland 1988, Edinburgh, 20. DRISCOLL, Stephen T. (1991) 'The Archaeology of State Formation in Scotland', in Scottish Archaeology: New Perceptions, ed. William S. Hanson & Elizabeth Slater, Aberdeen U.P., 81-111. — , & Margaret R. Nieke (1988) (ed.) Power and Politics in Early Medieval Britain and Ireland* Edinburgh: EUP. DUM VILLE, David N. (1976) 'A Note on the Picts in Orkney', Scottish Gaelic Studies, 12 part 2, 266. DUNCAN, A. A. M. (1957) 'Documents relating to the priory of the Isle of May c. 1140-1313', PSAS 90 (1956-7), 52-80. EDWARDS, Arthur J. H. (1926) 'Excavation of a number of graves in a mound at Ackergill, Caithness', PSAS 60 (1925-6), 160-79. , (1927) 'Excavation of graves at Ackergill and of an earth house at Freswick Links, Caithness, and a discovery of a Viking grave at Reay, Caithness', PSAS 61 (1926-7), 196-209. ELLIOT, R. W. V. (1959) Runes, Manchester U.P. EVANS, D. Ellis (1967) Gaulish Personal-Names: A Study of Some Continental Formations, Oxford: Clarendon. EVANS, D. Simon (1964) A Grammar of Middle Welsh, (Mediaeval and Modem Welsh Series, Supplementary Volume), (reprinted 1989), Dublin Institute for Advanced Studies. FARRELL, Roben (1982) (ed.) The Vikings, London: Phillimore. FENTON, Alexander & Hermann Pálsson (1984) (ed.) The Northern and Western Isles in the Viking World: Survival, Continuity and Change, Edinburgh: John Donald. FERGUSON, Samuel (1887) Ogham Inscriptions in Ireland, Wales and Scotland, Edinburgh.
BIBLIOGRAPHY
538
FERGUSON, William (1991) 'George Buchanan and the Picts\ Scottish Tradition 16 (1990-1), 18-32. FERNIE, Eric (1986) 'Early church architecture in Scotland', PSAS 116 (1986, publ. 1988), 393-411. FOJUT, Noel (1993) The Brochs ofGurness and Midhowe, (Historic Scotland Guide book), Edinburgh: HMSO. FORBES, Alexander Penrose (1872) (ed.) Kaiendars of Scottish Saints, Edinburgh. FORSYTH, Katherine (1996a) 'The Inscriptions on the Dupplin Cross', in From the Isles of the North: Medieval Art In Ireland and Britain (Proceedings of the Third International Conference on Insular Art), ed. Cormac Bourke (Belfast: HMSO, 1996), 237-44. — , (1996b) 'Symbols: the third Pictish writing system ?', in Festschrift for Isabel Henderson, ed. David Henry, Balgavies, Forfar: Pinkfoot Press (forthcoming). , (1996c) Language in Pialand: The Case Against Non-Indo-European Pictish (A. G. van Hamel lecture (1995), Utrecht: Stichting A. G. van Hamel voor keltische Studies (forthcoming). — , (1996d) 'Literacy in Pictland', in Literacy in Medieval Celtic Societies, ed. Huw Pryce, Cambridge U.P. (forthcoming). FOSTER, Sally (1990) 'Pins, Combs and the Chronology of Later Atlantic Iron Age Settlement', in Beyond the Brochs: the Later Iron Age in Atlantic Scotland, ed. Ian Armit, Edinburgh U.P., 143-74. — , (1996) Picts, Gaels and Scots: Early Historic Scotland, London: Batsford/Historic Scotland. FRASER, Rev. William (1793) 'Parish of Gigha and Cara', in OSA. FRIELL, J. G. P. & W. G. Watson (1984) (ed.) Pictish Studies: Settlement, Burial and Art in Dark Age Northern Britain, (British Archaeological Répons, British Series 125), Oxford. GERCHO W, Jan ( 1988) Die Gedenküberlieferung der Angelsachsen (Arbeiten zur Frühmittelalterforschung 20), Berlin: Walter de Gruyter. GORDON, C. A. (1956) 'Carving Technique on the symbol stones of North-East Scotland', PSASS6, 4060. GOUDIE, Gilbert (1878) 'On Two Monumental Stones with Ogham Inscriptions Recently Discovered in Shetland', PSAS 12 (1876-78, publ. 1879) 20-32. , (1879) ' On Rune-Inscribed Relics of the Norsemen in Shedand' PSAS 13 (1878-79), 136-64. — , (1881) 'Notice of aScultpured Slabfromthe Island of Burra, Shetland' PSAS 15 (1880-81), 199-209. — , (1904) The Celtic and Scandinavian Antiquities of Shetland, Edinburgh: William Blackwood. GRAHAM-CAMPBELL, James (1973) 'A ninth century Anglo-Saxon horn-mount from Burghead, Morayshire', Medieval Archaeology 17, 43-51. — , (1991) 'Nome's Law, Fife: on the nature and dating of the silver hoard', PSAS 121, 241-59. GRANT, Alexander (1993) 'Thanes and thanages, from the eleventh to the fourteenth centuries', in Crown, Lord and Community. Essavs presented to G. W. 5. Barrow, ed. Alexander Grant & Keith J. Stringer, Edinburgh U. P., 39-81.
BIBLIOGRAPHY
539
GRAVES, Charles (1855) [note on the Bressay stone] Gentleman's Magazine July 1855 80-1. — , (1858) 'On the ogham inscriptions found on monument at Bressay, Shetland [paper read at meeting of the Academy, 28 May 1855] ', Proceedings of the Royal Irish Academy 6 (1853-7), 248-9. GREENE, David ( 1977) 'Archaic Irish ', in Indogermanisch und Keltisch: Kolloquium der Indogermaischen Gesellschaft am 16. und 17. Februar 1976 in Bonn, ed. Karl Horst Schmidt, Wiesbaden: Ludwig Reichert, 11-33. HALLEN, Ywonne ( 1994) 'The use of bone and antler at Foshigarry and Bac Mhic Connain, two Iron Age sites on North Uist, Western Isles', PSAS 124, 189-231. HAMP, Eric (1954) [review of I. J. Gelb (1952) A Study of Writing: the foundations of grammatology) Zatschriftftir celtische Philologie 24, 308-12. — , (1986) 'Scottish Gaelic moraif, Scottish Gaelic Studies 14, 138-41. HARBISON, Peter (1991) Pilgrimage in Ireland: The Monuments and the People, London: Barrie & Jenkins. — , (1992) The High Crosses of Ireland: An Iconographie and Photographic Survey (RömischGermanisches Zentralmuseum, Forschungsinstitut für Vor- und Frühgeschichte, Band 17), Bonn: Habelt. — , (1994) 'The extent of Royal Patronage on Irish High Crosses', Stadia Celtica Japonca (New Series) 6, 77-105. HARDEN, Gillian (1986) 'Another part of a Pictish symbol stone from Little Ferry Links, Golspie, Sutherland', PSAS 116, 573-77. HARVEY, Anthony (1987) 'The Ogam inscriptions and their geminate consonant symbols', Ériu 38, 4571. — , (1989) 'Some significant points of Early Insular Celtic orthography', in Sages, Saints and Storytellers: Celtic Studies in Honour of Professor James Carney (Maynooth Monographs 2), éd. Donnchadh Ó Corráin, Liam Breatnach, & Kim McCone, Maynooth: An Sagart, 56-66. — , (1990a) 'Notes on Old Irish and Old Welsh Consonantal Spelling' in Celtic Linguistics: Readings in the Brythonic Languages, festschrift for T. Arwyn Watkins (ed.) Martin J BALL et al Amsterdam. — , (1990b) 'Retrieving the Pronunciation of early Insular Celtic scribes: Towards a Methodoology' Celtica 21, 178-90. , (1991) 'Retrieving the Pronunciation of early Insular Celtic scribes: The Case of Dorbbene' Celtica 22, 48-63. HAY, G. (1957) The Church of the Three Holy Brethren at Lochgoilhead. HEDGES, John W. (1987) Bu, Gurness and the Brochs of Orkney, 3 Vols., (British Archaeological Reports British Series 163-5), Oxford. — , (1990) 'Surveying the foundations: Life after "Brochs"', in Beyond the Brochs: the Later Iron Age in Atlantic Scotland, ed. Ian Armit, Edinburgh U.P., 17-31. HENDERSON, Isabel (1958) 'The Origin Centre of the Pictish Symbol Stones', PSAS 91, 44-60. , (1967) The Pias, (Ancient People and Places), London: Thames & Hudson. — , (1972a) 'The Picts of Aberdeenshire and their Monuments', Archaeological Journal 129, 166-74. — , (1972b) 'Pictish Sculptured stones', Archaeological Journal 129, 198-90 — , (1978) 'Sculpture north of the Forth after the take-over by the Scots', in Anglo-Saxon and Viking Sadpture: Papers from the Collingwood Symposium on insular sculpture from 800-1066 (BAR British Series 49), ed. James Land, Oxford, 47-73.
BIBLIOGRAPHY
540
HENDERSON, Isabel (1982) 'Pictish An and the Book of Kells\ in WHITELOCKet al. (1982), 79-105. , (1983) 'Pictish Vine-Scroll Ornament', in From the Stone-Age to the 'Forty-Five: Essays presented to R. B. K. Stevenson, ed Anne O'Connor & David V. Clarke, Edinburgh, 243-68. , (1986) 'The "David Cycle" in Pictish An', in Early Medieval Sculpture in Britain and Ireland (British Archaeological Reports, British Series 152), 87-123. , ( 1987) 'Early Christian Monuments of Scotland Displaying Crosses but no other ornament', The Pias - A New Look at Old Problems, ed. Alan Small, Dundee, 45-58. , (1993) 'The shape and decoration of the cross on Pictish cross-slabs carved in relief, in SPEARMAN & HIGGITT, 209-218. , (1994) 'The Insular and Continental Context of the St Andrews Sarcophagus' in Scotland in Dark Age Europe (St John's House Occasional Papers 5), (ed.) Barbara E. Crawford, St Andrews: The Committee for Dark Age Studies, University of St Andrews, 71-102. HENNESSEY, W. M. (1866) (ed.) Chronicon Scottorum Rolls series 46, London. HERITY, Michael (1993) 'The forms of the tomb-shrine of the founder saint in Ireland', in SPEARMAN & HIGGITT, 188-95. HICKS, Carola (1993) 'The Pictish Class I Animals', in SPEARMAN & HIGGIT, 196-202. HIGGINS, James G. (1987) The Early Christian Cross Slabs, Pillar Stones and Related Monuments of County Galwayt Ireland (British Archaeological Reports, International Series 375) 2 vols. HIGGITT, John (1982) 'The Pictish Latin Inscription at Tarbat in Ross-shire', PSAS 112, 300-21. , (1986) 'Words and Crosses: The Inscribed Stone Cross in Early Medieval Britain and Ireland', in Early Medieval Sculpture in Britain and Ireland (BAR British Series 152), ed. John Higgitt, Oxford, 12552. ' HOGAN, Edmund (1910) Onomasticon Goidelicum. Locorum ettribuum Hiberniae et Scotiae: An Index, with Identifications, to the Gaelic Names of Places and Tribes, (repr. 1993, Dublin: Four Courts Press). HOLDER, Nick (1990) 'Aspects of Ogam with Particular Reference to the Ogam-Inscribed Portable Objects', Unpublished M.A. dissertation, University of London. HULL, Vernon (1941), 'The Exile of Conall Core', Proceedings of the Modern Languages Association 56, 937-50. , (1947) 'Conall Core and the Corco Luigde', Proceedings of the Modern Languages Association 62, 887-909. HUNTER, John R. (1985) [notice of discovery of Pool ogham-inscribed stone], Discovery and Excavation in Scotland 1985, Edinburgh, 66. , (1986) Rescue Excavations on the Brough ofBirsay 1974-82 (Society of Antiquaries of Scotland Monograph Series 4), Edinburgh. — , (1990) 'Pool, Sanday: a Case Study for the Late Iron age and Viking Periods', ïnBeyondthe Brochs: Changing Perspectives on the Later Iron Age in Atlantic Scotland, ed. Ian Annit, Edinburgh U.P., 175-93. INGLIS, Jim (1987) 'Saints and symbols in Cromar', TheDeeside Field 19, (Deeside Field Club), 146-53. JACKSON, Anthony (1984) The Symbol Stones of Scotland: A social anthropological resolution of the problem of the Pias, Stromness: The Orkney Press.
BIBLIOGRAPHY
541
JACKSON, Kenneth H. (1953) Language and History in Early Britain, Edinburgh UP. , (1955) The Pictish Language', in The Problem of the Pias, ed. F. T. Wainwright, Edinburgh: Nelson (repr. 1980, Perth: Melven Press), 129-160, Appendix I 161-6 (1980). , (1960) The St Ninian's Isle Inscription: A Re-appraisal1, Antiquity 34 (1960), 38-42. , (1965) 'The Ogam Inscription at Dunadd', Antiquity 39, 300-302. , (1972) The Gaelic Notes in the Book of Deer (Osborn Bergin Memorial Lecture 1970), Cambridge U.P. , (1973a) 'An Ogam Inscription near Blackwaterfoot', Antiquity 47 (1973), 53-54 pl.ix. , (1973b) The Inscriptions', in St Ninian's Isle and its Treasure, ed. Alan Small, Charles Thomas, & David M. Wilson, Oxford U.P., 167-73. , (1977) 'The ogam inscription on the spindle whorl from Buckquoy, Orkney', PSAS 108 (1976-7), 221-2. [= Appendix 6 of Ritchie 1977] , (1983) 'Ogam stones and early Christian Latin inscriptions', in The Companion to Gaelic Scotland,ed. Derick S. Thomson, Oxford: Blackwell, 220-1. JENKINS, Dafydd & Morfydd E. Owen (1983) 'The Welsh marginalia in the Lichfield Gospeis. Pan I', Cambridge Medieval Celtic Studies 5, 37-65. JESCH, Judith (1991) Women in the Viking Age: Woodbridge, Suffolk: Boydell. JONES, Barri, Ian Keillar & Keith Maude (1993) 'The Moray aerial survey: discovering the prehistoric and proto-historic landscape', in Moray: Province and People, ed. W. D. H. Sellar, Edinburgh: Scottish Society for Northern Research, 47-74. KARKOV, Catherine & Roben Farrell (1991) (ed.) Studies in Insular Art and Archaeology (American Early Medieval Studies 1), Miami. KELLY, Dorothy (1993) 'The Relationships of the Crosses of Argyll: the Evidence of Form', in SPEARMAN & HIGGITT, 219-29. KERMODE, Philip M. C. (1907) Manx Crosses, London: Bemrose & Son [reprinted 1994, The Pinkfoot Press, Balgavies, Forfar]. LAING, Lloyd (1993) The Papil, Shetland, stones and their significance', Pictish Arts Society Journal 5, (Autumn 1993), 28-36. LIND, Erik Henry (1905) Norsk-Isléndska Dopnamn ockfingerade namnfran medeltiden (Norwegian Icelandic Christian Names), published in 9 pans 1905-15, Uppsala. KOCH, John T. (1983) The Loss of Final Syllables and Loss of Declension in Brittonic', Bulletin of the Board of Celtic Studies 30, 201-33, (section on Pictish 214-20). , (in preparation) A Grammar of Old Welsh. LAING, D. (1854) (ed.) Breviarium Aberdonense, 2 vols. (Bannatyne Club 96), London (originally published 1509, 1510, by William Elphinstone, bishop of Aberdeen). LAING, Lloyd L. (1993a) A Catalogue of Celtic Ornamental Metalwork in the British Isles cAD 4001200, (British Archaeological Reports, British Series 229), Oxford. , (1993b) The Papil, Shetland, stones and their significance', Pictish Arts Society Journal 4 (Autumn 1993), 28-36.
BIBLIOGRAPHY
542
LAING, Lloyd L. & Jenny LAING (1987) 'Scottish and Irish Metalwork and the conspiratio barbarica\ PSAS 116, (1986X211-221. LAMB, Raymond G. (1980) The Archaeological Sites and Monuments ofSanday and North Ronaldsay, Orkney (RCAHMS Archaeological Sites and Monuments Series 11) Edinburgh: RCAHMS. — , (1993a) 'Church and Kingship in Pictish Orkney: A mirror for Carolingian Continental Europe1, Medieval Europe (1992: Pre-printed papers 6: Religion and Belief, 101-6. — , (1993b) 'Carolingian Orkney and its Transformation*, inBATEY, JESCH&MORRIS (ed.), 260-71. LAMBERT, Pierre-Yves (1994) La Langue gauloise: Description linguistique, commentaire d'inscriptions choisés (Collection des Hesperides), Paris: Editions Errance. LANE, Alan (1980) Excavations atDunadd, Mid Argyll 1980, Interim Report, Dept. of Archaeology, Cardiff. — , (1981) Excavations at Dunadd, Mid Argyll 1981, Interim Report, Dept. of Archaeology, Cardiff. — , (1984) 'Some Pictish Problems at Dunadd', in FRIELL & WATSON, 43-62. — , (1987) 'English migrants in the Hebrides : 'Atlantic Second B' revisited', PSAS 117 (1987), 47-66. — , (1990) 'Hebridean Pottery: Problems of Definition, Chronology, Presence and Absence', in Beyond the Brochs: the Later Iron Age in Atlantic Scotland, ed. Ian Armit, Edinburgh U.P., 108-30. LANG, Andrew (1905) The Clyde Mystery. A study in forgeries and folklore, Glasgow: James Maclehose. LANG, James T. (1974) 'Hog-back monuments in Scotland', PSAS 105, 206-35. LAWRIE, Archibald (1905) Early Scottish Charters prior to 1153, Glasgow: James Maclehose. LEJEUNE, Michel, A. Vernliet, Leon Fleuriot, & Pierre-Yves Lambert (1985) Le plomb magique du Larzac et les sorcières gauloises, Paris: Centre National de recherches scientifiques [repr. in Études celtiques 22 (1985), 95-177]. LIEST0L, Aslak (1984) 'Runes', in The Northern and Western Isles in the Viking World, ed. Alexander Fenton & Herman Pálsson, Edinburgh: John Donald, 224-238. LINES, Mariaona (1989) 'Logie Stone by Inverurie, Aberdeenshire (P2)' (commercial greetings card), Edinburgh: Stoneline designs. LLOYD, J. E. (1888) The personal name-system in Old Welsh', Y Cymmrodor 9, 39-55. LöWE, Heinz (1986) 'FindanvonRheinau: Eine Irische Peregrinatio im 9. Jahrhundert', Deutsches Archiv ßr Erforschung des Mittelalters 42 Heft 1, ed. Horst Fuhrmann & Hans Martin Schaller, Köln: Böhlau Verlag, 25-85. LYNN, Chris (1989) 'Deer Park Farms: A visit to an Early Christian settlement', Current Archaeology 113, 193-8. MACALISTER, R. A. S. (1897) Smdies in Irish Epigraphy, Vol. 1, Dublin. — , (1901) 'Miscellanea: The Gigha Ogham', Journal of the Royal Society of Antiquaries of Ireland 31, 440. , (1902) Studies in Irish Epigraphy, Vol. 2, Dublin. , (1907) Studies in Irish Epigraphy, Vol. 3, Dublin.
BIBLIOGRAPHY
543
MACALISTER, R. A. S. (1910) The Ogham graffito in the Bodleian "Annals of Innisfallen"', Journal of the Royal Society of Antiquaries of Scotland, 20, 250. , (1932) 'An Ogham Inscription', in CRAW 1932, 448-50, fig. 2. , (1935) 'The Newton Stone', Antiquity 9, 389-98. — , ( 1937) The Secret Languages of Ireland with special reference to the Origin and Nature of the Shelta Language (partly based upon Collections and Manuscripts of the late John Sampson), Cambridge U.P. , (1940) 'The Inscriptions and Language of the Picts', in Feíl-Sgríbhinn Eóin mhic Néilt, Essays and Studies Presented to Professor Eoin MacNeill, ed. John Ryan, Dublin, 184-226. , (1945) Corpus Inscriptionem Insularum Celticarum, Vol. 1, Dublin: Stationery Office. [=CIIC] — , (1949) Corpus Inscriptionem Insularum Celticarum, Vol. 2, Dublin: Stationery Office. [=CIIC] MACAULEY, Donald (1976) 'Place-names', in The Moray Book, ed. Donald Omand, Edinburgh: Paul Harris, 248-63. MacBAIN, Alexander (1892) 'Ptolemy's Geography of Scotland' Transactions of the Gaelic Society of Inverness 18, (1891-2), 267-88, 288. MacDONALD, Aidan D. S. (1977) 'Old Norse 'Papar', Names in N and W , in Studies in Celtic Survival, BAR Brit. Ser. 37, ed. Lloyd Laing, Oxford 25-30. — , (1979) 'Gaelic Cill (Kil(l)-) in Scottish Place-Names', Bulletin of the Ulster Place-Name Society, Ser. 2, Vol. 2, 9-19. , & Lloyd R. LAING (1968) 'Early Ecclesiastical Sites in Scotland: a Field Survey, Part I', PSAS 100 (1967-8) (publ. 1969) 123-45. MACFARLANE, Walter (1908) see MITCHELL 1908 (edition of collection made 1748). MacGREGOR, Arthur(1974) 'The Broch of Burrian, North Ronaldsay, Orkney', PSAS 105 ( 1972-4, publ. 1976), 63-118. , (1985) Bone, antler, ivory and horn: the technology of skeletal materials since the Roman period, London: Croom Helm. MacGREGOR, Morna (1976) Early Celtic Art in North Britain, 2 Vols., Leicester U.P. MacKENZIE, W. M. (1914) (ed.) The Book of Arran, Vol.2, Glasgow: The Arran Society of Glasgow. MACKIE, Euan(1971), 'English migrants and Scottish brochs', Glasgow Archaeological Journal 2 (1971), 39-71. MacLEAN, Douglas Grant (1985) 'Early Medieval Sculpture in die West Highlands and Islands of Scotland', 2 Vols., unpublished PhD thesis, University of Edinburgh. McMANUS, Damian (1986) 'Ogam : Archaizing, Orthography and the Authenticity of the Manuscript Key to the Alphabet', Ériu 37, 1-32. , (1988) 'Irish Letter-names and dieir Kennings', Ériu 39 127-68. , (1991) A Guide to Ogam, Maynooth Monographs 4, Maynooth: An Sagart. MAC NEILL, John (Eoin) (1909) 'Notes on die distribution, history, grammar, and import of the Irish Ogham inscriptions'. Proceedings of the Royal Irish Academy 27, 329-370. (Eoin), (1919) Phases of Irish History, Dublin. , (1939) 'The Language of the Picts1, Yorkshire Celtic Studies 2 (1938-9), 3-45.
BIBLIOGRAPHY
544
McNEILL, Peter & Ranald Nicholson (1975) An Historical Atlas of Scotland c.400-c.I600 (Atlas Committee of the Conference of Scottish Medievalists), St. Andrews. MacQUARRIE, Alan (1992) 'Early Christian religious houses in Scotland: foundation and function', in BLAIR &SHARPE 1992a, 110-136. McQUEEN, John & Winifred (1989) (ed. and transi.) Scotichronicon by Walter Bower, Vol.2 Books III and IV, (D. E. R. Watt, general editor), Aberdeen U.P. McROBERTS, David (1963), The Ecclesiastical Significance of the St. Ninian's Isle Treasure', PSAS 94, 301-14. , (1965) 'The Ecclesiastical Character of the St. Ninian's Isle Treasure', The Fourth Viking Congress, ed. A Small, 224-46, Edinburgh and London. MANNING, Conleth & Fionnbarr Moore (1991) 'An Ogham stone find from Clonmacnoise', Archaeology Ireland, Vol. 5, No. 4 (Winter), 10-11. MARSTRANDER, Carl (1911) 'Ogham XOI', Ériu 5, 144. MARTIN, Martin (1703) A Description of the Western Islands of Scotland..., London. (2 ed. 1716, repr. 1980). MARWICK, Hugh (1952) Orkney Farm-Names, Kirkwall: W. R. Mackintosh. MEANEY, A.L. (1981) Anglo-Saxon Amulets and Curing Stones (British Archaeological Répons, British series 96), Oxford. MEID, Wolfgang (1992) Gaulish Inscriptions: Their interpretation in the light of archaeological evidence and their value as a source of linguistic and sociological information (Archaeolingua, Series Minor 1), Budapest: Archaeological Institute of the Hungarian Academy of Sciences/Linguistic Institute of the University of Innsbruck. MEYER, Kuno ( 1914) Über die älteste irische Dichtung II Rhythmische alliterierende reimlose Strophen, Abhandlungen der königlich Preussischen Akademie der Wissenschaften, Philos.-Hist. Klasse 10, Berlin. MILLAR, Alexander H. (1895) Fife: Pictorial and Historical, Its People, Burghs, Castles and Mansions, 2 vols., Cupar, Fife: Westwood. MITCHELL, A. (1906-8) Walter Maqfarlane: Geographical collections relating to Scotland, (Scottish History Society), 3 vols., Edinburgh. MOAR, Peter (1952) 'Two Shetland finds', PSAS 86, 206. — , & John Stewart (1954) 'Newly Discovered Sculptured StonesfromPapil, Shetland', PSAS IS (194344), 91-99. MOORE, George (1865) Ancient Pillar Stones of Scotland: their significance and bearing on ethnology, Aberdeen and Edinburgh.
BIBLIOGRAPHY
545
MORRIS, Christopher D. (1981) 'Excavation at Birsay, Orkney', in University of Durham, University of Newcastle upon Tyne Archaeological Reports for 1980, Durham, 35-40. — , (1982) 'The Vikings in the British Isles: Some aspects of their settlement and economy', in FARRELL, 70-94. , (1989) The Birsay Bay Project, Vol. L Coastal Sites Beside the Brough Road, Birsay, Orkney. Excavations 1976-82, (Dept. of Archaeology, University of Durham, Durham Monograph Series, no. 1), Stroud: Alan Sutton. , (1990) Church and Monastery in the Far North: An Archaeolgocial Evaluation (Jarrow Lecture (1989), Jarrow. , (1991) 'Native And Norse In Orkney And Shetland', in KARKOV & FARRELL (1991), 61-80. , (1993) "The Birsay Bay Project: A Résumé', in BATEY, JESCH, & MORRIS, 285-307. — , (1996) 'From Birsay to Tintagel: A Personal View', in Scotland in Dark Age Britain, ed. Barbara E. Crawford (St. John's House Papers 6), Aberdeen: Scottish Cultural Press, 37-78. MORRIS, R. W. B. (1981) The Prehistoric Rock Art of Southern Scotland, BAR Brit. Ser. 86, Oxford. MUIR, Richard (1975) Thanages', in An Historical Atlas of Scotland c.400-c.l600, ed. Peter McNeill & Ranald Nicholson, St. Andrews: Atlas Committee of the Conference of Scottish Medievalists, 27-8. MURRAY, Gordon (1986) 'The declining Pictish symbol - a reappraisal', PSAS, 116, 223-53. NASH-WILLIAMS, V. E. (1950) The Early Christian Monuments of Wales, Cardiff: Univ. Wales Press. NATIONAL MUSEUM OF ANTIQUITIES OF SCOTLAND (1892) Catalogue (New and Enlarged), Edinburgh. NELSON, Janet ( 1987) 'The Lord's anointed and the people's choice: Carolingian royal ritual', in Rituals of Rovalty: Power and Ceremonial in Traditional Societies, ed. D. Cannadine and S. Price, Cambridge U. P., 137-80. NICHOLSON, Edward (1896) The Vernacular Inscriptions of the Ancient Kingdom ofAtban, London: Bernard Quaritch. NICOLAISEN, W. F. H. (1970) 'Gaelic place-names in Southern Scotland', Studta Celtica 5, 15-35. , (1972) 'P-Celtic Place-Names in Scotland: a reappraisal', Studio Cûtica 7, 1-11. — , (1983) 'Names in the landscape of the Moray Firth', in Moray: Province and People, ed. W. D. H. Sellar, Edinburgh: Scottish Society for Northern Research, 253-62. NICOLL, Eric H. (1995) (ed.) A Pictish Panorama: The Story of the Pias, and a Pictish Bibliography, Balgavies, Forfar: Pinkfoot Press on behalf of the Pictish Ans Society. NIEKE, Margaret (1984) 'Settlement Patterns in the Atlantic Province of Scotland in the 1st Millenium AD: A study of Argyll', 3 Vols., Unpublished PhD thesis, University of Glasgow. NIGRA, Constantino (1872) Reliquie celtiche, I, II manoscritto irlandesedi S. Gallo, Turin. O'BRIEN, Michael A. (1962) Corpus Genealogiarum Hiberniae, Vol.1, Dublin: Dublin Institute for Advanced Studies. [ = CGH\ Ó CATHASAIGH, Tomas (1986) 'The sister's son in early Irish literature', Peritia 5, 128-60.
BIBLIOGRAPHY
546
Ó CORRÁIN, Donnchadh & Fidelma Maguire (1990) Irish Names, Dublin: Lilliput. Ó CUÍV, Brian (1986) 'Miscellanea: 1. 'Boicht' Chorcaige\ Celtica 18, 105. O'DELL, Andrew C. et al (1959) The St. Ninian's Isle Silver Hoard*, Antiquity 33, 241-68. , and C. Cain (1960), St. Ninian's Isle Treasure (Aberdeen University Studies 141), Edinburgh and London. O'DONOVAN, John (1868) (transi, and annot.) Cormads Glossary (ed. Whitley Stokes), Calcutta. OKASHA, Elisabeth (1971) Hand-List of Anglo-Saxon Non-Runic Inscriptions, Cambridge U.P. , (1985) fThe Non-Ogam Inscriptions of Pictland1, Cambridge Medieval Studies 9, 43-69. — , (1990) Vernacular or Latin?: The Languages of Insular Inscriptions, AD 500-1100', in Epigraphik 1988: Fachtagungßr nûttelalterliche und neuzeitliche Epigraphik Graz 10-14 Mai 1988, (Österreichische Akademie der Wissenschaften, Philosophisch-Historische Klasse, Denkschriften, 213 Band), ed. Walter Koch, Wien: Österreichischen Akademie der Wissenschaften, 139-47. O'KEEFE, J. G. (1934) 'OnMael Shechlainn, King of Ireland, f 1022, and his contemporaries\ Irish Texts 4, ed. J. Fraser, P. Grosjean, and J. G. O'Keefe (1931-4), 30-3. OLSEN, Magnus (1954) 'Runic Inscriptions in Great Britain, Ireland and the Isle of Man1, in Viking Antiquities in Great Britain and Ireland 6, ed. Haakon Shetelig Oslo, 151-233. OMAND, C. J. (1986) The Life of Findan Translated from the Latin', in The People of Orkney, (Aspects of Orkney 4), R. J. Berry & Howie N. Firth, Kirkwall: Orkney Press, 284-7. Ó MURCHADHA, Dómhnall (1980) 'Rubbings taken of the inscriptions on the Cross of the Scriptures, Clonmacnois', Journal of the Royal Society of Antiquaries of Ireland 110, 47-51. , & Giollamuire Ó MURCHÚ (1988) 'Fragmentary Inscriptions from the West Cross at Durrow, the South Cross at Clonmacnois, and the Cross of Kinnitty', Journal of the Royal Society of Antiquaries of Ireland 118, 53-66. O'RAHILLY, Cécile (1967) Táin Bo Cuailngefrom the Book ofLeinster, Dublin Institute for Advanced Studies. — , (1976) Táin Bô Cuailnge: Recension I, Dublin Institute for Advanced Studies. O'RAHILLY, Thomas F. (1946) Early Irish History and Mythology, (repr. 1984, Dublin Institute for Advanced Studies). ORDNANCE SURVEY (1968) Place names on maps of Scotland and Wales, (revised edition 1981), Southampton. Ó RIAIN, Pádraig (1985) Corpus Genealogiarum Sanctorum Hiberniae, Dublin: Dublin Institute for Advanced Studies. PADEL, Oliver P. (1972) 'Inscriptions of Pictland', Unpublished M.Litt. Thesis, University of Edinburgh. , (1972b) 'A Note on the ogham inscriptions on the stones at Newton and Logie House', Archaeology Journal 129 (1972) 196-8.
BIBLIOGRAPHY
547
PAGE, Raymond I. (1973) An Introduction to English Runes, London: Methuen. — , (1987) Reading the Past: Runes, London: British Museum PÂLSSON, Hermann & P. Edwards, (1978) (transi.) Orkneyinga Saga: The History ofthe Earls of Orkney, London. PEERS, Charles & C. A. R. Radford (1943) The Saxon Monastery of Whitby Abbey', Archaeologia 89, 27-88. PENNANT, Thomas (1772) A Tour in Scotland 1769, Second ed., London. — , (1774) A Tour in Scotland and Voyage to the Hebrides, 1772, Part I, Third ed., Chester. PINKERTON, John (1814) An Enquiry into the History of Scotland, Vol. 1, Edinburgh. POKORNY, Julius (1959, 1969) Indogermanishes Etymologisches Wörterbuch, 2 vols., Bern & München: Franke Verlag. PRICE, Glanville (1984) The Languages of Britain, London: Edward Arnold, 'Pictish' 20-27, 'Celtic Pictish' 155-7. PURSER, John (1992) Scotland's Music: A History of the Traditional and Classical Music of Scotland from Early Times to the Present Day, (BBC Scotland) Edinburgh: Mainstream. RADFORD, C. A. Ralegh (1942) The Early Christian Monuments of Scotland', Antiquity 16, 1-18. , (1953) The Pictish Symbols at Trusty's Hill, Kirkcudbrightshire', Antiquity 27. , (1959) The Early Christian and Norse Settlements at Birsay, Orkney Dept. of the Environment Official Guide, Edinburgh (2nd ed. 1978). — , (1962) 'An and Architecture: Celtic and Norse', in The Northern Isles (Studies in History and Archaeology), ed. Frederick T. Wainwright, Edinburgh: Nelson, 163-187. RAFTERY, Barry (1984) La Tène in Ireland, Marburg: Sonderband. RALSTON, Ian B. M. & Jim Inglis (1984) Foul Hordes: the Pias in the North-East and their background, Anthropological Museum, University of Aberdeen. REDFORD, Morag (1988) 'Commemorations of the Saints of the Celtic Church in Scotland', Unpublished MLitt dissertation, University of Edinburgh. RENFREW, Colin (1985) (ed.) The Prehistory of Orkney B.C. 4000 - 1000 A.D., Edinburgh U. P. (reprinted 1990 with postscript by Colin Richards). RHIND, William (1839) Sketches of the Past and Present State of Moray. A
RHYS, John (1892) The Inscriptions and Language of the Northern Picts', PSAS 26, 263-351. , (1898) 'A Revised Account of The Inscriptions of the Northern Picts', PSAS 32, 324-98. , (1899) The Gigha Ogham', in [n.a] 'Scottish Archaeological Tour of the Royal Society of Antiquaries of Ireland in Conjunction with the Cambrian Archaeological Association', Journal of the Royal Society of Antiquaries of Ireland 29, 134-218; 266-366; at 346-49. , (1901) The Gigha Ogham', Journal of the Royal Society of Antiquaries of Ireland 31, 18-23.
BIBLIOGRAPHY
548
RICHARDSON, James S. (1948) The Broch of Gurness, Aikerness, West Mainland, Orkney Ministry of Works Guide, Edinburgh: HMSO. RITCHIE, Anna (1972) 'Painted pebbles in early Scotland1, PSAS 104 (1971-2), 297-301. 1 (1977) 'Pictish and Viking-Age Farmsteads at Buckquoy, Orkney', PSAS 108 (1976-7, published 1979), 174-227. , (1983) 'Birsay around A.D. 800', in THOMSON, 46-66. , (1985a) 'Orkney in the Pictish kingdom', in RENFREW, 183-204. , (1985b) Exploring Scotland's Heritage: Orkney and Shetland, (RCAHMS), Edinburgh: HMSO. , (1986) Brough of Birsay (Historic Scotland guide-book), Edinburgh: HMSO. , (1987) 'The Picto-Scottish-Interface in Material Culture', in SMALL, 59-67. , (1988) Scotland BC: An Introduction to the Prehistoric Houses, Tombs, Ceremonial Monuments and Fortifications in the Care of the Secretary of State for Scotland, Edinburgh: HMSO. — , (1989) The Pias: An introduction to the life of the Pias and the carved stones in the care of the Secretary of State for Scotland, Edinburgh: HMSO. , & David J. Breeze (1991) Invaders of Scotland: An Introduction to the Archaeology of the Romans, Scots, Angles and Vikings, Highlighting the Monuments in the Care of the Secretary of State for Scotland, Edinburgh: HMSO. RITCHIE, J. N. Graham (1969) 'Two new Pictish symbol stones from Orkney', PSAS 101 (1968-9), 1303. , & Anna Ritchie (1981) Scotland: Archaeology and Early History (Ancient Peoples and Places), London: Thames & Hudson. , & Iain Fraser (1994) Pictish Symbol Stones. A Handlist 1994, (RCAHMS), Edinburgh: HMSO. RITCHIE, James (1916) 'Description of sculptured stone at Rayne and small cross at Culsalmond, Aberdeenshire', PSAS 50, 279-87. , (1923) 'Stone-circles at Raedykes, near Stonehaven, Kincardineshire, PSAS 57 (1922-3), 20-8. RITCHIE, P. Roy (1984) 'Soap-Stone Quarrying in Viking Lands', in FENTON & PÁLSSON, 59-84. RIVET, A. L. F. & Colin SMITH (1979) The Place-Names ofRofnan Britain, London: Batsford. ROBERTSON, Anne S. (1970), 'Roman finds from non-Roman sites in Scotland', Britannia 1, 198-213. ROBERTSON, Joseph (1843) Collections for a History of the SMres of Aberdeen and Banff (Spalding Club), Aberdeen. ROBINSON, Main (1985) (éd.) The Concise Scots Dictionary, Aberdeen U.P. ROSS, Sinclair (1983) 'The Culbin Sands - a mystery unravelled', in Moray: Province and People, ed. W. D. H. Sellar, Edinburgh: Scottish Society for Northern Research, 187-204. Royal Commission on the Ancient and Historic Monuments of Scotland ( = RCAHMS) RCAHMS (1911a) Second Report and Inventory of Monuments and Constructions in the County of Sutherland, Edinburgh: HMSO. , (191 Ib) Third Report and Inventory of Monuments and Constructions in the County of Caithness, Edinburgh: HMSO. , (1928) Inventory of the Ancient and Historic Monuments of The Outer Hebrides, Skye and the Small Isles, Edinburgh: HMSO.
BIBLIOGRAPHY
549
RCAHMS (1933) Eleventh Report and Inventory of Monuments and Constructions in the Counties of Fife, Kinross and Clackmannan, Edinburgh: HMSO. , (1946) Inventory of the Ancient and Historic Monuments of Orkney and Shetland 3 Vols., Edinburgh: HMSO. , (1971) Argyll: An Inventory of the Monuments. 1 Kintyre, Edinburgh: HMSO. , (1982) Argyll: An Inventory of the Monuments. 4 Iona, Edinburgh: HMSO. , (1985) Pictish Symbol Stones. A Handlist 1985, Edinburgh: HMSO. , (1988) Argyll: An Inventory cfthe Monuments. 6 Mid-Argyll and Cowall - Prehistoric and Early Historic Monuments, Edinburgh: HMSO. , (1992) Argyll: An Inventory of the Monuments. 7 Mid-Argyll and Cowall - Medieval and Later Monuments, Edinburgh: HMSO. , (1994a) South-East Perth: An Archaeological Landscape, Edinburgh: HMSO. , (1994b) Pictish Symbol Stones. A Handlist 1994, Edinburgh: HMSO [= Ritchie & Fraser 1994) RUTHERFORD, Anthony & Graham Ritchie (1974) The Catstane\ PSAS 105, 183-8. SAMSON, Ross (1992) 'The Reinterpretation of the Pictish Symbols', Journal of the British Archaeological Association 145, 29-65. SAWYER, Peter (1981) Roman Britain, (Oxford History of England), Oxford: Clarendon. SCOTT, Jack (1966) South-West Scotland, (Regional Archaeologies), London: Cory, Adams & MacKay. SCOTT, Lindsay (1948) 'Gallo-British Colonies: The Aisled Round-House Culture in the North', Proceedings of the Prehistoric Society 14 (n,s), Cambridge, 46-125. SHEPHERD, Ian A. G. (1983) 'A Grampian Stone Circle Confirmed' PSAS 113, 630-4. , (1993) 'The Picts in Moray', in Moray: Province and People, ed. W. D. H. Sellar, Edinburgh: Scottish Society for Northern Research, 75-90. , & Alexandra N. Shepherd (1978) 'An incised Pictish figure and a new symbol stone from Barflat, Rhynie, Gordon District', PSAS 109, 211-22. 214-5. , & Ian B. M. RALSTON (1979) Early Grampian: A Guide to the Archaeology, Aberdeen: Dept. of Physical Planning, Grampian Regional Council. SHETELIG, Haakon (1954a) 'Viking Graves', in Viking Antiquities in Great Britain and Ireland, Part IV, ed. Haakon Shetelig, Oslo, 65-112. , (1954b) 'Runic Inscriptions: Shetland, Orkney, Scotland', in Viking Antiquities in Great Britain and Ireland, Pan IV, ed. Haakon Shetelig, Oslo, 158-178. SIMPSON, W. Douglas (1944) The Province qf Mar (lúúná Lectures 1941), Aberdeen University Studies 121, Aberdeen U.P. SIMS-WILLIAMS, Patrick (1992) The Additional Letters of the Ogam Alphabet', Cambridge Medieval Celtic Studies 23 (Summer 1992), 29-75. , (1993) 'Some Problems in Deciphering the Early Irish Ogam Alphabet', Transactions of the Philological Society 91.2, 133-180. SKENE, William F. (1865) 'Notes on the ogham inscription on the Newton Stone', PSAS 5, 289-98. , (1874) 'Notice of an ogham inscription in the churchyard of Aboyne, Aberdeenshire', PSAS 10 (1872-74), 602-3.
BIBLIOGRAPHY
550
SMALL, Alan ( 1987) (ed.) The Pias: A New Look at Old Problems, Dundee: Graham Hunter Foundation. , (1973b) 'The site: its history and excavation1, in SMALL 1973a, 1-7. , Charles Thomas & David M. Wilson ( 1973a) St. Ninian 's Isle and its Treasure (Aberdeen University Studies 152), London: Oxford U.P. SMYTH, Alfred P. ( 1984) Warlords and Holy Men: Scotland AD 80-100, (The New History of Scotland), London: Edward Arnold. SOUTHESK, Rt. Hon. Earl of [James Carnegie] (1883) The Newton Stone', PSAS 17, 21-45. , (1884) The ogham inscriptions of Scotland1, PSAS 18 (1883-4), 180-206. , (1886) 'The oghams on die Brodie and Auquhollie Stones, with Notes on the inscriptions of the Golspie and Newton Stones, and a list of the oghams in Scotland1, PSAS 20 (1885-6), 14-40. — , (1893) The Origins of Pictish Symbolism, with notes on the sun boar, and a new reading of the Newton inscriptions, Edinburgh. , (1895) 'An Ogam Inscription at Abemethy, 1895\ PSAS 29, 244-51. SOUTHWICK, Leslie (1981) The So^Called Sueno's Stone at Forres, Elgin: Moray District Libraries. SPEARMAN, R. Michael (1993a) The Mounts from Crieff, Perthshire, and their wider Context1, in SPEARMAN & HIGGITT, 135-42. , & John HIGGITT (1993) The Age of Migrating Ideas: Early Medieval Art in Northern Britain and Ireland, Edinburgh: NMS. SPROULE, David (1985) 'Politics and Pure Narrative in the Stories about Core of Cashel', Erin 36, 1128. ST. JOSEPH, John Kenneth Sinclair (1978) The camp at Durno, Aberdeenshire, and the site of Mons Graupius', Britannia 9, 271-87. STEER, Kenneth (1969) Two unrecorded Early Christian stones', PSAS 101, 127-9. — , & John Bannerman (1977) Late Medieval Monumental Sculpture in the West Highlands, Edinburgh: HMSO. (RCAHMS) STEPHENS, George (1875) 'Note of a Fragment of a Rune-Inscribed Stone from Aith's Voe, Cunningsburgh, Shetland, now in the Museum', PSAS 10 (1872-3 & 1873-4), 425-430. , (1901) The Old-Northern Runic Monuments of Scandinavia and England, Vol.4 (4 vols. 1866-1901), London and Copenhagen. STEVENSON, Robert B. K. (1952) 'Long cist burials, particularly those at Galson (Lewis) and Gairloch (Wester Ross), with a symbol stone at Gairloch', PSAS 86, (1951-2), 106-11. , (1955) 'Pictish An*, in The Problem of the Pias, ed. F. T. Wainwright, Edinburgh: Nelson (repr. 1980, Penh: Melven Press), 97-128. , (1959) 'The Inchyra Stone and some other Unpublished early Christian Monuments', PSAS92,33-55. , (1981) 'Christian sculpture in Norse Shetland', Frodskaparrit (Heidwrscrit til Sverra Dahl), Annales Societatis Scientiarum Faeroensis 28/29, 283-92. [STEVENSON, R. B. K.] (1947) [Note], in 'Donations to and purchases for the Museum', PSASSl (19467, publ. 1949) 191-3, pl.XXII.4.
BIBLIOGRAPHY
551
STOKES, Whitley (1872) 'The Manumissions in the Bodmin Gospels', Revue celtique 1, 332-45. — , (1890) 'On the linguistic value of the Irish Annals', Transactions of the Philological Society (188890), 365-433, (includes 'Pictish names and other words', 390-418). , (1895-7) (ed.) 'Annals of Tigernach', Revue celtique 16 (1895) 374-419; 17 (1896) 9-33, 119-263, 337-420; 18 (1897) 9-59, 150-97, 267-303. — , (1900) Accallamh no Senôrach (Irische Texte 4th series, vol.1, ed. Ernst Windisch & Whitley Stokes), Leipzig. — , (1905) Accallam in dá Thuraid 'The Colloquy of the Two Sages', Revue celtique 26, 4-64. — , & John STRACHAN (1901-3) (ed.), Thesauruspalaeohibernicus (2 vols, and supp., Cambridge and Halle a.S, repr. in 2 vols. Dublin 1975). STUART, John (1818) 'An account of some sculptured pillars in the northern part of Scotland', Archaeologia Scotica: Transactions of the Society of Antiquaries of Scotland 2, 314-23. — , (1856) Sculptured Stones of Scotland, Vol. I, Aberdeen: Spalding Club. — , (1867) Sculptured Stones of Scotland, Vol. II, Edinburgh: Spalding Club. — , (1869) (ed.) The Book of Deer, Edinburgh: Spalding Club. SUTHERLAND, Elizabeth (1994) In Search of the Pias, a Celtic Dark Age Nation, London: Constable. TAIT, E. S. Reid (1938) [Note], in 'Donations to the Museum', PSAS 71 (1937-38), 369-70. TAYLOR, Alexander Bun (1931) 'Some Saga Place-Names', Proceedings of the Orkney Archaeological Society 9 (1930-1), 41-5. TAYLOR, Simon (1995) Settlement Names in Fife, unpublished PhD dissertation, University of Edinburgh. THOMAS, Charles A. (1961) 'The Animal Ait of the Scottish Iron Age and its Origins', Archaeological Journal 118, 14-64. — - , (1963) 'The Interpretation of the Pictish Symbols', Archaeological Journal 120, 31-97. — , (1971) The Early Guistian Archaeology of North Britain, (The Hunter Marshall Lectures 1968), London: O.U.P. — , (1973) 'Sculptured stones and crosses from St. Ninian's Isle and Papil', in SMALL et al., 8-44. — , (1990) '"Gallici nautae de Galliamm prvinciis" - a sixth/seventh century trade with Gaul, reconsidered', Medieval Archaeology 34, 1-26. — , ( 1994) And Shall These Mute Stones Speak ?: Post-Roman Inscriptions in Western Britain, Dalrymple Archaeological Monographs 2, Cardiff: University of Wales Press. THOMAS, F. W. L. (1879) 'Dunadd, Glassary, Argyllshire; The Place of Inauguration of the Dalriadic Kings', PSAS 13 (1878-9) 28-47. THOMPSON, Alexander (1865) 'Notice of the various Attempts which have been made to Read and Interpret the Inscription on the Newton Stone, Garioch, Aberdeenshire', PSAS 5, 224-34. THOMSON, William P.L. (1983) (ed.) Birsay: A Centre of Political and Ecclesiastical Power (Orkney Heritage 2), Kirkwall: Orkney Heritage Society. — , (1986) 'St. Findan and the Pictish-Norse Transition (Appendix B)', in The People of Orkney (Aspects of Orkney 4), (ed.) R. J. Berry & Howie Firth, Kirkwall: Orkney Press, 279-87.
BIBLIOGRAPHY
552
THURNEYSEN, Rudolf (1891) (ed.) 'Mittelirische Verslehren1, Irische Texte 3, Leipzig. , (1912) [review of Mac Neill in PRIA 27 (1909)], Zeitschrift für celtische Philologie, 8, 184-5. , (1946)4 Grammar of Old Irish (Revised and enlarged edition with supplement), transi. D. A. Binchy & Osboro Bergin, repr. 1980, Dublin Institute for Advanced Studies. TRAILL, William (1890) 'Results of Excavations at the Broch of Bunian, North Ronaldsay, Orkney, during the summers of 1870 and 1871', Archaeologia Scotica: Transactions of the Society of Antiquaries of Scotland 5 (1874-90), 341-64. TURNER, Val ( 1991) How to be a Detective: an introduction to Shetland's archaeology for bairns of all ages, Shetland Amenity Trust (second edition 1993). , (1994) 'The Mail Stone: an incised Pictish figure from Mail, Cunningsburgh, Shetland', PSAS 124, 315-25. TYLECOTE, Ronald Frank (1962), Metallurgy in Archaeology: a prehistory of matallurgy in the British Isles, London. UHLICH, Jürgen (1989) 'DOV(A)- and lenited -B- in Ogam', Ériu 40, 129-133_ , (1993) Die Morphologie der komponierten Personnamen des Altirischen (Beiträge zu Sprachwissenschaften 1), Witterschlick/Bonn: Wehle. VENDRYES; Joseph (1959) Lexique étymologique de l'Irlandais Anden A, Paris: Centre Nationale du Recherche Scientifique (and Dublin: Dublin Institute for Advanced Studies). , (1960) Lexique étymologique de l'Irlandais Anden M NOP, Paris: Centre Nationale du Recherche Scientifique (and Dublin: Dublin Institute for Advanced Studies) (repr. 1983). WAINWRIGHT, Frederick T. (1955a)(ed.) The Problem of the Pias, Edinburgh: Nelson (repr. 1980, Perth: Melven Press), , (1955b) 'Houses and graves', in Wainwright 1955a, 87-96. , (1961) The Inchyra Ogam (Dundee Archaeological Studies 1) [first published in Ogam 11 (1959), 269-78 (Rennes)]. , (1962a) (ed.) The Northern Isles (Studies in History and Archaeology), Edinburgh: Nelson. , (1962b) 'Pictsand Scots', in WAINWRIGHT 1962a, 91-116. , (1962c) The St. Ninian's Isle Excavation', in WAINWRIGHT 1962a, 193-4. WALKER, Bruce & Graham Ritchie (1987) Exploring Scotland's Heritage: Fife and Tayside, (RCAHMS), Edinburgh:HMSO. WARNER, G.F. (1906, 1915) The Stowe Missal (Henry Bradshaw Society 31-2), 2 vols. WARNER, Richard B. (1983) 'Ireland, Ulster and Scotland in the earlier Iron Age', in From the Stone-age to the 'Forty-Five: Essays in Honour ofR. B. K. Stevenson, ed. Anne O'Connor & David V. Clarke, Edinburgh: John Donald, 160-87. WATSON, William J. (1926) The History of the Celtic Place-Names of Scotland: being the RhindLectures on Archaeology (expanded) delivered in 1916, Edinburgh: William Blackwood. WAUGH, Doreen J. (1985) The Place-Names of Six Parishes in Caithness, Scotland, Unpublished PhD thesis, University of Edinburgh. (1993) 'Caithness: An Onomastic Frontier Zone', in BATEY, JESCH, & MORRIS, 120-8.
BIBLIOGRAPHY
553
WEBSTER, Graham (1987) Celtic Religion in Roman Britain, Totowa, N.J.: Barnes & Noble (British edition = The British Celts and Their Gods Under Rome, 1986). WHATMOUGH, Joshua (1940) 'Gaulish uimpi\ Language 25, 388-91. WHITE, T. P. (1873) The Ecclesiastical Antiquities of the District of Knapdale, Argyllshire, and the Islands of Gigha and Cara\ PSAS 10, 383-8. , (1875) Archaeological Sketches in Scotland - Knapdale and Gigha. WHITELOCK, Dorothy, David Dumville & Rosamund McKitterick (1982) (ed.) Ireland in Medieval Europe: Studies in Memory of Kathleen Hughes, Cambridge U.P. WILLIAMS, Ifor (1937) The personal names in the early inscriptions', in An Inventory of the Ancient Monuments of Wales and Monmouthshire, VIll, Anglesey, London: Stationery Office, Appendix VI, cxivcxvii. , (1938) Canu Aneirin: gyda Rhagymadrodd a Nodiadau, Cardiff: University of Wales Press. — , (1949) The Towyn Inscribed stone* (Presidential Address to the Cambrian Archaeological Association), Archaeologia Cambrensis C, 160-72. WILSON, Daniel (1863) Prehistoric Annals of Scotland, (2nd ed.) vol. II, London. WOOD, Nicola (1989) Scottish Place Names, Edinburgh: Chambers. WOODHAM, A. A. & John MacKenzie (1957) Two cists at Golspie, Sutherland', PSAS 90, 234-38. YOUNGS, Susan (1989) (ed.) 'The Work ofAngels ': Masterpieces of Celtic Metalwork, 6th - 9th Cenmries A.D., London: British Museum. ZIMMER, Heinrich (1898) 'Matriarchy Among the Picts', inLeabharnan Gleann, ed. George Henderson, Edinburgh, 1-42, (transl. from the original which appeared in the Zeitschrift der Savigny-Stiftung fiir Rechtgeschichte, Romanistische Abteilung 15 (Weimar).