New Perspectives on English Historical Linguistics: v.2: Lexis and Transmission: Vol 2 (Current Issues in Linguistic Theory)

NEW PERSPECTIVES ON ENGLISH HISTORICAL LINGUISTICS II

AMSTERDAM STUDIES IN THE THEORY AND HISTORY OF LINGUISTIC SCIENCE General Editor E.F. KONRAD KOERNER (Zentrum für Allgemeine Sprachwissenschaft, Typologie und Universalienforschung, Berlin) Series IV – CURRENT ISSUES IN LINGUISTIC THEORY Advisory Editorial Board Lyle Campbell (Christchurch, N.Z.); Sheila Embleton (Toronto) Brian D. Joseph (Columbus, Ohio); John E. Joseph (Edinburgh) Manfred Krifka (Berlin); E. Wyn Roberts (Vancouver, B.C.) Joseph C. Salmons (Madison, Wis.); Hans-Jürgen Sasse (Köln)

Volume 252

Christian Kay, Carole Hough and Irené Wotherspoon (eds) New Perspectives on English Historical Linguistics Selected papers from 2 ICEHL, Glasgow, 2–26 August 2002 Volume II: Lexis and Transmission

NEW PERSPECTIVES ON ENGLISH HISTORICAL LINGUISTICS SELECTED PAPERS FROM 12 ICEHL, GLASGOW, 21–26 AUGUST 2002 VOLUME II: LEXIS AND TRANSMISSION

Edited by

CHRISTIAN KAY CAROLE HOUGH IRENÉ WOTHERSPOON University of Glasgow

JOHN BENJAMINS PUBLISHING COMPANY AMSTERDAM/PHILADELPHIA

8

TM

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences — Permanence of Paper for Printed Library Materials, ANSI Z39.48-984.

Library of Congress Cataloging-in-Publication Data International Conference on English Historical Linguistics (2th : 2002 : Glasgow University) New persprectives on English historical linguistics : selected papers from 2 ICEHL, Glasgow, 2–26 August 2002. p. cm. -- (Amsterdam studies in the theory and history of linguistic science. Series IV, Current issues in linguistic theory, ISSN 0304-0763 ; v. 25–252) Includes bibliographical references and index. Contents: v. . Syntax and morphology / edited by Christian Kay, Simon Horobin, Jeremy Smith -- v. 2. Lexis and transmission / edited by Christian Kay, Carole Hough, Irené Wotherspoon. . English language--Grammar, Historical--Congresses. 2. English language--History--Congresses. I. Kay, Christian. II. Horobin, Simon. III. Smith, J. J. (Jeremy J.) IV. Title. V. Series. PE075.I57 2002b 427--dc22 2004047943 ISBN 90 272 4763 3 (Eur.) /  588 54 3 (US) (Hb; Volume : alk. paper) ISBN 90 272 4764  (Eur.) /  588 55  (US) (Hb; Volume 2: alk. paper) ISBN 90 272 4765 X (Eur.) /  588 528 3 (US) (Hb; Set) © 2004 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. • P.O.Box 36224 • 020 ME Amsterdam • The Netherlands John Benjamins North America • P.O.Box 2759 • Philadelphia PA 98-059 • USA



"ack"> "smi"> "ben"> "col"> "dar"> "dia"> "dur"> "fis"> "hac"> "hic"> "hou">

Table of contents

Acknowledgements Introduction Chancery Standard Michael Benskin

vii ix 1

Cant and slang dictionaries: A statistical approach Julie Coleman

41

DOST: A significant instance of historical lexicography Marace Dareau

49

Image schemata and light: A study in diachronic lexical domains in English Javier E. Díaz Vera

65

Loanword etymologies in the third edition of the OED
: Some questions of classification Philip Durkin

79

“Non olet”: Euphemisms we live by Andreas Fischer Intrusive [h] in present-day English accents and ·hÒ-insertion in medieval manuscripts: Hypercorrection or functionally-motivated language use? Martina Häcker

91

109

Mergers, near-mergers and phonological interpretation Raymond Hickey

125

New light on the verb “understand” Carole Hough

139

< /R/TREARGET E FF

"ker"> "mil"> "pon"> "pou"> "sco"> "sve"> "tis"> "wel"> "si"> "toc">

vi

Table of contents

Homophones and the stabilization of orthography in nineteenth-century and early twentieth-century English Susan Kermas

151

Kailyard, conservatism and Scots in the Statistical Accounts of Scotland
Robert McColl Millar

163

A sociolinguistic approach to the Norse-derived words in the glosses to the Lindisfarne and Rushworth Gospels Sara M. Pons-Sanz

177

Haplology in English adverb-formation Amanda Pounder

193

Uses of Scottish place-names as evidence in historical dictionaries Margaret Scott

213

On the stressing of French loanwords in English Ann-Marie Svensson

225

Like like love: Comparing two modern English words diachronically Heli Tissari

235

Spirantisation and despirantisation Jerzy Welna

251

Subject index

267

DOCINFO AUTHOR ""TITLE "Acknowledgements"SUBJECT "Current Issues in Linguistic Theory, Volume 252"KEYWORDS ""SIZE HEIGHT "220"WIDTH "150"VOFFSET "4">

Acknowledgements

The Editors would like to record their thanks to the contributors to this volume and to those who helped at various stages of the reviewing process: Julie Coleman, Andreas Fischer, Olga Fischer, Margaret Laing, Matti Kilpiö, Bettelou Los, April McMahon, Carol Percy, Jane Roberts, Irma Taavitsainen and Keith Williamson. Our thanks are also due to Konrad Koerner series editor of CILT, for his support and, Anke de Looper of John Benjamins for her enthusiasm, patience and advice. We are also grateful to all those who helped to set up and run the Twelfth International Conference on English Historical Linguistics from which these papers derive, notably colleagues at the Glasgow English Language Department, the STELLA project, the Edinburgh Institute for Historical Dialectology, and all our student helpers. A special thank you goes to Ian Hamilton for his tireless work on conference organisation and the resulting volumes of papers. We would also like to acknowledge with thanks the financial support of the British Academy, the Linguistics Association of Great Britain, Oxford University Press, the University of Glasgow, and Thomas Tunnock Ltd. Glasgow, December 2003 The Editors



Introduction

The papers collected in this volume are a selection of those given at ICEHL12, which was held at Glasgow University in August 2002. They are concerned, very broadly, with the historical lexicology and transmission of English. A first volume, concerned with grammatical issues, is being published simultaneously, and together the two volumes provide a fair summary of many if not most issues that are currently engaging practitioners of English historical linguistics. The editors spent some time deciding on a title for these volumes, and the final choice, New Perspectives on English Historical Linguistics, is a somewhat sober compromise. But, as I point out in the preface to Volume I, this is of course a sober title with an ambitious claim. The historical study of the English lexicon and sound-/writing-system has a long pedigree, and it is sometimes hard to see one’s work as much more than simply footnotes to the Oxford English Dictionary and the works of Karl Luick or Eric Dobson. Is it possible, in other words, to say something really new on these subjects? Is it possible to justify the historical study of the English lexicon and its transmission as something worth doing — something for which, to put it bluntly, it is legitimate to be paid to pursue? We think it is, and the evidence is supplied by the papers printed here. ICEHL12 did, of course, coincide happily with one event whose newness cannot be disputed: the publication of the Dictionary of the Older Scottish Tongue. DOST is one of that “select band of historical dictionaries” (as Philip Durkin described them in his celebratory speech) envisaged by Sir William Craigie back in 1919 as the obvious next steps after the completion of the Oxford English Dictionary. As an intellectual achievement DOST is rivalled only by the Middle English Dictionary which, with much greater resources, was completed the previous year. Together with the Scottish National Dictionary, which appeared in 1976 and covered the period since 1700, DOST enables scholars to trace the complete history of the Scots lexicon from its origins in the twelfth century up to the present day. Marace Dareau’s paper, which derives from her plenary presentation at the conference, will become a classic description of the vicissitudes of undertaking a long-term research project during a

x

Introduction

period of unparalleled social and intellectual change in the British — specifically Scottish — university system. It is tempting to see the completion of such projects as in some sense a final statement: a line is drawn, and scholarship turns to other research questions in other fields where answers are more easily and more quickly found. Certainly some would have been tempted to encourage such a development: the authorities of the ancient Scottish universities, perhaps, who (with occasional gaps and with occasional protests) kept DOST in funds for so many years. However, the other papers in this book demonstrate very clearly that lexicographical monuments such as OED, MED, SND and DOST are rather like the Victorian infrastructures on which modern cities depend. First, they are never truly finished, and need constant renewal. Thus, for instance, all are being or have been revised or reconfigured to take advantage of advances in computer technology — something, incidentally, on which the greatest editor of DOST, the late Jack Aitken, was a world authority. Secondly, they are also indispensable foundations for further scholarly endeavour, and most of the remaining papers in this volume show how the disciplines of historical lexicology can be pushed forward in the light of such achievements. In so doing they constitute a definition of what historical lexicology should be about, and it is perhaps appropriate in this introduction to ask such a defining question. The study of the lexicon has suffered in some circles from under-theorising, perhaps because its study requires engagement with all levels of language. Lexicology is sometimes identifed with semantics, but this is wrong; semantics is the underpinning for all levels of language (grammar, lexicon and transmission). The lexicologist needs to engage with semantic notions in relation to the lexicon, but also with other levels of language: morphology, syntax and transmission. Onomastics, a distinct branch of lexicology which raises complex questions of semantics and morphology, forms another strand of research. The lexicologist needs to develop an historical sense through, for example, the study of etymology, and engage with questions of evidence, and of style and register, both diachronically and diatopically. This will include insights from modern sociolinguistics, whereby the “present can explain the past”. More generally, lexicologists find themselves confronted constantly with questions of classification and organisation, often, until the recent renewal of interest through cognitive semantics, without the formalist luxuries afforded to some schools, say, of phonology or grammar. Above all, though, lexicology is the area of linguistic enquiry where data are of paramount importance. The historical lexicologist of English is confronted

Introduction

with literally millions of words recorded over 1500 years. In such circumstances, of course, it is only too easy for scholars to lose sight of the wood amidst the trees, and the task of the lexicologist in particular is to make sense of this mass of data. Many, if not all, of the papers in this volume begin with a close study of a piece of data, but all of them — it was something the editors tried to emphasise in making their selection from those delivered at ICEHL — attempt to draw conclusions with a more general significance for linguistic theory. Many use new research tools, such as machine-readable corpora, or new intellectual tools, such as cognitive linguistics. More broadly, all the papers in this volume, whether concerned with lexicology or transmission, have a social orientation; neither lexicology nor phonology, according to the authors of the papers collected here, can be seen as divorced from their social setting. Although it is a claim which is also a cliché, it may be supposed that such developments constitute a new Kuhnian paradigm for English historical linguistics. This paradigm has been emerging for the last thirty years: the rapprochement between philology and linguistics. At one time, philology — data-centred, apparently innocent of theory — seemed to many in the academic community to have completed its task. Texts had been edited, dictionaries completed, dialectological surveys undertaken. The way was open for theoryrich linguistics to take over, and indeed some felt able to claim that all previous scholarship needed to be recouched in (say) the terminology of generative theory before further progress could be made. Philological reaction to such opinions was often hostile, and for some time the two approaches to the historical study of English diverged markedly. The focus on data which is characteristic of lexicology is not, of course, confined to that subject, and the essays on phonology and writing-systems in this volume are similarly data-centred. The divergence between philology and linguistics is now surely over. To paraphrase Michael Halliday, we have learnt “to build on our predecessors and move forward, instead of constantly staying behind where they were in order to trample them underfoot” (Halliday 1987: 152). The most exciting current work in the field draws on the strengths of both philological and linguistic traditions; both traditions have learnt to value each other, and the dialogue between these complementary approaches is one of the most encouraging characteristics of the discipline as it is currently practised. These essays demonstrate how productive this dialogue is proving to be. Jeremy J. Smith Glasgow, 2003

xi



xii

Introduction

Reference Halliday, M. A. K. 1987. “Language and the order of nature”. The Linguistics of Writing ed. by N. Fabb & A. Durant, 135–154. Manchester: Manchester University Press.



Chancery Standard Michael Benskin University of Oslo

1.

London English and the language of government

‘Chancery Standard’, otherwise ‘Type IV’, is the name given to a form of fifteenth-century London English identified as the direct ancestor of the modern written standard. The term was introduced by M. L. Samuels in his classic paper of 1963, “Some applications of Middle English dialectology”, to designate the type of language found in “that flood of government documents that starts in the years following 1430” (1963: 88). In many of its features, Type IV breaks continuity with the earlier mainstream London language (Samuels’s ‘Type III’) as it appears, for example, in the Ellesmere Chaucer, Hoccleve’s autographs, and a range of administrative writings from the London of their time. Among the salient differences are:

these their such should not through gave 1/3sg.ind. but

Type III

Type IV

thise hir(e) swich(e) sholde nat thurgh yaf bot

thes(e) theyr(e), þeir(e), þair(e), her such(e) shulde not thorough, þorow(e) gaf but

(Samuels 1963: 89, note 10)

These differences had long been known, but were taken to separate a Westminster ‘parliamentary’ dialect from that of London. Samuels held that they were to be explained chronologically. No less than Type III, which eventually it replaced,

2

Michael Benskin

Type IV was a variety of London language, but it did not appear until a good two generations after Type III had become established. Between times — and this is the difference between the two types — London language had changed, as the balance of immigration into the capital shifted from one part of the country to another. Type IV evolved: from a combination of spoken London English and certain Central Midland elements, which themselves would be transmitted via the spoken, not the written language. But the result was a written, not a spoken, standard, which was to spread considerably in use by 1470. (Samuels 1963: 93)

The Type IV forms, save for such(e) such, cannot have evolved from their Type III London equivalents, but came in from outside and more or less abruptly displaced them, a linguistic hypothesis that tallied remarkably with the conclusions of Ekwall’s onomastic inquiries (1951, 1956) into the local origins of London’s fourteenth-century population. That the ‘Chaucerian’ and ‘Westminster’ types represented primarily successive and not co-aeval stages of London language could hardly be gainsaid, and on this view, had the offices of state adopted English a generation or so before, then presumably Type III would have spread to become a national written standard. In the event, Type IV, an amalgam of that older usage with elements from the Central Midlands dialects, “was backed by the full weight of the administrative machine, and was certain to oust eventually (though by no means immediately) the other incipient standards” (Samuels 1963: 89). Taxonomically, the assemblage of forms in Samuels’s inventory for Type IV is illustrative only; as defining characteristics of the type, they are necessary but not sufficient, a difference commonly overlooked. These forms were listed as merely a few of the “more outstanding” replacements that distinguish later London usage from the ‘Chaucerian’ type, and they are characteristic in dialects of the Central Midlands as well; it is precisely their independent co-occurrence in midland language that enables the likely immigrant component in the London amalgam to be identified. Indeed, the Central Midland element in London language of the later fifteenth century goes well beyond the inventory for Type IV: among forms widespread though never dominant are sich such, mych much, ony any, and youen given. It is clear, too, that although some of the Type IV stigmata were worn by the written standard of the Central Midlands, they came into London English not from manuscripts but through immigrant speech: telling is the near-absence, in city and government documents alike, of ·ijÒ corresponding to ME /i˜/, one of the most distinctively

Chancery Standard

orthographical features characteristic in Central Midland Standard. There is nothing to support the idea (Fisher 1977:885) that Type IV first came into being as a written form, as a merger between London and Central Midland literary standards; none of the midland forms separating Type IV from Type III presupposes visual adoption into London use, and their entry, as will appear, is spasmodic. The terms ‘Type III’ and ‘Type IV’ are self-evidently shorthand, better applied, it may now be thought, to the forms that differentiate them than to discrete dialectal types, in that the presence of any one such form does not entail the presence or absence of any of the others.1 For example, both the Ellesmere Chaucer and Hoccleve’s autographs are canonically Type III, but even they show, as minor variants, Type IV noght (Ellesmere) and not (Hoccleve) for not, and for but they have only Type IV but. London administrative writings show that by 1416–1424 combination was becoming promiscuous:2 these IV thes(e), þes(e) their IV the(i)r, þeir, þair, her(e) such IV such(e) should IV shuld(e) predominant, but III shold not III nat predominant, but IV not beside noght

1.Lately, it has been held in conference proceedings that London writers deliberately chose Type III for their literary compositions against Type IV for their administrative work, as if they themselves apprehended the philologist’s construct. The only ground for this belief is that in the short time span considered, the literary texts are generally earlier than the administrative texts, by whose date the Type IV forms were increasingly in use. In administrative writings of the later fifteenth century, however, both from London and the state offices, some Type III forms persist, and even (to all appearances) may be revived (see below, p. 36). Similarly, later literary works, among which some of the London chronicles, show Type IV forms no less than does contemporary administrative usage. 2.City of London Corporation Records Office, Guildhall Letter-Books I and K, extracts printed by Chambers & Daunt 1931, 62–89 (‘The War with France’) and 92–115 (‘Guildhall Letter-Books’). The Letter-Books are registers containing copies of the Corporation’s correspondence, ordinances, articles submitted for the Corporation’s approval, etc.; in detail, their language may variously reflect either the originals or the registrars. The usage here reported is confined to the Corporation’s own compositions (cf. Atlas LP 6390, which draws additionally — and legitimately for cartographic purposes — on copies of royal missives as well as submitted articles). The present list draws on twenty-two different texts, none of them lengthy, but note that Type III and Type IV forms frequently co-occur in the same text: among others, ‘War with France’ XIV (1418) combines þese/thes, her, shulde, nat, but
; ‘Guildhall Letter-Books’ VII (1418) ther, nat, thorugh, but
; VIII (1419) þes, shold, but
; XII (?1423) her/their, suche, shuld(e), not/nat
; XIII (1423) suche, shuld, nat, bot
; XVI (1423–4) shuld, nat, yaf.

3

4

Michael Benskin

through III thurgh, IV thorugh gave III yaf but III bot predominant, but IV but

None of this is excluded by Samuels’s original account, though its brevity leaves it apt to be misunderstood as a history of sharp cut-offs in a compressed timespan (cf. Benskin 1991: 78); that its argument is otherwise bears restating. Thus, it plainly acknowledged that forms characterising the different Types were in many cases combined in the same texts (1963: 89): Naturally, not all manuscripts written in something approaching standard English in the early fifteenth century belong to Types I [Central Midland Standard], III or IV. Many … have a Middlesex or Surrey basis with a sprinkling of features from the main types.

Again, its claim that Type IV “consists of that flood of government documents that starts in the years following 1430” (1963: 88) does not entail that Type IV forms were wholly absent from London usage before 1430, or that Type IV ousted Type III once and for all at that date, or that after 1430 government documents in English were written exclusively in Type IV. In certain points, however, the article is itself mistaken. It is true that “any of the large classes of documents at the Public Record Office will show clearly that, until 1430–35, English is the exception rather than the rule”, but only for a few classes is it true that “after that, there is a sudden change, and the proportions are reversed, from a mere trickle of English documents among thousands in Latin and French, to a spate of English documents” (1963: 87). Where Latin had been the norm, it remained so, and that applies to most of the classes comprising the work of state officials. These must be distinguished from classes consisting partly or even mainly of documents whose originals, though not necessarily the extant copies, came into the state offices from outside. Among these are the Ancient Correspondence of the Chancery, Ancient Petitions, and, as Samuels himself recognised, Early Chancery Proceedings. Certainly there is a spate of documents in English, but at the expense mostly of French, and a main source is the huge increase of business in domains where the vernacular was already established. After 1430–35, that vernacular was very commonly Type IV, though it was by no means the sole form of government English. The idea that Type IV represents a ‘Chancery Standard’, however, cannot be maintained. Samuels’s 1963 paper indicated its documentary sources in general terms only, but in a later article he was more explicit. There, Type IV is described as “Typical forms

Chancery Standard

in documents written by Chancery-trained scribes, in PRO, Early Chancery Proceedings and Exchequer T. R. Council and Privy Seal” (Samuels 1983: 20), but though it is possible that Chancery-trained scribes wrote the bulk of the extant documents in Early Chancery Proceedings (C 1), they wrote precious few of those in Exchequer and T. R. Council and Privy Seal (E 28). In that class, most by far of the writings emanating from government are the work of the Privy Seal Office, and neither the institution nor its practices can be described as ‘Chancery’. This matters: the development of a written standard among several score clerks in a major department of state — Chancery — is a very different matter from convergence of usage within an office employing only a dozen (Brown 1971: 261–262). The works of John H. Fisher, however, have idealised ‘Chancery Standard’ as the product and deliberate cultivation of Chancery, with whose supposed history it has been largely identified (Fisher 1977, 1979, 1988, 1992, 1996; Fisher, Richardson & Fisher 1984). Into this ideology is woven a nexus of language and script: allegedly, ‘Chancery Standard’ goes with the Chancery hand, whence some far-reaching claims concerning the adoption of both in the propagation of literary works, Chaucer’s and Langland’s among them. Much of this now passes as orthodoxy, cited with approval by anglicists who have never troubled to scrutinise its foundations, which foundations, it will appear, comprise misrepresentations of language, of diplomatic, of palaeography, and of text. These are set out in seemingly canonical form as the introduction, largely a reworking of Fisher 1977 and Richardson 1980, to An Anthology of Chancery English, a collection of two hundred and forty-one allegedly illustrative texts (Fisher, Richardson & Fisher 1984, henceforth cited as ‘Anthology’).

2. The texts of the Anthology Anthology’s preface announces that “This book has been an adventure in autography” (ix). So it is, but hardly in the sense that its compilers supposed, as may be confirmed by anyone who compares their texts with “the velums in the Public Record Office” (sic, though one might have imagined that more of the animal was used). They are riddled with mistakes. From thirteen texts, most of them short, Lister Matheson (1986: 647–648) reported seventy-six errors, but even his twenty “more substantive or dialectal variations” are far from the worst. The editing is uninformed not only philologically but historically, and the annotation, brief though it is, can accordingly go far astray. This matters not

5

6

Michael Benskin

least because it is common ground that the development of standard English involved regional spread and the displacement of local usage, but whereas one might imagine that the evidence for the institutional and local origins of the illustrative texts had therefore been carefully scrutinised, it is apt to be traduced. To rehearse in full the failings of Anthology’s apparatus and texts would be to write a book, but lest anyone be inclined to rely upon them, consider, as mere tokens, the following:3 No. 164 (PRO C 1/1/21) is said to be a petition “of Rauf atte Ree of Lancashire” (204), though the document gives no indication that Rauf had ever set foot in the county: it describes him as a tenant of the king in the Duchy of Lancaster, and ·fermourÒ of Ramsden Hall in Essex.4 Apparently, the editors just assumed that the lands of the Duchy must have been co-extensive with the county of Lancashire, whereas they lay very largely outside it, as consultation of the standard work would have shown (Somerville 1953). So would the Guide to the Contents of the Public Record Office (I: 183), a resource unnoticed by these writers, which reports that “Nearly every county in England and Wales is in some way represented in the territories and jurisdiction of the Duchy of Lancaster”, among them “the vast estates of the earldoms of Hereford and Essex”. The petition itself shows Rauf atte Ree only at Ramsden Hall; the party he first complains of is John Wethy of Essex, whose maintainers seem to be local men; and the dispute concerns the king’s ·Maner’ of Ramesden Hall in EssexÒ. From the index to The Place-Names of Essex, hardly an obscure reference, the surname atte Ree can be traced to Ree in Loughton, in the south-west of the county; it was the home of William atte Ree in 1327 (Reaney 1935: 66–67). Great Burstead, Ramsden Bellhouse and Ramsden Crays are adjacent parishes, a little north of Basildon and lying some fifteen to twenty miles east from Loughton; in Great Burstead anno 1484 is recorded Danyell land, possibly the

3.
For the sake of clarity, citations of Middle English from Anthology are enclosed in double quotation marks, as are all other citations (unless indented) from published works. Citations from manuscript are enclosed in angle brackets. Anthology’s texts have for some time been available on the Internet, at http://etext.lib.virginia.edu. At the time of writing, some twentyfour discrepancies with the printed version are said to have been corrected. Collation with the MSS is apparently not envisaged. 4.
Anthology prints “fermor”, with “r” for the abbreviation of final ur, as throughout the text. So “Chartor”’ 8, “modor”’ 9, “othor” 10, etc. The Glossary (s.v. fermor) explains the word as “usually tax farmer”; but commonly it means ‘leaseholder’ or ‘tenant’ (cf. MED s.v. fermour), and clearly does so here, for Rauf is later described as “the kingis tenant of his Maner of Ramesdon hall” (Anthology 206, 35–36).

Chancery Standard

“Danyell tenement” from which Rauf claims to have been ejected (Reaney 1935: 147). The language of the petition is recognizably of Essex; there is nothing to link it with Lancashire, and its past participles in I-, beside es for is, need not be wondered at.5 Anthology No. 178, PRO “E28/51/A6” (reference ill-formed, read ‘E 28/51/6’). The document is a petition against the abbot of “Stratflere”, in the names of the poor tenants and commons of “Vchayron per
cell of the Shyre of Cardygan”. Anthology makes them “the residents” without distinction, promoted by the Index to “citizens”, and of a place called “Uchayron Percell”, which supposed municipality is unidentified. The petition’s wording, however, with “of” rather than “in” the shire of Cardigan, indicates that “per
cell” is not the second element of a place-name, but means ‘parcel, part’: Vchayron is the district north of the River Aeron, in which lies the oppressor’s “Stratflere” (Strata Florida, correctly indexed).6 The petitioners request a privy seal addressed to the chamberlain of South Wales, commanding him, in Anthology’s text, “to respite the leve of the comune
s” (17); but this “comune
s” commons is for MS ·som
mesÒ sums, in which the first letter cannot possibly be c, and ·o¯meÒ follows clear as daylight.7

5.Anthology gives the date of the petition as 1416–1421; it is endorsed 10 July 9 Henry [VI]. The supplication from Rauf atte Ree cited in a Signet letter of 20 July [1420] (printed as Anthology no. 82) may be this document. For the distribution of es, see Atlas I: 338, Map 134. The Atlas reference to this document (“C1/1/4”in I: 95b, 119a), is in error. 6.For a map, see Rees 1967, Plate 28 (Uwch Aeron). For the ME form, cf. Welsh Vch Ayron in fourteenth-century MSS of Brut y Tywysogyon (Jones 1955: 264, sub anno 1277). 7.The date of the petition is not declared, though Anthology gives it as 1429 without comment. The evidence is the “(Action recorded in French)”: ·lett
res ent feurent fai
tes le .x. iour de december lan &c’. viijeÒ, i.e. letters [under the Privy Seal], as requested by the petitioners, were made on 10 December in the eighth year of the reign, which internal evidence establishes as that of Henry VI. The text is not complete: between “Bath” 4 and “Cromewell” 4, the torn right edge of the document intervenes, and about fifteen millimetres of text are missing. Anthology prints the text as continuous, without annotation. Loss of text is reported between “comaundement” 5 and “lordys” 5, with the explanation “torn”, but after “comaundement” are thirteen millimetres of writing before the tear: ·the …Ò, whence illegible without ultraviolet light. In 1 and 2, “owr” is MS ·yowr’Ò, with ·yÒ not very thoroughly erased. Anthology’s text ignores the flourish on final ·rÒ in ·yowr’Ò 3, 4, 8 (×2), 9, 10 (×3), 11, 12 (×2), 13, 14, 16; and in ·answher’Ò 7 (×2). Final ·eÒ is ignored in ·mater
eÒ 15. Capitals are misread as minuscule in ·RuleÒ 6, ·ComaundyngÒ 12. The standard abbreviation for ur in ·afturÒ 8, ·fadurÒ 8, is misread as “e
r”. “determynd” 18 is for ·deter
myndÒ, “afore” 13 for ·a foreÒ. “tho” 1 is for ·theÒ (the theta-like ·eÒ is unambiguous), “lawe” 12 for ·lawesÒ.

7

8

Michael Benskin

No. 237 (“C146/C3584”, read ‘C 146/3584’) is an indenture of sale between (i) Robert Clopton, citizen and alderman of London, and (ii) John Gerveys of Bury St Edmunds; the subjects are Robert’s landed property in Bury and neighbouring parishes, with security of an annuity on Robert’s manor of “wyndey” in Cambridgeshire. The document contains no statement as to where it was written or sealed. Its language is not markedly local, but “sweche” (9, 11), with “wheche” (11) and “ony” (13, 21), suggests Suffolk rather than London, so agreeing with the weight of its local content. Anthology confidently styles it “London indenture”. No. 240 (“E40/A7651”, read ‘E 40/7651’) is an indenture of lease, between (i) two citizens of London, and (ii) a merchant of Northampton. The subjects are property in the town of Northampton, owned by the Londoners, which the merchant takes for twenty years. There is no statement in the contract as to where it was drawn up or sealed, whether London or Northampton; anno 1455 (“1456”) its language would be out of place in neither. Anthology labels it, without comment, “Northampton indenture”, which may or may not be right; the scrivener’s name, Plumer, is known from the London company (Steer 1968: 10, 21). At least the place-name is correct; whereas it is not helpful to be told that a contract dated at Stamford and concerning property in Stamford is a “Lincoln indenture” (Anthology no. 238, PRO E 101/504/19). Lincoln is a city; its county is Lincolnshire, in which Stamford lies some forty miles away. The writer of the Stamford document is even charged with not knowing his own language. Anthology reports the sealing clause as “In wittenes where of the seid Iohn & william has putte yeir seles…”, which Fisher and Richardson call “a downright solecism”, a lapse “from what must have been regarded as the formal eth ending for the 3rd person” (45). They would corrupt a well-formed text. In the dialects of the north and the North Midlands, and of parts of the East Midlands, the suffix of the present indicative plural ends normally in -s, unless its subject is a directly adjacent personal pronoun.8 The contract, between parties respectively of “wisbeche” (Cambs.) and “(beche) Saifterton in ye Counte of lincoln” (not found: for Bridge Casterton by Stamford?), is dated at Stamford, for building work at Stamford, and there is nothing to link it with

8.
Wright 1905, §435; Mustanoja 1960: 48–52; Bennett & Smithers 1968: xxxvi; Benskin & Laing 1981: 93–95; McIntosh 1983. No. 223 in Anthology, a petition from a former sheriff of Westmorland, shows the same construction (19) in ·which’ charges amountes to þe somme of iiijxx xiiij l’iÒ, but its writer escapes censure; the language is “Chancery … (except ‘arn’)” (75).

Chancery Standard

any other part of the country. For the writer, therefore, Iohn & william has rather than Iohn & william haue was in all likelihood the correct form.9 No. 241, PRO “C146/C273” (reference ill-formed, read ‘C 146/273’). Anthology describes the document as an indenture, as does its author, but as may be expected of a will and intent (here, instructions to feoffees), in physical form it is a deed poll. It is said to begin “Be hit y knowe to all maner mene that wher. aŠ Iohn Cheyne. Esquyer. haue by my dede …” That “aŠ’” should have given pause: for the supposed as, neither OED nor MED lists any such form.10 The assumed syntax also jars, in that the subject (“Iohn Cheyne”) and verb (“haue”) are disjunct. Against which, the MS has ·.Y.Ò not “aŠ’”, that is, a commonplace form of the first person pronoun, in concord with first singular ·haueÒ.11 In “doe some soth to the church” (11), the editors think “soth” means “good deed” (Glossary, p.384, s.v.). This looks like homespun etymologising from the primary sense ‘truth’ listed alongside; OED gives no ground for it. The MS, however, has ·sethÒ, with ·eÒ theta-like and clear. Clear also is the meaning: (as)seth is ‘satisfaction, amends’ (OFr. as(s)et), and the expression is a set phrase (see OED, s.vv.; MED s.v.). Seemingly absent from the Glossary is “vre” (12); there is no indication as to what the editors thought it meant. The form in MS is rubbed and hard to read, but vre is indefensible; the reading appears rather to be ·NeÒ. Certainly a conjunction is required: “… with oute subsidi of the sayde londis vre as well tho seruauntes …”. In the next line (13), “al tho” is ·as thoÒ, as the syntax implies. For “divise” (15) the MS reads ·areiseÒ, just possibly ·aroiseÒ (see OED and MED s.vv. araise(n), senses 4 & 10 respectively): the feoffees are to levy and raise the revenues and profits of the manors, and ·tho to take & deliuer
e to my Executour
s soe that they ther with mowe pay all my sayd detteÒ, which instruction of itself precludes the feoffees from “devising” them. For “rentholds” (9), the syntax alone ought to have been warning. Anthology’s text reads (p. 302):

9.See Atlas I.467, Map 653. The domain of the pres. indic. pl. in -s extends southward well into the area about Wisbech. 10.So Glossary, s.v. as. Neither is it in Atlas (IV: 61–62), which reports the variant forms of as from several hundred northern and midland sources. 11.The first loop cannot stand for a, for in this hand the form has only a single-compartment; it is made with the upper stroke slanting, and the resultant lobe is spurred. ·.Y.Ò is written over a crease in the document, and a photograph would lose most of the letter.

9

10

Michael Benskin

… in all my landis and tenemen
ts withinne the paryssh of Pynhoe in the Schere (9) of Deuon
sshire in the forme as hit apperith in the sayde dede] & rentholds for as moche as I am in dette to diuer
s per
sons …

It is unclear what Fisher and Richardson think “rentholds” means; their glossary records it merely as “n[oun]”, and they omit it from their section on vocabulary “Not in OED or MED”, although it appears in neither.12 If a noun, the word is seemingly qualified by “sayde”, and so ought to repeat or restate, but no antecedent is to be found. In MS the word is a conjunction, ·Neuer
thelesÒ. In the dating clauses of this document, the “fourtieth” year of Henry VI is twice reported (2,22). That would indeed give Anthology’s date of 1462 (the months are April and May), but in defiance of the accession of Edward IV. The MS, however, has ·fourtethÒ, which is not ‘fortieth’ but ‘fourteenth’, a normal reflex of OE feowerteoþa.13 The year of the document, written on 1 May, is accordingly 1436; it makes a difference.14

12.So p. 36. At the time of publication, MED had reached only preven, but the relevant fascicule later published indeed has no record of it. OED’s closest congener is rentholders, cited from a text of 1657 (s.v. rent sb1 4b). 13.For fortieth, OE feowertigoþa, MED lists forms in -eoþe but not -eþe
; the vowel of the suffix is otherwise i(e) (s.v. fourtithe ord. num.). OED records fourteth ‘fortieth’ from the sixteenth century, in view of which its absence from the ME record may be accidental. 14.There are many other errors in this text. Editorial treatment of flourished final ·gÒ is mistaken and inconsistent: usually the flourish is reported as ·eÒ, but not italicised: so “beringe” 1, “longe” 3 & 6, “berynge” 6, “kynge” 2, 7 & 21, “thinge” 14; once it is ignored, “kyng” 2. So also the flourish on final ·rÒ: it is printed as if MS ·eÒ in “Schere” 4 and “sshere” 6, but ignored in ·yeer’Ò 7 & 21, ·modir’Ò 10 & 13, ·memoir’Ò 12, “hir” 13 (·hur’Ò). So much for the rules stated in “Editorial Principals” (sic) on p. 79: “Every expanded e has been italicized but not every flourish has been expanded into final e”. The standard abbreviation for ur is misread as “er” in “apper
tena
ncis” 4 and “apper
tena
nce” 5 (recte ·appur
tenaun
ceÒ, with au clearly spelled out, and a tilde for n). It is misread again in “Executor
s” 15 for ·Executour
sÒ, though it is the same sign as in the correctly-reported ·endentur
isÒ 20 and ·endenturÒ 21. The saint’s name in line 7 is inserted between the lines, and has a capital: ·MathewÒ, not “mathew”. Three personal names are misread: “Palain” 7 is ·PaltonÒ; “James Chideleoh” and “william Chideleogh” 8 are in MS surnamed ·ChideleghÒ, with the lower loop of ·gÒ in James’s surname partly erased by wear. (Why “James” has “J”, but “Iohn Cheyne” 1 has “I”, is unclear: the MS letters are the same.) The dots printed after “Chideleogh” 8, without annotation, correspond to an erasure immediately following the surname, and then the words ·and diuer
s othir’Ò. “mene” 1 is for ·men
neÒ, “feoffed” 7 for ·enfeoffedÒ, “knyght” 8 for ·knyghteÒ, “landis” 8 for ·londisÒ, “withinne” 8 for ·with inneÒ, “Deuon
sshire” 9 for ·Deuen
sshereÒ, “per
sons” 10 for ·personisÒ, “some” 11 for ·som
meÒ, “hir” 13 for ·hur’Ò, “tharre” 13 for ·thaireÒ, “detts” 16 for ·detteÒ. In 20, “made” looks rather to be ·maid’Ò, which ends clearly in ·d’Ò not ·eÒ, apparently preceded by a minim.

Chancery Standard

Dating is not one of Anthology’s strengths. The year of no. 240 (“E40/ A7651”, recte E 40/7651) cannot be “1456” (p. 300) if the dating clause reads, as indeed it may, “the .iiij. day of Decembre In the xxxiiijth yere of the reigne of kyng henry the sext” (30–31): anno domini, the year must be 1455. (Reference in the same text (24) to the “vxth” year of the reign of Henry V is perplexing; by Roman count, “vx” is ill-formed,15 and fifteen years after his accession this king had been dead for nearly five.) In no. 234, 24 November 4 Henry VI should be rendered 1425, not “1426” (p. 295). And so on. The arithmetic is sometimes confusing, to be sure, but it can be avoided altogether by consulting a table of regnal years. Cheney’s Handbook of Dates for Students of English History, and the Handbook of British Chronology are the standard recourse, neither listed by Anthology.

3. Chancery language, Chancery hand Fisher would have it believed that ‘Chancery Standard’ is bound up with Chancery hand. In principle there is no reason why it should not be, though the facts, as will appear, are otherwise. The claim as Fisher presents it, however, is immediately and fatally flawed, because what Fisher means by ‘Chancery hand’ is not the hand of Chancery documents. Of Anthology’s four plates alleged to illustrate the Chancery hand — a variety of anglicana — only one approximates to it (Plate VI); the three others are in secretary (Plates II–IV). Anthology’s opening section, ‘The Chancery Hand’ (3–5), is an extraordinarily confused account: Fisher (the responsibility is his: Anthology ix, Fisher 1988: 268–270) thinks he knows better than the palaeographical works he cites, but has not understood them. He writes with assurance about “the business script that had developed in the 13th century” in England, but fails to recognise that it was a form of “this native cursive”, developed in peculiarly English fashion, that became the fifteenth-century Chancery hand. Chancery hand for Fisher is “this court or anglicana hand” of the thirteenth century “modified by the continental business hand” in the second half of the fourteenth century. He says that this is a new script that Malcolm Parkes calls ‘secretary’, but Parkes’s ‘secretary’, far from being a modified native script, is in Parkes’s own words “A new cursive script … introduced into this country, a script which possessed letter forms and 15.Cf. no. 223’s rendering of ·iiijxx xiiij l’iÒ, four score and fourteen pounds, as “xx
iiij xiiij li” (19).

11

12

Michael Benskin

stylistic features which had no counterpart in the traditional English handwriting” (Parkes 1969: xix). Heedless of which, Fisher declares that Parkes’s secretary so described “we shall refer to as Chancery”, claiming that the indebtedness of this English Chancery hand to the French may be seen by comparing the 1366 letter by Charles (Plate V) with the exemplar by Hammond … the signed letter by Henry’s [sc. Henry V’s] Signet clerk Robert Shiryngton … and the one probably by William Toly.

In truth, all are examples of the secretary hand, and for Hamond, as a clerk to the Privy Seal (see pp. 14–16 below), secretary was the accustomed usage.16 So it was for the clerks to the Signet, and as Parkes (1969: xx) had made clear, the practice of these two offices normally distinguishes at a glance their official writings from those of Chancery.17 Parkes’s ‘secretary’ differs from anglicana — of which the Chancery hand is a variety — “in the duct of the script and in the treatment of strokes” as well as in its letter forms. Parkes 1969: xix: First, the duct of the new script was based upon the regular antithesis of broad strokes and hairlines placed in different diagonals according to the angle of the slanted pen, thus giving to many of the hands a characteristic ‘splayed’ appearance. Secondly, angular broken strokes appeared in places where in most other scripts one would expect to find curved ones.

The main differences between the letter forms of the two scripts are here illustrated in Figure 1. In the course of the fifteenth century, the secretary forms infiltrate anglicana, in bookhands as well as in documentary hands. Fisher, for reasons that can only be wondered at, having renamed the secretary script as ‘Chancery’, accordingly claims that the spread of secretary letter forms betokens the influence of Chancery, and with them, naturally, went ‘Chancery Standard’. This is set out in some detail in his article of 1988, which claims to demonstrate Chancery’s influence in the language and hands of the Piers Plowman manuscripts. Palaeographically this supposed demonstration is nothing of the sort, because the letter shapes in question do not belong to Chancery hand; Matheson (1986), reviewing the Anthology, had by then made clear that they do not, but

16.The contrast between these and the hand of B. L. Cotton Vespasian F III, fol. 8 (Anthology Plate V), possibly that of Henry V himself, is said to be a “contrast between these set professional hands and a non-professional hand” (3). So it is, but no less is it a contrast between secretary and anglicana. 17.Cf. Matheson 1986: 648, with additional references to published facsimiles.

Chancery Standard

Anglicana (Chancery)

Secretary (Signet, Privy Seal)

‘a’

‘g’

‘r’ short ‘s’

‘w’

Figure 1.

the article’s bland reference to Matheson’s “strictures” (268, note 6) shows Fisher still oblivious to the difference between the two modes of script. All that the Piers Plowman article shows is the spread of secretary forms into anglicana book hands, which is a commonplace. As the fifteenth century wears on, these forms start to appear even in Chancery documents, but Chancery was a slow recipient, not an initiator in their use. If royal clerks were responsible for their dissemination at all, then they were the clerks of Signet and Privy Seal. The supposed link between Chancery hand and Chancery language is one of Anthology’s abiding themes, and the hand of each of its texts is labelled ‘Chancery’ or ‘Non-Chancery’ accordingly.18 It might be imagined, in view of Fisher’s redefinition of the term, that documents said to be in ‘Chancery hand’ would at any rate all be in secretary, but they are not. Anthology’s labellings are capricious at best, seemingly dependent as much on evaluations of language and documentary origin as of script. Others imply perceptions of handwriting not easily reconciled with palaeographers’ practice hitherto, as when the anglicana currens of the petition of the Folk of Mercerye is described as “Textura/ Chancery” (197), and that of the Parliament Rolls is said to be “very much like that which M. B. Parkes [1969: 14–15] designates bastard secretary” (24).

18.Texts 106–241 are individually so labelled. On p.4, texts 1–105 are said collectively to be in “Chancery hand”.

13

14

Michael Benskin

It should not be thought that this argument is about terminology for its own sake, objection to departure, merely because it is novel, from established usage. The issue here is formal categories and their defining characteristics, not the names by which they are called. One such category is a mode of script — Parkes’s ‘secretary’ — almost never used by Chancery clerks for official business. To treat this as if it were the same as the script that they did use, and call both ‘the Chancery hand’ indifferently, is a licence for discerning the influence of Chancery where in truth there was none.

4. Documentary categories Anthology’s texts begin with ‘The Signet Letters of Henry V’, by page over a quarter of the whole. Of these, no. 40, a letter from the abbot and convent of St Mary Graces, London, is acknowledged as an exotic (52, n. 2), but its forms should not have been allowed to contaminate the Signet content of the Glossary. The editors claim (6) that Only the 103 Signet letters preserve the original work of his [sc. Henry V’s] personal clerks. Our collection has been restricted to these and to other primary documents.

Why, then, do they include the text of a Signet letter from a version entered by a clerk of Chancery on the Close Rolls (no. 81, C 54/270 m. 17d)? It is no more original than the Signet letters copied by “civic and ecclesiastical institutions”, and which Anthology (6) abjures, but as a Chancery enrolment this is the only letter in the collection that is properly described as in Chancery hand. Failure to distinguish between this and the secretary hand of all the Signet clerks’ originals is further testimony to the editors’ palaeographical deficit: in the eyes of Fisher and Richardson, the hand is Chancery throughout. Text no. 1 in Anthology, PRO C 81/1326/36, is a more revealing case. The document is a letter of credence for the king’s commissioners, dated ·þe .xx. day of Iuyll’Ò; the year is lacking. Their business is not stated, though it looks to have been a national raising of forced loans. The commissioners were to treat with diverse private persons, with corporations, and with the heads of religious houses. They had therefore to be provided with variant copies of the letter, its salutation altered according to the recipient’s status; and below the text of the letter, and in the same hand, the appropriate styles are set out in formulary fashion, concluded by the flourished signature ·HamondÒ. A facsimile of the

Chancery Standard

document appears as Anthology’s Plate II, where it it is described as “Exemplar (from the Signet of Henry V?)”; the edited text repeats this designation, adding “Early Date?” (84); and in the ‘List of Documents’ (67), it is “The exemplar”, first among ‘The Signet Letters of Henry V’. An unwary reader might imagine that it was in some way the model for those that follow. Be it noted, therefore, that it is an exemplar only in the sense that any letter of which multiple copies were ordered is an exemplar; most such, and they are commonplace, list the addressees at the foot. This one is unusual only in that it lists styles of address, and the apparently long list of intended recipients was written presumably on a separate sheet. Anthology assigns this letter to the reign of Henry V (early or otherwise), against Hall’s alleged “c 1483”,19 because “it is in the same sort of hand as that of the Signet clerks of Henry V and is in the same C81 collection in the PRO” (59, note 12). Against which, systematic palaeographical comparison with the Signet letters of Henry VI and Edward IV might have made for a less confident judgement of the handwriting, and even a cursory examination of the PRO class lists or the Guide to the Contents of the Public Record Office would have shown that C 81 consists largely of Privy Seal warrants, its 1,796 files ranging in date from 1230 to 1485.20 The file containing this “exemplar” (C 81/1326) is late, consisting otherwise of documents from 1482–1483, whence Hall’s rejected date.21 All of these, however, are Latin warrants under Privy Seal for writs of Fiat protectio cum clausula (safe conducts), and have nothing to do with the “exemplar”. Hamond’s signature appears on none of them, they need therefore have no bearing on the “exemplar’s” date, and we may accept that the document has been misfiled. But it beggars belief that Anthology should count it among the Signet letters at all, for the letter’s own dating clause proclaims it as Privy Seal, ·Yeuen’ vndre our’ priue seelÒ. What, then, of Hamond? That he is absent from Otway-Ruthven’s monograph on The King’s Secretary and the Signet Office in the XV Century, and from Kirby’s Calendar of the Signet Letters of Henry IV and Henry V, might have given Anthology’s compilers pause. Indeed, he was not a Signet clerk, but, as the

19.Hall (1908: 98) is misrepresented: he gives “21 July, ? 22 Edward IV”, which is 1482, and ‘?’ does not mean ‘circa’. 20.Guide I: 43. 21.The date of only one document is illegible. The dates of the others are from 12 December 22 Edw. IV to 26 March 23 Edw. IV.

15

16

Michael Benskin

“exemplar” implies, a clerk to the Privy Seal, and he was active long after Henry V’s demise: his signature, identical with that in Anthology’s “exemplar”, appears on five Privy Seal warrants issued in July and August 1446.22 The standard reference works enable a little of his biography to be established. The Calendar of Patent Rolls records only one fifteenth-century Privy Seal clerk of that name, John Hamond; he is first noticed in 1446, in connexion with his presentation to a canonry and prebend at the church of St John the Baptist beside Clare in the diocese of Norwich, benefits implying a career by then well-established.23 Council documents show him in office at least until 1455.24 In the Calendar of Close Rolls, his only certain appearance belongs to 16 February 1462, when he was sent to the abbey and convent of St Benet Hulme, “to take, etc., as any other heretofore had”, presumably for his retirement.25 There is no certain later reference in the index to either Calendar. Suppose him, then, to have retired in 1462. It is possible that his career lasted more than forty years, and so began in the reign of Henry V, but it is unlikely that he rather than his seniors would have been laying down the “exemplar”s’ protocols at that time — a time, moreover, at which English for a Privy Seal letter would in any case have been extraordinary. No doubt by oversight, but properly as it turns out, Hamond finds no place in Anthology’s account of the Signet clerks of Henry V “whose names are signed” (9–11), and save in the Glossary his language is ignored. Anthology’s treatment of the “exemplar” is not an isolated blunder. Fisher and Richardson, for all their seemingly informed discussion of departments of

22.PRO C 81/1250, no’s 1, 24, 28, 34–35, for writs of Fiat protectio cum clausula. These were found by accident; an organised search would doubtless yield many more. 23.Calendar of Patent Rolls for 1441–1446, 435. Related documents are PRO E 28/76/1 (king to archdeacon of Sudbury et al., under Privy Seal, Westminster, 15 Sept. 24 Henry VI (1445), describing John Hamond as one of the king’s clerks in the office of Privy Seal); and PRO E 28/76/28, a supplication from Hamond dated 18 July 24 Henry VI (1446) (the schedule is E 28/76/29). Note also E 28/68/58 (undated, but the file is of 19 Hen. VI, i.e. 22 Sept. 1440 to 21 Sept. 1441), a petition of Henry Benet, John Hamond, priest, John Brewester, Richard Langport, and John Foston, clerks in the office of Privy Seal, for a grant of 50 marks as reward for services rendered during the last five years. 24.So in PRO E 28/87/3, draft of a Privy Seal warrant [19 Nov. 34 Hen. VI, i.e. E 28/87/23, draft of a letter under Privy Seal to the chamberlain of Chester and the auditors there, Westminster 12 Dec. 34 Hen. [VI], (in the same hand as E 28/87/19, and possibly 23); E 28/87/44–45, drafts of a warrant under Privy Seal, for payment of 25 marks to Robert Kyrkham, keeper of the hanaper in the Chancery, endorsed by T[homas] K[ent], 8 July 34 Hen. VI, i.e. 1456 (Hamond’s signature). 25.Calendar of Close Rolls 1461–1468, 105.

Chancery Standard

government and uses of seals, are contradicted again and again by the very documents they claim to elucidate. For example, they hold (54, note 18) that among those printed by Chambers and Daunt, in A Book of London English, “eight are missives from Henry V (i.e., Signet) and two others are from the Dukes of Clarence and Bedford (i.e., Privy Seal)”. Clarence’s letter is dated “at Maunt, Vnder our signet”, Bedford’s “vnder oure signet, at Rouen”: how can either of them be read as ‘Privy Seal’?26 Consider now the purported ‘Later Signet and Privy Seal Papers’ (no’s 106–160). Among these (17) The later Signet letters … [twenty enumerated] … are written in the name of the King by secretaries or members of the Council. As such, they fall in with the materials classified as Privy Seal.

So they do, nearly all of them, for the Council’s secretariat was not the Signet Office, but the Privy Seal. Of the twenty enumerated, two have dating clauses declaring them to be letters under signet, namely no. 108 and no. 157, but only no. 108 is certainly under the king’s; no. 157 is in the name of Queen Margaret, not the king, and she had her own secretary. Possibly under the king’s signet is no. 109 (which is damaged). For the rest, even in lack of knowledge of the workings of the Council, passing them off as letters under the Signet cannot be excused. Eight of the twenty are explicitly dated under Privy Seal, and are even labelled thus in the headings of Anthology’s texts (no’s 132–134, 136, 151–152, 154–155). Four more, correctly though for reasons unstated, are likewise headed ‘Privy Seal’ (no’s 117, 130, 135, 137). Though unrecognised in Anthology, form, content and preservation show the five remaining to be Privy Seal as well (no’s 106, 107, 148, 150, 156).27 Thus of twenty purported Signet letters,

26.Text cited from Chambers & Daunt 1931: 80–81 (no. XIX), 85–86 (no. XXIV). The letters survive only as copies in the Guildhall Letter-Books. 27.No. 106, said to be “Signet”, is addressed to the treasurer and barons of the Exchequer directly, and so must be Privy Seal: the Exchequer did not accept the Signet as authority for its issues, but required a warrant under either the Privy Seal or the Great Seal (cf. Guide to the Contents of the Public Record Office, II: 237). The normal practice was hence for grants initiated under the Signet to take the form of a warrant to the Keeper of the Privy Seal, instructing him to make a like warrant addressed to the treasurer. No. 107, “Signet or Privy Seal”, is a Council draft. No’s 148 and 150 were written by Adam Moleyns, as noted; but he was then clerk to the Council and secondary in the Privy Seal. No. 156 is an indenture for royal service, and hence issued from Privy Seal. All of these are from files of documents received by the Exchequer (PRO E 28, Exchequer Treasury of the Receipt, Council and Privy Seal), which of itself all but excludes the Signet as their proximate source.

17

18

Michael Benskin

seventeen are under Privy Seal, and twelve of them are so headed in Anthology. The offices thus confounded, it may be wondered why the editors observed any distinction at all.28

5. The language of the clerks to the Signet of Henry V According to Anthology, Chancery Standard first appears in the Signet letters of Henry V, “all of which are by Chancery clerks” (46). Chancery and Signet thus confounded, the language of these letters is discussed clerk by clerk, and in some detail. Nineteen letters, signed by Shiryngton, are accepted as being in his own hand; no letters signed by him are ascribed to any other clerk. Seven unsigned letters are regarded as “probably” in his hand, and four more as “possibly” so.29 The date range is given as 12 August 1417 to May 1422. Shiryngton’s language is said to display “only a few non-standard forms: þey, þaire, þeym, but hem (2.3, 57.10, 64.10)”. Apparently the þ-forms are standard whereas hem is not, and hem is found thrice against regular þeym. The Anthology’s texts, however, tell a different story:30

28.In the glossary, the “Later Signet and Privy Seal Papers” no’s 110–160 are lumped together as ‘P’, so excluding the one certain and the one probable Signet letter, besides two Privy Seal documents. But even then, “Later Signet and Privy Seal Papers” cannot be trusted: it includes documents having nothing to do with Privy Seal, among them Queen Margaret’s letter to the chancellor (no. 157), and letters to the king from Thomas Longley, bishop of Durham (no. 112), Henry Beaufort, bishop of Winchester (no’s 113, 127), Humphrey, duke of Gloucester (no. 125), and Richard, duke of York (no’s 126, 129). There is even a petition from the Early Chancery Proceedings (no. 160), in the name of a Chancery clerk (Anthology “Clerk of the Crown”), Richard Sturgeon. 29.So p. 10. Even here, there are discrepancies. In the survey of Shiryngton’s œuvre, the hand of no. 41 is “probably” Shiryngton’s (10), but is only “possibly” so in the note to the edited text (105). Against which, the other letters thought in the survey to be “probably” in Shiryngton’s hand are labelled “Shiryngton’s hand” without qualification in the notes to the texts (no’s 54, 65, 72, 77, 78, 79). 30.Unless otherwise stated, the forms here cited are those of the Anthology, which may or may not agree with the MSS.

Chancery Standard

Letters signed ‘Shiryngton’ (×19) þay ×1 (no. 64) þei ×1 (no. 57) þey ×1 (no. 64) þaire ×3 (no 57 ×2, no. 64) þaym ×1 (no. 62) hem ×4 (no. 57 ×2, no. 64 ×2)

Letters unsigned, but Shiryngton’s hand “probable” (×7) or “possible” ((×4))

þey ×1 ((no. 6)) þaire ×1 (no. 77) hem ×1 ((no. 2))

These attestations are hardly enough to fix norms of any kind. Best-recorded is them, which occurs six times, and the allegedly non-standard hem accounts for five of them. The survey’s þeym, which the Glossary lists from none of Henry V’s signet letters (391), is þaym in the edited text; þay the survey ignores. Shiryngton’s form for not, the survey continues, is “usually nat, but not (60.7, 72.4)”. The texts show otherwise: Letters signed ‘Shiryngton’ (×19)

Shiryngton’s hand “probable” (×7) or “possible” ((×4))

nat ×5 (no. 59 ×4, no. 64) nat- ×2 (no’s 59, 76) not ×3 (no’s 60, 61, 62) not- ×1 (no. 75)

nat ×2 ((no. 39)) (no. 77) not ×1 (no. 72) not- ×2 (no’s 54, 78)

In face of not(-) seven times, nat(-) needs more than nine occurrences to count as “usual”; four of the nine are in the same letter. Allegedly, Shiryngton’s though is “always þow/þoo” (10); but þow is found twice only, in no. 62 (signed) and no. 6 (“Hand possibly Shiryngton’s”), þoo is found nowhere in these letters, and not even the Glossary records it. Next to Shiryngton, William Toly is the most frequent signatory to Henry V’s signet letters. To the fifteen he signed, Anthology adds eight as “probably” in his hand (though the texts are labelled “Toly’s hand” without qualification), and one as “possibly” so; the date range is given as 22 September 1417 to 18 June 1422 (10). In Toly’s usage “There are a few non-standard forms: … þay, þaim, þaire, but hem (50.4, 58.5)” (11).

19

20

Michael Benskin

Letters signed ‘Toly’ (×15) þay ×3 (no’s 67, 84, 93) thay ×2 (no’s 84, 93) þaire ×1 (no. 66) here ×1 (no. 93) þaim ×1 (no. 66) hem ×2 (no’s 50, 105) hem ×1 (no. 93)

Toly’s hand “probable” (×8) or “possible” ((×1)) thai ×1 (no. 46) they ×3 (no. 4) þaire ×1 (no. 58) taire ×1 (no. 58) hem ×1 (no. 58)

Again, the survey conceals tallies that are nugatory and misrepresents them. Toly writes hem not twice but four times, against þaim only once, yet hem, the dominant them-form among Henry V’s signet clerks, is again labelled “nonstandard”. Ignored are they (thrice) and the singletons thai and here, as also is taire (58/7), which evades the Glossary as well (the MS has ·þaireÒ). We are to believe Toly’s “style and orthography nearly identical” to Shiryngton’s (10), but the not-forms escape notice: letters signed by Toly show not six times (no’s 50, 67 ×2, 82, 93, 105), letters “probably” in his hand show it thrice (no’s 30, 46 ×2), and nat appears only twice over his signature (no. 93). Even this much vitiates the survey (11) of Thomas Andrew’s usage, whose “forms are quite regular: þaire, þayme in the postscript to 53, not in Andrew’s hand; not (74, 80) vs Shiryngton’s and Hethe’s nat”.31 So much for Shiryngton’s seven instances of not(-); and what is the relevance of þaire and þayme when they are not even in Andrew’s text? As to “Hethe’s nat”, there are two occurrences, both in letters only “possibly” in his hand (no’s 18, 45); against which, not(-) is found five times in four letters “in Hethe’s hand” (no’s 13 ×2, 29, 34, 38).32 There are four letters signed by John Depeden, all said to be in his hand; their date range is given as 7 March 1421 to 31 July 1421 (11).

31.The texts show not(-) in both of the signed letters (no. 52, no. 80 ×2) and both of the letters probably (11, but only “possibly” in the notes to the texts) in Andrew’s hand (no. 53 ×2, no. 74 ×2). 32.So styled in the notes to the texts. Hethe’s hand is reported only as “probable” in the survey (11). He is there said to write “only hem (13.3)” for them, but no. 88, signed and in “Hethe’s hand” (132), has þaym.

Chancery Standard

Depeden has several non-standard forms: her [sc. their] (90.2, 96.6) but their (96.8); swich (90.3) but such (90.6); youen (90.7) but yeuen elsewhere; Wurshipful (94.1); whech (95.2); ar comen (96.3).

Depeden’s given is “elsewhere” yeuen thrice (no’s 94, 95, 96), but also yiuen (no. 94), which is not reported. Beside Wurshipful, note Worshipful (no’s 90, 95); beside whech, note þe whiche (no. 96). Ignored are the six writings of them, each the “non-standard” hem (no’s 90, 96 ×5).33 Thus on the evidence of four brief letters, Depeden’s usage shows less internal consistency even than Anthology makes out; had the accidents of employment been otherwise, and more of his letters been known, his language might have commanded the same attention as Shiryngton’s and Toly’s. The examination could be prolonged, but its conclusion is sufficiently clear. Anthology’s survey is not merely inaccurate scholarship: it reports an ideology rather than its texts. The Signet clerks’ diversities of usage are represented as mere incidentals in an implied uniformity, ‘Chancery Standard’ not so much in the making as almost ready-made. An ostensibly judicious evaluation is in truth haphazard and tendentious, concealing and playing down such differences as might be expected between the usages of any dozen or so London residents of the time, whether from diverse institutions or from none. Had Henry V’s Signet clerks really been concerned with institutional spelling norms, then the word England would surely have been a prime candidate for fixity, whereas their letters show at least seven variants: Eng(e)lond, England(e), Engeland, Ingelond, Ingeland. The most remarkable feature in all of this, however, is yet to come: namely, that in spite of the confident labelling of this or that form as ‘Chancery’ or ‘non-Chancery’, the criteria for what counts as ‘Chancery’ and what does not, are nowhere defined. Tacit reliance on the eight categories of Samuels’s Type IV inventory is not at issue. A few of the “more outstanding” features that distinguish Type IV from Type III are far from constituting a description of Type IV, by whatever name it is called, and among the forms of even those eight categories, Anthology does not accept all: gaf 3sg. gave is dismissed as belonging to the sixteenth century rather than the fifteenth, and her their is described as “non-Chancery” for no stated reason at all. It might be supposed that recognition of a form as ‘Chancery’ or ‘non-Chancery’ would depend on a far-reaching examination of documents

33.For completeness, in terms of the present analysis: the they-forms are they (no. 96 ×2) and þey (no. 96 ×2), the not-form is not (no. 95).

21

22

Michael Benskin

initiated in Chancery and in the hands of Chancery clerks: how else, indeed, could Chancery usage be established? There is no sign that any such work informs either Anthology or any other of Fisher’s and Richardson’s writings.

6. ‘Chancery English’ and parliamentary petitions The compilers of Anthology claim (xii) that We follow M. L. Samuels in calling the official written English of the first half of the 15th century “Chancery English” although it emerged from at least four offices, Signet, Privy Seal, Parliament, and the emerging Court of Chancery itself. The generalized term is valid in an historical sense. By the end of the 15th century the term “Chancery” had come to be restricted to the royal courts of law, but until the departmentalization of the national bureaucracy into various offices, which began in the reign of Henry VII, Chancery comprised virtually all of the national bureaucracy except for the closely allied Exchequer. Thomas Frederick Tout, who made a life-time study of the workings of Chancery and its affiliated offices, begins his Chapters in The Administrative History of Medieval England [1920–33] by quoting Palgrave’s [1834] observation that “Chancery was the Secretariat of State in all departments of late medieval government”.

Again, the editors have not understood what they have read. R. G. Collingwood’s dictum is apposite: a body of knowledge consists not of ‘propositions’, ‘statements’, ‘judgements’ … but of these together with the questions they are meant to answer; … a logic in which the answers are attended to and the questions neglected is a false logic. (Collingwood 1939: 30–31; cf. MacIntyre 1985: 11)

Tout was concerned to rebut the claims current in his own day that there was conflict between different parts of the administration — that is, between the Privy Seal Office, controlled by the Council, the Signet Office, being the king’s personal secretariat, and the Chancery — and, read in context, it is plain he did not mean that these offices were identical. It ought to have been plain even out of context: ‘Secretariat of State’ does not mean “virtually all of the national bureaucracy except … the Exchequer”, and Palgrave’s “departments of government” does not mean ‘institutions’, but ‘activities’. It is not true that “the departmentalization of the national bureaucracy into various offices … began in the reign of Henry VII”: it had begun centuries before, with the Exchequer

Chancery Standard

during (if not before) the reign of Henry III. For the fifteenth century, “the emerging Court of Chancery” can refer only to the Chancery’s equity jurisdiction, and so excludes the whole administrative side; but English emerged from this side as well, as appears not least among the records of parliament. The statement that “By the end of the 15th century the term ‘Chancery’ had come to be restricted to the royal courts of law” is astounding, whether as defiance of the facts or mere ignorance,34 but by such means is the English of government identified as that of Chancery.35 “By 1430”, according to Fisher (1977: 881), “Chancery English had assumed its mature form”. (The criteria for recognising Chancery English, mature or otherwise, be it remembered, are nowhere stated in his work.) He presents “the beginning and end of a typical sample”, PRO SC 8/25/1238, a private petition in the name of Levin le Clerc, burgess of Ghent, to the king. It is not certain that it is the work of a Chancery clerk, but for the moment let that pass. The copy entered on the Parliament Roll (PRO C 65/90, m. 7, no. 21),36 however, is undoubtedly so, for the Rolls of Parliament were indeed “entirely in the hands of the Chancery clerks” (880). The text cited from the petition amounts to 209 words, which for ease of reference are individually numbered (881–882). A comparison with the enrolled version follows. For this purpose it is helpful to distinguish between the letters i and u, and o and e. Among Fisher’s errors of transcription are “Besechith” (7) for ·BesechuthÒ, “whiche” (74,133) for ·whucheÒ; and “rehersed” 209 for ·rehersodeÒ. Since, merely by misreading the document, he deprived it of those

34.Compare the introduction to the PRO Class List for C 66, the Patent Rolls (unpublished, but probably available by now on the Public Record Office website): The Chancery was the main instrument of general medieval central government administration. It embraced everything not the province of the Exchequer or Courts of Common Law, but also included a substantial common law administration of its own. In addition, from the fourteenth century, an equity jurisdiction (proceeding according to the rules of ‘equity and conscience’) began, to adjudicate where there was no common law remedy. The administrative machinery created was developed and amended over the centuries, but continued basically intact until 1873. (For “where there was no common law remedy”, however, read ‘where the court accepted the petitioner’s claim that there was no common law remedy’.) 35.No less culpable is the present writer’s reference, which delete, in Benskin & Laing 1981 (101 n. 14), to Chancery as a court of final appeal, whence the prestige of Chancery Standard. 36.Fisher’s “C65/90/21” (882) is ill-formed, “21” failing to specify whether capitulum or membrane is intended.

23

24

Michael Benskin

provincial stigmata, it was no doubt easier for him to regard its language as ‘Chancery English’.37 Examination of the full text confirms that it is even less ‘Chancery’-like than he imagined, and that to present it as a “typical example” of the species (881) makes nonsense of the idea that there was any such thing. The 3sg. pres. indic. -uth is confirmed in ·absentuthÒ, and u in which by four other examples; -ode wk. ppl. is found additionally in ·arerodeÒ, ·conteignodeÒ, ·payodeÒ, against -ed in ·conteignedÒ, ·deliuer
edÒ, ·passedÒ twice, ·priuilegedÒ, ·onsweredÒ; the ending of the noun pl. is -us twice, in ·goddusÒ; answer vb. is ·onswer-Ò in all four occurrences; after is ·aftur’Ò, once; (-)upon is ·oponÒ thrice, against -·vponÒ once (201).38 On this evidence, the westerly origins of the language are undeniable. The differences between the petition and its enrolment, Fisher claims, “give us a glimpse of the drift towards standardization”. Having purged the text by his own misreadings, he leaves the enrolling clerk with little to change, and so implies that he made only minor alterations. Even these are inaccurately reported. Allegedly, seid is changed to saide “in every instance” (882), which is true for the fifteen occurrences of said in Fisher’s extracts; but the enrolled text shows twenty-two more, among which are ·seidÒ twice and once ·seydÒ. The comparison is not helped by reliance on the printed version of the Parliament Roll in Rotuli Parliamentorum (IV: 372), a work long recognised as unreliable in detail even for historical purposes.39 Fisher, however, thinks it “a

37.Additionally, MS ·MÒ is misread as “m”, in ·Maner
sÒ 34, and ·MoneyÒ 153. “attor
ney” 52 is for ·attour
neyÒ; “aboue” 106 is for ·a boueÒ, “bonde” 134 for ·boondeÒ, “shall” 166 for ·schallÒ, “Chanceler” 174 for ·ChauncelerÒ, “thervpon” 201 for ·ther’ vponÒ, “fourme” 207 for ·four
meÒ. “xijc” 69 is for MS ·xijcÒ twelve hundred, but the page format perhaps forbade superscripts. A full text of the petition, variously inaccurate, appears in Anthology (no. 183). 38.The list is not exhaustive: nearly a quarter of the text is rubbed or torn, and in lack of ultraviolet light could not be reliably read. 39.Only those who have been forced to rely on it have held otherwise (e.g., Morsbach 1888:16). Contrast Weale 1931: 273, note 1: “The transcripts in the printed Rotuli Parl. are of little value for philological purposes, as the spelling in them often does not represent that in the actual documents”. The afterword in Rotuli Parliamentorum (vol. VI: 558) is disarming: There are also some Typographical Errors; from which no Degree of Attention on the part of the Printer could possibly secure a Work of this Nature and Magnitude: — Although they are not to be found in every Copy; and where they do occur, will be obvious to a careful Reader, it may be necessary to enumerate them. There follow forty-six corrections. A new edition, funded by the Leverhulme Trust, is to be published on the Internet.

Chancery Standard

reasonably faithful transcription of the enrolled version”, and that the editorial changes “do not affect the validity of the printed text as a basis for linguistic study” (883). He notices four departures from the manuscript, and as if to defend the printed text, discounts three as “probably lapses in transcription”. It appears that what matters is the implied pronunciation: the one discrepancy that attracts critical comment is Rot. Parl. “sich” for the manuscript’s alleged “suche”, which “is probably an accident, and in no sense an example of northern dialect”. The difference between an accident and a lapse is unexplained, but either way, Fisher is its ‘onlie begetter’. It is just not true that manuscript “u is in ligature [sic] with the s and was misread” (883): the manuscript unquestionably has ·sich’Ò.40 As to the dialectal implications, sich is more commonly midland than northern, and is one of the characteristic forms of Central Midland Standard (Samuels 1963: 85–86; cf. Atlas I: 321, Map 68, and IV: 17–19). Were Fisher right in holding that the petition is the work of a Chancery clerk, “prepared from dictation or based upon a preliminary draft brought in by the petitioner” (882), then it would follow, and in contradiction of his own claims, that at least one Chancery clerk of around 1430 was tolerant or even heedless of many non-Chancery forms. As it happens, however, there is no good reason to suppose that it was written by a Chancery clerk at all. Fisher’s evidence that “it had to be copied over and presented by a sworn clerk as attorney, since the Chancery clerks had a monopoly on presentations to Parliament and to the Chancellor” (882), is merely an essay in anachronism. He cites Holdsworth to the effect that such a monopoly came to be concentrated in the office of the Six Clerks, and that their control of the flow of Chancery business had become a scandal by the eighteenth century (882, n. 41, and so Anthology 58, n. 7), but none of this has any bearing on the mediaeval practice.41 For the fifteenth century, the most thorough examination of the parliamentary petitions is that of A. R. Myers, who concluded that professional scriveners were very often the authors of private petitions, and he reported some cases

40.The apostrophe represents the bar through the ascender of ·hÒ, possibly for e but probably otiose; Fisher expands it as “e”, without italics. 41.In the historical endeavours of Fisher and Richardson, anachronism is a besetting but hardly original sin. Cf. Tout’s review of Maxwell-Lyte’s massively authoritative Historical Notes on the Use of the Great Seal: “to illustrate a fourteenth-century practice a reference is made to a custom not established until centuries later … the book contains terrible pitfalls for the beginner” (Tout 1927: 632). Such practice has been pervasive in the historiography of the mediaeval Chancery: for a comprehensive review, see Haskett 1996, esp. 245–246, 251, 268–280.

25

26

Michael Benskin

where they were demonstrably so (Myers 1937: 387–388). His work is unnoticed in Fisher’s article, but Anthology makes amends, blithely dismissing it on the grounds that Maxwell-Lyte, Wilkinson, McIlwain, and Baildon “all agree that the similarity in style, form, and appearance indicates that the Ancient Petitions in the PRO (SC8) were drawn up by the Chancery clerks” (Anthology 58, n. 7). Even were they accurately reported, these authorities would be less than they appear: published between 1896 and 1929, they had no opportunity to consider Myers’s arguments, and for them as not for Myers, the topic was incidental to other inquiry. More recent work, by Lister Matheson and (independently) the present writer, confirms Myers. In Matheson’s words (1986: 649): The evidence suggests that the scribes of private petitions could be the petitioners themselves, professional scriveners (probably specializing in such documents), or moonlighting lower-level government clerks from different departments, as Thomas Hoccleve [Privy Seal] is known to have been, while Commons petitions probably drew more readily upon the resources of government clerks.

For the historian of standard English, the dissociation of such petitions from Chancery makes their language of more interest, not less, in so far as a larger constituency of writers, belonging to no one institution, is seen to converge, however unevenly, on broadly the same usage. So much appears from the Repository List in A Linguistic Atlas of Late Mediaeval English, which notices around 250 of the Chancery’s Ancient Petitions, most from before 1460 (Atlas I: 131–136). Of these, only a dozen are reported as in wholly or predominantly local language; in another thirty or so the local element is fairly muted; the remainder, well over two hundred, are described as “standard” (i.e., Samuels’s Type IV, or an approximation to it), “not in local language”, or “of no dialectal interest”. Even local language does not always tally with the domicile of the petitioner(s), and may thus reflect a provincial writer, active in the capital, with a diverse provincial clientele (cf. 7 below, and Atlas I: 49–50, on Early Chancery Proceedings). The language of the extant petitions is variously preserved or modified in the corresponding versions on the Parliament Rolls. In his earlier account of the characteristics of Chancery English, Fisher (1977: 880–881) held that The English of the Rotuli Parliamentorum may be taken as a yardstick against which to measure the evolution of Chancery Standard. The Rolls of Parliament were entirely in the hands of the Chancery clerks, and nothing they wrote could be more official or prestigious.

Chancery Standard

From “official” and “prestigious” to “Chancery Standard”, however, is a non sequitur, as would have appeared from better acquaintance with the manuscripts. The enrolled petitions, as with copies of Middle English texts generally, attest to varying scribal practice. Some clerks copied literatim or nearly so, like the enroller of the common petition concerning non-resident parsons and vicars presented to the Leicester parliament of 1424.42 Though some of his spellings differ, he reproduces its north-west midland language almost without exception, and outbids it in ·monerÒ for ·manerÒ: the petition has ·onÒ corresponding to his familiar ·anÒ in words of Germanic origin, and in Romancederived manner he assumed it once too often. The king’s reply to the petition, enrolled below and in the same hand, confirms that the language is not the clerk’s own: the petition’s ·hafesÒ has answers to ·hathÒ in the reply, as does ·beÒ by to ·byÒ. That the exemplars generally determined the language of this clerk’s copies appears from his other entries on the same roll (C 65/86), among which: m. 7, no. 18 ·swichÒ, ·swicheÒ such; no. 19 ·swichÒ; m. 13, no. 35 ·swechÒ, and ·mykelÒ, ·mychelÒ, ·mechelÒ much; king’s answer to no. 35, ·sucheÒ, ·mochÒ; no. 36 ·suchÒ, ·mocheÒ; m. 14, no. 46 ·suchÒ, ·sucheÒ, ·mychÒ; the enrolled petition concerning non-resident parsons (m. 13, no. 38) has ·sichÒ, ·sycheÒ, ·sucheÒ, and ·mykellÒ, ·mycheÒ. In so far as private or common petitions are the work of scriveners from outside government offices, the copies on the Parliament Rolls need not be expected to display the language of Chancery unalloyed. Petitions presented by the Commons, for whom Chancery provided a secretariat, are in principle more likely to do so. The distinction is important, though neither Fisher 1977 nor Anthology is alive to it. Anthology does, however, take a changed view of petitions copied onto the Parliament Rolls as a yardstick for Chancery English, stating that “The principle was, obviously, to represent the original exactly”, and recognising that the enrolling clerks might introduce “non-standard forms” (24). It is perhaps for this reason that Anthology prints nothing from the Parliament Rolls, though offering collations of fifteen (“16”, 24) of its edited petitions. The omission is unfortunate, because the texts on the Parliament Rolls are not all of a piece. Petitions of whatever sort loom large among them, and may or may not be of external origin, but texts which must have originated within government are not hard to find. Some originated within Chancery, notably the records of proceedings in Parliament, the king’s responses to

42.PRO C 65/86, m. 13, no. 38, copied from PRO SC 8/135/6716 (cf. Weale 1931: 315, no. 105.1).

27

28

Michael Benskin

petitions, records of examinations in Chancery, and (though less certainly) diverse acts instigated by the Crown. For an account of Chancery’s language these are indispensable, because Chancery otherwise initiated very little writing in English for purposes of state.

7. Early Chancery Proceedings One of the main sources for Samuels’s Type IV London English was “Typical forms in documents written by Chancery-trained scribes” among the Public Record Office’s documentary class called ‘Early Chancery Proceedings’ (PRO C 1). The class consists mainly of private petitions (‘bills’) addressed to the Chancellor, seeking remedy for some wrong that is claimed not to be remediable in the courts of common law. A given bill may or may not be accompanied by the defendant’s answer, subsequent affirmations and denials, and ensuing writs and examinations. By about 1440, such bills are normally in English. From the late 1430s until 1466, well over nine thousand are preserved; from 1467 to 1484, nearly fourteen thousand; and from 1485 to 1500, over twelve thousand.43 Save in rare cases, the names of their writers are unknown. Samuels was far from assuming that “Chancery-trained scribes” accounted for all or even the bulk of these petitions. Some he attributed to provincial writers, as evidence not of the origin, but of the spread of the written standard (1963: 93): These include a number of appeals addressed to the Chancellor that were evidently written in more remote counties, yet, apart from stray dialectal forms, they are written in some approximation to Chancery Standard.

By implication these writers were not Chancery-trained, and though Samuels’s criteria for recognising Chancery training or its lack have proved less secure than he supposed, it is evident that by no means all of the petitions in Early Chancery Proceedings were written by clerks of Chancery (cf. Matheson 1986: 649). Fisher identified those written in what he regarded as non-Chancery

43.These figures, necessarily approximate in that many bills are of uncertain date, are based on the List of Early Chancery Proceedings in vols. XII, XVI and XX of the Public Record Office series Lists and Indexes. Consistent with these counts is Haskett’s (1996: 282) “approximately 61,000 cases” for the period 1417–1532. His “perhaps 90,000 cases” for the period before 1500 (Haskett 1991: 11) seems excessive.

Chancery Standard

hand as non-Chancery documents, apparently “the originals of complaints submitted to Chancery, which had then to be re-written by the clerks [sc. of Chancery] before being officially presented” (1977: 888–889, and so Anthology 21). The supposed evidence for routine rewriting is less the existence of duplicate versions of the same bill, though some are to be found, than the postmediaeval Chancery’s practice (cf. note 41 above), but the defects of the FisherRichardson texts and analyses may be left aside. Rather, some new work will be reported: for the history of standard English, as Samuels wrote forty years ago (1963: 93), it is true indeed that “more could be learnt than hitherto from an intensive study of the Early Chancery Proceedings”. This is not the place for a comprehensive account, and what follows is merely preliminary notice of an examination of over two thousand such petitions, several hundred of them in fine detail. As noted above, it is only in rare cases, if at all, that the name of a petition’s writer is declared. One such may be John Faukeswell, who signed several bills on behalf of diverse petitioners during the late 1430s and 1440s. Those in the same hand and language, with the signature in the same ink, make a presumptive case that Faukeswell had written them; in another, ·per Iohann
em Faukeswell’Ò may be an authorisation.44 Whether he was a Chancery clerk is uncertain. The PRO card index records one or more of that name, but from the 1470s to the late 1480s, too old to be this man though quite possibly his descendant(s); Chancery writs bearing the Faukeswell signature, and of appropriate date, have yet to be found. Personal names are in themselves of little use for establishing whether a petition is the work of a Chancery clerk as opposed to an outsider, but there may be evidence of other sorts. Petitioners in Chancery were required to name two persons who guaranteed payment of damages to a defendant in the event that the charges proved false. When the names are recorded (and on many bills they are not), it is usually as plegii de prosequendo, without elaboration, entered below the text of the bill and to its right, and commonly by a different hand.45

44.In the same hand and spelling are C 1/11/380, 381, 382, 383 and 389 (all signed), and 392 (unsigned). In a different hand and spelling, and with different drafting, is C 1/11/388 (which is also the hand of C 1/11/379). 45.The reports in Anthology are variously garbled: “Plegium de pro
secutio” (no. 128, though two are named), “Pleg’ de per
sequendo” (no’s 172 and 173), cf. no. 171 “pleg’ de prosequendo”. Here as elsewhere, the editors seem unable to recognise the standard abbreviations of mediaeval Latin, and in some cases just ignore them (so in no. 81 “Cancellar Angliae directi” 24, in no. 223 “De per
ficuo Com” and “sub nom
in
e vic” 15–17).

29

30

Michael Benskin

Sometimes, however, the guarantee is recorded in full form, as clauses of mainprise. These are always in Latin, and though the drafting varies, they are in set form. In PRO C 1/18/134, for example, they appear thus: Memorandum quo
d Primo die Aprilis Anno Regni Regis henrici sexti vicesimo nono Iohann
es Avdeley Armiger et Iohann
es Vernoun’ Armiger coram domi
no Rege in Cancellaria sua per
sonaliter constituti manuceperunt videlice
t vter
que eorum pro kater
ina que fuit vxor willielm
i Dorset quo
d in casu quo ipsa mater
iam in hac supplicaci
on
e specificatam veram pro
bare non poter
it . tunc pre
fato Iohann
i lowys omn
ia dampna et expensas que ips
e ea occasi
one sustinebit eidem satisfacient iuxta formam statuti in hac parte editi et pro
uisi. It is to be remembered that on the first day of April in the twenty-ninth year of the reign of King Henry the Sixth [1451], John Audeley, esquire, and John Vernoun’, esq., being personally present in the king’s court in his Chancery, stood surety, namely that each and either of them on behalf of Katherine [the complainant] who was the wife of William Dorset, in the event that she is unable to prove that the matter specified in this bill is true, then to the aforesaid John Lowys [the defendant], for all damages and costs that he himself shall incur in this case, they will make satisfaction to the same according to the terms of the statute declared and provided in this regard.

The mainprise clauses are written always in Chancery hand — not Chancery hand as Fisher and Richardson understand it, but the hand as it appears in the writings of Chancery. The clauses constitute a record of the chancellor’s court, and it is hence unlikely that they were entered by anyone other than a clerk of Chancery. When the hand of the mainprise clauses is the same as that of the bill above them, then likewise the English may be presumed to be of Chancery. Admittedly, a document written by a Chancery clerk need not of itself be evidence for the Chancery writer’s own English, a sample of what he would produce were he composing instead of copying: all it may show is the English of an exemplar, as demonstrated from the Parliament Rolls in Section 6 above. When, however, a Chancery clerk writes diverse bills on behalf of diverse petitioners from diverse parts of the country, and the language is consistent from one bill to the next, that language can hardly be anyone’s but his own. Otherwise we must suppose that the complainants had all used the same agent to draw up their original bills, that the Chancery clerk had copied them more or less exactly, and that so far as the evidence of handwriting goes he never copied bills composed by any other agent acting for complainants in Chancery. Failing which, it follows that the clerk in question either converted the English in the course of copying original versions handed in to Chancery, or that he himself

Chancery Standard

was the author and that the bills extant are the original compositions. Either way, the language is that of a clerk of Chancery. So far, a dozen or so Chancery clerks have been identified as the writers of English bills, among whom is the clerk responsible for the mainprise clauses cited above.46 The same hand, variously mannered or informal, appears to be that of well over forty other bills, to ten or more of which it added the mainprise clauses.47 Its language is plausibly metropolitan for the most part, and need not detain us; of more interest for present purposes are the hands who write English of another sort. One such is the hand of C 1/9/121, a petition in the name of William Caser of Wiltshire; it appears in eight other bills, from the years 1437 to 1439.48 Its language is self-consistent, but it hardly reflects the Home Counties: they them their such many pres. indic. pl

þai hom hor soche mony -on, -en

friend give have own adj. soul two

fryndgyf, gyve, yif hafe avne, ovne saule twoo

The evidence of these and other forms establishes dialectal origins in the northwest Midlands: in terms of the Atlas configuration, the focal area extends from the west of mid-Derbyshire into north-east Staffordshire.49

46.In this case the identity is blatant, but it is commonplace that a single hand may change aspect as it shifts between English and Latin, and some do so almost beyond recognition. Chancery hands that admit secretary forms in their English, sparingly or otherwise, may exclude them altogether from their Latin. Formality is also at issue: the English bill, though formulaic, was in essence a private letter of complaint addressed to the chancellor, but its Latin mainprise clauses were an official memorandum of Chancery. 47.PRO C 1/9/44, 47, 60, 104, 115, 117, 125, 142, 181, 190, 252, 254, 260, 262a, 267, 271, 273, 274, ?342, ?340b, 404, 406, 408, 409, 413, 419, 422, 434, 457, 459, 460 & 465; C 1/17/54, 55a, 57, 65, 79, 87 & 91; C 1/18/133; C 1/19/4, 155, 166, 173, 181, 458 & 468. The hand of the mainprise clauses is that of the petition in C 1/9/115, 125, ?142, 190, 252, 254, 260, 262a, 267, 271, 274 & ?457. The dates range from 1437 or before to 1451 or later. The case for identity is reinforced by the spelling; note especially ·gra
ceousÒ gracious, ·reuerentÒ reverend, and ·cheriteeÒ charity, combined with the distinctive form of the nota for and. 48.PRO C 1/9/105, 131, 134, 140, 259, 263, 272 & 335; mainprise clauses in all save 335, but those in 259 and 272 appear to be by different hands. In the same hand is a petition addressed to Parliament, PRO SC 8/25/1246 (cf. p. 26 above). 49.Between LPs 227 and 718 (Staffs.), 498 (Derbys.), 54, and 61 (see key Map 2, Atlas II: 384, and IV: 335). Note also it
; which, whech; ony
; moche
; 3sg. will, woll; froo; ayenst
; nat

31

32

Michael Benskin

A more diluted north midland language, this time easterly, appears in a group of ten petitions by the one hand, which also added their clauses of mainprise. The dates are from 1438 to 1443. The same is responsible for a memorandum of a viva voce witness examination in Chancery, dated 19 January 17 Henry VI (i.e., 1439), a type of record unnoticed in previous studies of Chancery English.50 Among this clerk’s forms are: such each much any

sich, syche ich mych ony

beseech goods 3sg. pres. indic.

besek-, besechgudes, godes -s, -th

There are other northerly indications. The letters y and þ are written ·yÒ without distinction. The suffix of the present indicative plural is usually -th, sometimes -n, but also, as in the witness examination, sometimes -s or zero; and throughout, the selection of a consonantal as opposed to zero suffix conforms with the northern personal pronoun rule.51 Such cases do little to support the belief that Chancery, as a matter of policy, trained its clerks to write English of a certain type, and more is at issue here than concerns historians of the language. Much of the history of the mediaeval court of Chancery is unwritten, but investigation on a very large scale is now in progress, namely The Early Court of Chancery in England Project, 1417–1532, directed by Timothy S. Haskett (see Haskett 1996). Such a history can be recovered only from the bills of complaint, for “Most of the primary documentary survivals of the medieval court are not the court’s own records

(not-); nor nor; throwe, -thorwe through; pres. indic. 3sg. -eth, -ith; gode (good); hie, hegh high; luf love sb.; adv. -le, -ly. The petitions concern persons or property in Kent, Wiltshire, Warwickshire, Devon, Lincolnshire, Sussex, Norfolk, and Lancashire; PRO SC 8/25/1246 concerns property in Peterborough. Atlas’s assessments of their language are cursory and mostly wrong, though plausible when each document is considered apart: only when the texts are conflated can their language be properly evaluated, and that depends on recognising that they are all in one hand. To see that they are so is natural enough for a palaeographer or student of drafting formulae who works intensively on the Early Chancery Proceedings; but when no one text of itself provides sufficient material for localisation, a collaborative dialect survey could hardly recognise their common origins save by chance. 50.The petitions are PRO C 1/9/11, 16, 23, 76, 79, 94, 107, 111, 456 & 474; C 1/11/362 and C 1/12/199 are in the same hand, but lack mainprise. In C 1/9/485, the mainprise but not the petition is in this hand. The examination in Chancery is C 1/9/486. 51.Cf. p. 8 above, 3rd para.; Benskin 1991: 85–88.

Chancery Standard

but the petitions brought to it”; and they compel or ought to compel the interest of mediaeval historians at large, for they touch on very many areas of mediaeval life (Haskett 1996: 280–281, 284). Understanding of these documents will be advanced by analysis not only of their contents but of their forms, and in this, philology must be brought to bear as well as palaeography and diplomatic. Most historians have to take the philology on trust, and Fisher’s seemingly authoritative claims as to what was and was not Chancery English are no help to them at all.52

8. The English of government and the spread of standard English Prominent in Fisher’s Speculum article of 1977 is the idea that the forms of Chancery English were subject to deliberate control, a view re-stated uncompromisingly in “Chancery Standard and modern written English” (Fisher 1979, esp. 139–140) and promoted more recently in The Emergence of Standard English (Fisher 1996). As N. F. Blake wrote (1997: 512) in a review: Fisher’s position on the rise of standard English is that it was the result of deliberate planning and management; it was not something which just happened.

Fisher’s belief in such control, however, rests mainly on misunderstandings of what he has read. He claims (1977: 891) that Because of the tight hierarchy in Chancery, the idiom and form of the official documents was controlled by a very few people. In theory only the twelve masters in Chancery were empowered to originate new language or to sign important letters.

The main authority here is “Statutes of the Realm, 2 Inst. 405”, cited from an article of 1885. Fisher does not give the statute’s date, but perhaps considered it of no account, assuming as in other matters that Chancery’s practice was largely

52.It is untrue, for example, that seid is a non-Chancery form (cf. Haskett 1991: 16): it appears repeatedly in enrolments of royal proclamations (for which origins outside government are not in question) on the Close Rolls of the 1460s and 1470s. It must be set on record that the philology represented in the writings of Fisher and Richardson is most emphatically not to be trusted (cf. pp. 7–10 above); that “sho [for she] in the derivative versions is more archaic than sheo in the original” (Anthology, 20) could be written only in ignorance of one of the most celebrated etymologies in the history of English (cf. OED s.v. she; Samuels 1972: 114–116; Britton 1991).

33

34

Michael Benskin

unchanging (see above, pp. 23, 25 and note 41); and his excerpt refers not to masters in Chancery, twelve or otherwise, but to Clerks (“let the Clerks of Chancery agree in making a writ … And …where they cannot agree … let the Clerks refer the cases to the next Parliament”). These, however, are minor infelicities: the whole citation is irrelevant to Fisher’s case, because it is about legal remedy and legal form, about the making of Chancery writs, instruments that during the fifteenth century were composed in Latin, and for which Latin remained the accustomed language until the Commonwealth. New writs could have unforeseen consequences, and so could customary writs issued in novel wording; that is why their forms had to be so carefully controlled. Certainly “There is ample evidence of the supervision exercised over the writing of the under clerks”, but it has yet to be shown that any of it has the slightest bearing on Chancery English. Fisher’s other primary source is from Sanders’s Orders of the High Court of Chancery, namely the Ordinaciones of 1388–1389, though he is silent as to their date. Sanders (1845, I: 1–2, note (b)) recorded the claim that in the time of Henry V (1413–1422) they were renewed with additions and reformations, and thought it possible that such emendations are incorporated in the surviving text of the 1388–1389 ordinances, but there is no certain evidence as to how long the relevant provision remained in force. It is at all events lacking in the ordinances that can be assigned to Henry V’s reign, namely the Renovacio Ordinum Cancellarie cum novis addicionibus et reformacionibus eorundem (Sanders 1845, I: 7b–d), and until 1545 there are no more. Fisher reports (1977: 892) that the Ordinaciones (in which read allii not allei) forbad Chancery clerks to mingle with “omni allei clerici qui dicte cancellarie propter doctrinam et scripturam adherere voluerint …” “Scripturam” no doubt here means the technical forms of writs, and perhaps the formal Chancery hand, but it must inevitably have meant, also, the correct use of the official language, be it Latin, French, or — after 1420 — English.

It is a question whether the maker(s) of the Ordinaciones even considered the possibility of English as an official language, but that scripturam here embraces English morphology and spelling — the defining characteristics of Chancery Standard — depends less on accurate reading than imagination. In contrast to Fisher’s other writings, Anthology is curiously reticent about the putative Chancery policy, referring neither to the statute nor the Ordinaciones. The foreground is dominated by Henry V, whose importance in establishing English as an official written language cannot be gainsaid. Even so,

Chancery Standard

Anthology’s lengthy analysis of the dictaminal form and personal style of his letters is hardly relevant to the development of a written standard, for in principle they could be replicated in any dialect of English, or, as indeed they had been, in French. In Anthology’s view (xv), “It would appear that the Signet office of Henry V was most influential in establishing the conventions of Chancery Standard”. Their consolidation and spread, however, are still presented as the work of Chancery clerks, who “by the 1440’s and 1450’s … had achieved a comparative regularization” (27). This achievement is supposed to be borne out by a review (27–51) of the forms in Anthology’s edited documents, “which show the [Chancery] clerks trying to eliminate the kind of orthographic eccentricity found in the Privy Seal minutes”, etc. (27), but because Fisher and Richardson have failed to understand the sources of their documents, their designations of forms as ‘Chancery’ or ‘non-Chancery’ are worthless. Less than a quarter of the documents in An Anthology of Chancery English were written by clerks of Chancery, but even at that Chancery may be overrepresented in so far as the collection is designed to illustrate the forms of English propagated by government. The editors are right to say (17) that “It was documents by the professional clerks — what we today [1984] might call the typing pool — that circulated throughout the country and helped to establish the model for official English”, but those clerks were predominantly the clerks of Signet and Privy Seal, and their training was not, in spite of Fisher’s repeated assertions, in the households of Chancery (Otway-Ruthven 1935: 133–134; Brown 1969, 1971: 265–266; cf. Anthology, 17). Moreover, the clerical muster of these offices, the ‘typing pool’, was small: in the Signet Office four or five, rising to eight under Edward IV, and in the Privy Seal Office perhaps twelve (OtwayRuthven 1935: 110–113; Brown 1971: 261–262). At any one time, therefore, the king’s English was represented to the country at large by relatively few of his clerks, whose language, like John Depeden’s, was not always mainstream in all respects (see pp. 20–21 above). Anthology’s collection of state documents includes by intention nothing later than 1455. It would be a mistake, however, to imagine that the forms of English prevailing in government writings of that date continued to dominate it for the rest of the century, save in so far as they were replaced by more modern forms. So much appears, for example, from letters under the signets of Edward IV (1461–1483) and Richard III (1483–1485). About nine hundred of these, at a rough estimate, with as many more in Latin, are to be found in the PRO class PSO 1, conveniently filed in order of date. Each bears a clerk’s signature, and that a given name goes usually with the same handwriting

35

36

Michael Benskin

indicates that in general the signatures were those of the clerks themselves, rather than a superior’s authorisations (cf. Otway-Ruthven 1935: 26–27). Among forms that anglicists would perhaps not expect for this period, with the names and dates of their perpetrators, are nat not and thise these: nat 1463– Bele, Harpisfeld, Craft; 1474– Slefeld; 1478– Freman. thise 1477– Herbert, King; 1478– Freman; 1481– Bedenne, Gregory, Pery, Robyns, Wylde; 1484/5 Kendale.

Both forms look back to the Ellesmere Chaucer, and separate Samuels’s Type III from Type IV, Chancery Standard. Belonging to neither Type is thies or þies, a form of these not uncommon in provincial legal documents of the later fifteenth century, and from parts of the country in which it was not the earlier local usage (cf. Atlas vol. I: 306, Map 5). It enters in company with various other non-local forms, and the language of the relevant documents is typically semi-standard,53 but thies is neither Chancery Standard as hitherto described, nor even prominent in London English, though its spelling is generally assumed to imply the same pronunciation as Type IV thes(e). A source is not far to seek: thies (sometimes þies) 1463– Bele, Craft, Harpisfeld, Herbert; 1469– Profoot; 1474– Robyns, Slefeld; 1477– King; 1478 Freman, Wylde.

For a time, these clerks were among the foremost disseminators of the king’s English, and it is little wonder if their forms were adopted by lawyers and administrators in the shires: no-one then could know that nat and thise and thies would not survive into the later standard, and some of the clerks to Henry VIII did not know it either. Only in the long view can such forms be recognised as deviations from some true path, and it is a question how far any writers during the fifteenth century thought in terms of a written standard as fixed for even the foreseeable future. A standard need be no more than a fashion widely accepted, and changes in fashion proceeded often enough from the royal court. The development of a written standard, even in the offices of government, was more complex and less determined than it has sometimes been made to appear, and government English is not the whole story. Somehow, the London

53.As, e.g., in Cumbria Record Office (Carlisle) Ca. 5/1/28, dated 1474 (Atlas LP 1170), beside fully northern ther
; cf. been are beside er, and pres. part. -yng not -and. The language is substantially local, but had better sources been available for Carlisle, it would not have been utilised for the maps.

Chancery Standard

origins of Chancery Standard have been left aside. London is a peculiar quantity in Fisher’s account, its language initially heterogeneous compared with his Chancery Standard; yet Chancery Standard, first attested in the Signet letters of Henry V, is then recognised in Guildhall documents, and at an earlier date than it appears in those of Chancery. Critical here, but yet to be given proper attention, are (1) the chronology, (2) how far English writings of given type and from a given source are available at a given date, and (3) the number of scribes whose language they more or less inadequately represent. It remains of course a question whether London English in some form would have become a national standard even had government kept to Latin and French. As Morsbach recognised (1888: 167–168), London’s commercial pre-eminence counts for much in this history, in connexion with both origins and spread; so too does its sheer size, unrivalled by any city in England. From beyond London and the Home Counties, the perspectives are more complicated, in so far as ‘colourless regional standards’ arise independently of London or government English, but interact with them as well (Samuels 1981; Benskin 1991: 82–85). It must be remembered, too, that people as well as writings went out from London into the provinces, government officials among them. Through the Inns of Chancery and the Inns of Court passed many who returned to provincial careers, whether in law or administration; and after qualifying, some of the most successful lawyers practising in the Westminster courts were only term-time residents in the capital, looking to a clientele drawn largely from their counties of origin (Ives 1968, 1983). Historians of standard English have hardly touched on legal education so far, but the Inns of Court look increasingly to have provided a main conduit from the capital (Benskin 1989: 20–23, 1991: 88–90, 2001: 740–746). All this, however, is to anticipate a larger study of the beginnings of standard English, in which, as it will now appear, the Chancery cannot be assigned the leading part. Indeed, outside its equity jurisdiction, Chancery’s ordinary administrative practice did nothing to promote English of any sort, but rather, for the purposes of government, retarded it. The great bulk of Chancery’s routine administration was effected by writs, and Chancery writs were composed almost exclusively in Latin. Much has been made of Chancery’s hundred-and-twenty disciplined scribes, and their supposed conversion of the English in documents initiated outside Chancery into the official Chancery forms: English letters under the Signet were recast as letters under the Privy Seal, these in turn were recast as letters from the Chancery, and in the Chancery’s issue (it is said) the irregular usages of the other offices gave way to Chancery standard. They did, but that

37

38

Michael Benskin

standard was Latin. Drafts that the editors of Anthology imagine to have culminated in Chancery English culminated either in Chancery Latin, mostly as writs or letters patent, or in English under Privy Seal. Only exceptionally did drafts under Privy Seal result in English missives from the mediaeval Chancery. Chancery Standard was Latin, and save for nine years during the Commonwealth, it remained so until 1731, when for official purposes it was abolished altogether by Act of Parliament.

References Atlas: Angus McIntosh, M. L. Samuels & Michael Benskin, with the assistance of M. Laing & K. Williamson. 1986. A Linguistic Atlas of Late Mediaeval English. Aberdeen: Aberdeen University Press. Bennett, J. A. W. & G. V. Smithers, eds. 1968. 2nd edn; repr. 1974. Early Middle English Verse and Prose. Oxford: Clarendon Press. Benskin, Michael. 1989. “Some aspects of Cumbrian English, mainly mediaeval”. Essays on English Language in Honour of Bertil Sundby ed. by Leiv Egil Breivik, Arnoldus Hille & Stig Johansson, 13–46. Oslo: Novus Forlag. ———. 1991. “Some new perspectives on the origins of standard written English”. Dialect and Standard Language/Dialekt und Standardsprache, in the English, Dutch, German and Norwegian Language Areas ed. by J. A. van Leuvensteijn & J. B. Berns, 71–105. Amsterdam: Koninklijke Nederlandse Akademie van Wetenschappen Verhandelingen, Afd. Letterkunde, Nieuwe Reeks, deel 150. ———. 2001. “Lincoln’s Inn, Black Book vol. 1 (A1a1)”. Appendix 9(a) of The Records of the Honorable Society of Lincoln’s Inn. The Black Books Vol. VI A. D. 1914–A. D. 1965, ed. by P. V. Baker. London: Lincoln’s Inn. 740–746. ——— & Margaret Laing. 1981. “Translations and Mischsprachen in Middle English manuscripts”. Benskin & Samuels 1981, 55–106. ——— & M. L. Samuels, eds, 1981. So meny people longages and tonges. Philological essays in Scots and mediaeval English presented to Angus McIntosh Edinburgh: Benskin & Samuels. Blake, N. F. 1997. Review of Fisher 1996. Review of English Studies N. S. 48. 512–13. Brown, A. L. 1969. The Early History of the Clerkship of the Council. Glasgow: Glasgow University Publication NS 131. ———. 1971. “The Privy Seal clerks in the early fifteenth century”. The Study of Medieval Records: Essays in Honour of Kathleen Major ed. by D. A. Bullough & R. L. Storey, 260–281. Oxford: Clarendon Press. Calendar of the Plea & Memoranda Rolls preserved among the Archives of the Corporation of the City of London at the Guildhall a. d. 1458–1482. Philip E. Jones, ed. 1959. Cambridge: Cambridge University Press. Chambers, R. W. & Marjorie Daunt, eds. 1931. (repr. 1967). A Book of London English 1384–1425. Oxford: Clarendon Press.

Chancery Standard

Cheney, C. R., ed. 2000. Handbook of Dates for Students of English History rev. by Michael Jones. Cambridge : Cambridge University Press. Collingwood, R. G. 1939 (repr. 1970). An Autobiography. Oxford: Clarendon Press. Ekwall, Eilert. 1956. Two Early London Subsidy Rolls. Lund: Gleerup. ———. 1956. Studies on the Population of Medieval London. Stockholm Fisher, John H. 1977. “Chancery and the emergence of standard written English in the fifteenth century”. Speculum 52. 870–899. ———. 1979. “Chancery Standard and modern written English”. Journal of the Society of Archivists 6. 136–144. ———. 1988. “Piers Plowman and the Chancery tradition”. Medieval Studies presented to George Kane ed. by Edward Donald Kennedy, Ronald Waldron & Joseph S. Wittig, 267–278. Wolfeboro, NH: D. S. Brewer. ———. 1992. “A language policy for Lancastrian England”, Publications of the Modern Language Society of America 107. 1168–1180. ———. 1996. The Emergence of Standard English. Lexington: University Press of Kentucky. ———, Malcolm Richardson & Jane L. Fisher, eds.1984. An Anthology of Chancery English. Knoxville: University of Tennessee Press. Guide to the Contents of the Public Record Office, vols. I–II 1963, vol. III 1968. London: H. M. S. O. Hall, Hubert. 1908. A Formula Book of English Official Historical Documents. Part I, Diplomatic Documents. Cambridge: Cambridge University Press. Handbook of British Chronology, 3rd edn ed. by E. B. Fryde, D. E. Greenway, S. Porter & I. Roy. Cambrdige: Cambridge University Press. Haskett, Timothy S. 1991 “The presentation of cases in medieval Chancery bills”. Legal History in the Making ed. by W. M. Gordon & T. D. Fergus, 11–28. London: Hambledon Press. ———. 1996. ‘The medieval English court of Chancery”. Law and History Review 14. 245–313. Ives, E. W. 1968. “The common lawyers in pre-Reformation England”. Transactions of the Royal Historical Society. 5th series, 18. 145–173. ———. 1983. The Common Lawyers of Pre-Reformation England. Cambridge: Cambridge University Press. Jones, Thomas. ed. 1955. Brut y Tywysogyon, or The Chronicle of the Princes. Red Book of Hergest version. Cardiff: University of Wales Press. Kirby, J. L. 1978. Calendar of the Signet Letters of Henry IV and Henry V (1399–1422). London: H. M. S. O. McIntosh, Angus. 1983. “Present indicative plural forms in the later Middle English of the North Midlands”. Middle English Studies Presented to Norman Davis in Honour of his Seventieth Birthday ed. by D. Gray & E. G. Stanley, 235–244. Oxford: Clarendon Press. MacIntyre, Alasdair. 1985. After Virtue. A Study in Moral Theory. 2nd edn, 1st edn 1981. London: Duckworth. Matheson, Lister M. 1986. Review of Fisher, Richardson & Fisher 1984. Speculum 61. 646–650. Maxwell-Lyte, Sir H[enry] C. 1926. Historical Notes on the Use of the Great Seal of England. London: H. M. S. O.

39



40

Michael Benskin

Morsbach, Lorenz. 1888. Ueber den Ursprung der neuenglischen Schriftsprache. Heilbronn: Henninger. Mustanoja, T. 1960. A Middle English Syntax, vol. I, Parts of Speech. (= Mémoires de la Société Néophilologique de Helsinki vol. XXIII.) Helsinki: Société Néophilologique. Myers, A. R. 1937. “Parliamentary petitions in the fifteenth century”. English Historical Review 52.385–404, 590–613. Otway-Ruthven, [A.] J. 1939. The King’s Secretary and the Signet Office in the XV Century. Cambridge: Cambridge University Press. Parkes, M. B. 1969. English Cursive Bookhands 1250–1500. Oxford: Clarendon Press. Reaney, P. H. 1935. The Place-Names of Essex. (= English Place-Name Society vol. XII.) Cambridge: Cambridge University Press. Rees, William. 1967. An Historical Atlas of Wales from Early to Modern Times, 3rd edn. London: Faber & Faber. Richardson, Malcolm. 1980. “Henry V, the English Chancery, and Chancery English”. Speculum 55.726–50. Samuels, M. L. 1963. “Some applications of Middle English dialectology”. English Studies 44.81–94. ———. 1972. Linguistic Evolution, with special reference to English. Cambridge: Cambridge University Press. ———. 1981. “Spelling and dialect in the late and post-Middle English periods”. Benskin & Samuels 1981.43–54. ———. 1983. “Chaucer’s spelling”. Middle English Studies presented to Norman Davis in Honour of his Seventieth Birthday ed. by Douglas Gray & E. G. Stanley, 17–37. Oxford: Clarendon Press. Sanders, George Williams, ed. 1845. Orders of the High Court of Chancery, and Statutes of the Realm relating to Chancery, from the Earliest period to the Present Time. 2 vols. London: Maxwell. Somerville, R. 1953. History of the Duchy of Lancaster, vol. I [1265–1603]. London: Chancellor and Council of the Duchy of Lancaster. Steer, Francis W., ed. 1968. Scriveners’ Company Common Paper 1357–1628. With a continuation to 1678. London: London Record Society. Tout, T. F. 1927. Review of Maxwell-Lyte 1926. English Historical Review 42.630–635. Weale, Magdalene M. 1931. “A descriptive list of English manuscripts in the Public Record Office, London”. Chambers & Daunt 1931.269–317. Wright, Joseph. 1905. The English Dialect Grammar. Oxford: Clarendon Press.



Cant and slang dictionaries A statistical approach Julie Coleman University of Leicester

1.

Introduction

Like their mainstream counterparts, cant and slang dictionaries often owe much, if not everything, to earlier works. Relationships between the dictionaries discussed in this paper have long been known (Hotten 1859: 152; Burke 1939: 65, 72–4; Starnes & Noyes 1946: 212–227; Partridge 1950: 62, 70). What has been harder to describe is how each lexicographer used his source, and how compilers’ claims reflect their actual practice. To explore one way in which selection methods can be identified, I will concentrate on dictionaries derived from B. E.’s New Dictionary of the Terms Ancient and Modern of the Canting Crew, published in London by W. Hawes in c1698.

2. B. E.’s dictionary It is not surprising that B. E.’s dictionary was used as a source by later lexicographers, because it is far more extensive than any that preceded it, and also innovative in a number of respects. For example, B. E. is the first lexicographer of English slang: unlike earlier lists, his includes general slang, jargon, and colloquialisms, as well as cant terms. This contributes to a marked increase in size of the word-list: Richard Head’s Canting Academy listed a few hundred terms; B. E. has over 4000 entries. His is the first dictionary of non-standard English to be published as a volume in its own right, rather than as a mere glossary appended to another work. None of the cant lexicographers had

42

Julie Coleman

included etymologies in their word-lists; B. E. provides a small number.1 None had offered usage labels; B. E. labels almost a quarter of his entries. Head had included two illustrative citations; B. E. provides over three hundred and fifty. He also incorporates almost three hundred proverbs and proverbial expressions into his dictionary: as headwords, citations, and even definitions. Crossreferences indicate useful further information listed under related terms. In short, B. E.’s dictionary is so broad in scope and in approach, that it offers something to everyone. For almost a century, until it was superseded by Grose’s Classical Dictionary of the Vulgar Tongue, this was the fullest record of nonstandard English available. B. E. did not produce his dictionary in isolation from all that had gone before (Coleman 2001: 400–1). He used the cant-English and English-cant lists in Head’s Canting Academy (1673) and the list of debtors’ slang in Thomas Shadwell’s Squire of Alsatia (1688). He also used the glossaries of specialized jargon in John Guillim’s Display of Heraldrie (1632, or later editions) and W. Hope’s Compleat Fencing Master (1691). For definitions of standard English senses, he sometimes turned to Elisha Coles’s English Dictionary (1676). In analysing B. E.’s dictionary, I used Microsoft Access databases. Each entry in the database has fields for separate features of the dictionary entries. For example, B. E.’s entry for backt
: Backt, dead, as he wishes the old Man backt, he longs to have his Father upon six Mens shoulders; or as his Back’s up he is in a fume, or angry.

incorporates two separate definitions: one for the headword, which is illustrated by a citation, and one for his back’s up. These each constitute a row in the database. The first belongs to the semantic field of death, the second to the field of emotion. Usage labels, where they occur, are normalized, so the entries: Adrift, loose. I’ll turn ye adrift, a Tar-phrase; I’ll prevent ye doing me any harm. Antient, at Sea, for Ensign, or Flag.

are both entered into the database as ‘naut’ for nautical or naval. Categories in the database for which a yes/no response was required included etymology, authority, citation, cross-reference, and proverb. Dictionaries derived in whole or in part from B. E.’s include Alexander Smith’s Thieves New Canting Dictionary (1719), the anonymous New Canting

1.Thomas Dekker discusses etymology in his Belman of London, but he does not incorporate that information into his word-list.

Cant and slang dictionaries

Dictionary (1725), Francis Grose’s Classical Dictionary of the Vulgar Tongue (1785), and an American edition of B. E.’s dictionary published in 1813 under his name, and with the same title. Many later lists were to use these works, so B. E.’s indirect influence continued well into the nineteenth century. 2.1 Alexander Smith’s Thieves New Canting Dictionary Smith’s Thieves New Canting Dictionary contains 286 entries from B. E.’s 4052 (7.1%). His motivation for choosing most of them is clear, despite Gotti’s (1999: 81) assertion to the contrary: 262 (91.6%) are labelled as ‘cant’ in B. E. (p = 0.01).2 This is not all of B. E.’s cant terms, however. Smith also selected in favour of non-cant terms illustrated by citation: 90 (32.5%) of his entries include citations, which is a marked increase on B. E.’s 8.8% (p = 0.01). Smith also selected against terms belonging to the semantic field of pleasure and pastimes (p = 0.05). In short, Smith produced a specialized dictionary of cant from B.E’s general slang dictionary. His is livelier, in that a larger proportion of entries include illustrative citations. It is also more focused semantically. In excluding pleasure and pastimes terms, Smith separated the thieves from normal citizens, and made them seem more menacing: even their leisure activities are unfamiliar. 2.2 The New Canting Dictionary The New Canting Dictionary was published in 1725. Its preface is suitably promotional: With Regard to the Performance in general, we shall only say, There is nothing so complete of its Kind; and we have taken no small Pains to collect all the New Words made use of by Villains of all Denominations …

and promises great things to the student of its word-list: …By perusing and retaining many of which, an Honest Man, who is obliged to travel the Road, and to frequent Inns and Places of Publick Resort, (whereby he is often forced to mix with different Companies) may easily discover, by the

2.Where p = probability. If p = 0.01, the chance of the result occurring through random variation is less than 1%. Where p = 0.05, as below, the chance of the result occurring through random variation is less than 5%. If p is greater than 0.05, the result is not statistically significant, and is probably attributable to chance.

43

44

Julie Coleman

Cant Terms and Dialects of the Persons, their Profession and Intentions, and know how to secure himself from Danger; Which is the principal Design of compiling this Vocabulary …

In fact, this dictionary is about two thirds the size of B. E.’s, containing 2606 entries in comparison with B. E.’s 4052. It adds a few entries, but not enough to justify the claims of its preface. Like Smith’s dictionary, this claims to focus on the language of thieves, and it fulfils its promise: 94.1% of B. E.’s entries labelled ‘cant’ are in the New Canting Dictionary (p = 0.01). In total 924 entries (35.5%) are derived from entries labelled ‘cant’ in B. E. (p = 0.01). There is no significant difference in the semantic coverage of cant terms in the two dictionaries. However, most of the terms in the New Canting Dictionary are not cant. 51.9% of B. E.’s entries that are not labelled ‘cant’ are in the New Canting Dictionary. Here there is a significant difference in semantic coverage between the two dictionaries. The compiler of the New Canting Dictionary selected in favour of non-cant terms belonging to the semantic fields of crime & dishonesty, law & order, sex (all p = 0.01), and poverty (p = 0.05). There are, in addition, fewer miscellaneous entries, and fewer for animals & nature (both p = 0.01). In other words, the editor of the New Canting Dictionary confounded meaning and register in selecting from his source. Terms belonging to semantic fields related to the activities of criminals and beggars are implicitly labelled as cant by their inclusion in his word-list.3 2.3 Francis Grose’s Classical Dictionary of the Vulgar Tongue The first edition of Francis Grose’s Classical Dictionary of the Vulgar Tongue was published in 1785. Grose is the first cant and slang lexicographer to present himself as an antiquarian. Unlike his predecessors, Grose is happy to list his sources, although he is more honest about ancient sources than modern ones. He also claims to have collected terms amongst the low populace of London. In his preface, he writes: The many vulgar allusions and cant expressions that so frequently occur in our common conversation and periodical publications, make a work of this kind

3.Gotti (1999:86) discusses entries beginning with ‘A’ and ‘B’ that are in B.E.’s dictionary but not the New Canting Dictionary. He asserts that the compiler chose to omit terms that, “were considered obsolete … words belonging to the general language … specialized terms … literary or figurative expressions … colloquial expressions … [and] words of foreign origin”.

Cant and slang dictionaries

extremely useful, if not absolutely necessary, not only to foreigners, but even to natives resident at a distance from the metropolis, or who do not mix in the busy world …

The first edition of Grose’s dictionary contains 3893 entries in all. Grose used several word-lists, including some of B. E.’s own sources as well as dictionaries derived from B. E. This makes it rather difficult to determine exactly what use he made of B. E.’s dictionary, but 254 (6.5%) of Grose’s entries can only be from B. E., and 1904 (48.9%) could be from B. E. The following analysis assumes that all terms that could be from B. E.’s dictionary are, because that is a large enough sample for meaningful comparisons to be made. B. E. labelled as ‘cant’ 722 (37.9%) of the entries that Grose included in his dictionary (p = 0.01). There is no significant difference in the semantic coverage of these terms. However, 1182 of Grose’s entries from B. E. (62.1%) were not labelled as ‘cant’ by B. E. Among these terms there is a significantly different semantic coverage. Non-cant terms from B. E.’s dictionary were more likely to be included in Grose’s dictionary if they belonged to the fields of crime & dishonesty, body & health, sex (all p = 0.01), or poverty (p = 0.05). They were less likely to be included if they belonged to the fields of pleasure and pastimes, animals & nature, and geography & travel (all p=0.01). Like the compiler of the New Canting Dictionary, Grose was influenced by meaning as well as register. His preference for terms belonging to the semantic field of sex disproves his claim in the preface that: To prevent any charge of immorality being brought against this work, the Editor begs leave to observe, that when an indelicate or immodest word has obtruded itself for explanation, he has endeavoured to get rid of it in the most decent manner possible; and none have been admitted but such, as either could not be left out, without rendering the work incomplete, or, in some measure, compensate by their wit, for the trespass committed on decorum.

Grose also selected in favour of entries illustrated by citations. Two hundred and twenty-seven of Grose’s entries from B. E. (14.6%) include illustrative citations (p = 0.01). This preference for entries containing citations applies only to entries not labelled as ‘cant’ (p = 0.01). 2.4 An American edition of B. E.’s dictionary The last dictionary derived from B. E.’s that I am going to consider here is the American edition published in 1813. This curious dictionary contains 1658

45

46

Julie Coleman

entries, of which 1569 (94.6%) are from the earlier edition. This editor pared down and focused the material derived from his source. He selected against entries labelled as ‘cant’, illustrated by citation, including etymologies, and containing cross-references (all p = 0.01). The title of the dictionary is still A New Dictionary of the Terms Ancient and Modern of the Canting Crew, but only 221 of its entries (13.3%) had been labelled as ‘cant’ in the earlier edition. Among these ‘cant’ terms there are now more miscellaneous entries and more entries for emotion & temperament (both p = 0.01). 1348 of this dictionary’s entries (81.3%) were not labelled as ‘cant’ in the earlier edition. Among these entries too, there are marked differences in semantic coverage. The editor selected in favour of crime & dishonesty, emotion & temperament, geography & travel, poverty, sex, and violence (all p = 0.01). He included fewer miscellaneous entries, and fewer belonging to the fields of animals & nature, clothes & looks, fools & victims, pleasure & pastimes (all p = 0.01), money, and speech (both p = 0.05). His motives are explained by entries like: Fulsome-flattery; when even vanity itself is nauseated with its excess and absurdity. He-is-out; the polite refusal of admittance, at the doors of fashionable life. Quotation; the practice of quotation is either an ostentation of knowledge, or an honest effort to persuade. The omission of it is either an affectation of originality, or the prudential dictate of ignorance. The motive in either case may be judged of by the congruity or incongruity of the occasion.

This editor’s aim was to produce not a dictionary of the language of thieves and beggars, but a dictionary of the language of hypocrites and poseurs: it is still a dictionary of cant, but of cant understood in a different sense.

3. Conclusions In summary, the scope and innovations of B. E.’s dictionary recommended it as a starting-point for later cant and slang lexicographers. Each selected entries from the dictionary according to different criteria, depending on their individual aims. Two main selection methods emerge: by usage label and by meaning. Smith, Grose, and the New Canting Dictionary all show a preference for terms labelled as ‘cant’ in B. E.’s dictionary. Conversely, the American edition selects



Cant and slang dictionaries

against ‘cant’ terms. All of the dictionaries derived from B. E.’s are influenced by the meaning of the terms he includes: terms belonging to specific semantic fields, especially crime & dishonesty, poverty, and sex, have a better chance of being included in some dictionaries, and a worse chance in others. Both Smith and Grose show a preference for entries illustrated by citation in B. E.’s dictionary, while the American edition selects against entries containing citations and other extraneous material. These methods are applicable to studies of the relationships between any dictionaries. They make it possible to pin down differences between dictionaries that are largely derived from the same source. They produce verifiable generalizations about relationships between word-lists, and allow for comparison between lexicographers’ claims and their actual practice.

References Anon. 1725. New Canting Dictionary. London: The Booksellers of London and Westminster. Burke, W. J. 1939. The Literature of Slang. New York: New York Public Library. Coleman, Julie. 2001. “Some of the sources of B. E.’s New Dictionary of the Terms Ancient and Modern of the Canting Crew”. Notes and Queries 246.400–1. Coles, Elisha. 1676. An English Dictionary. London: Samuel Crouch. Dekker, Thomas. 1608. The Belman of London. London: Nathaniel Butter. E., B. c1698. A New Dictionary of the Terms Ancient and Modern of the Canting Crew. London: W. Hawes. ———. 1813. A New Dictionary of the Terms Ancient and Modern of the Canting Crew. Philadelphia: printed for the publisher. Gotti, Maurizio. 1999. The Language of Thieves and Vagabonds. 17th and 18th Century Canting Lexicography in England. Tübingen: Niemeyer. Grose, Francis. 1785. A Classical Dictionary of the Vulgar Tongue. London: S. Hooper. Guillim, John. 1632. A Display of Heraldrie. London: Ralph Mab. Head, Richard. 1673. The Canting Academy. London: F. Leach for Mat. Drew. Hope, W. 1691. The Compleat Fencing-Master. London: Dorman Newman. Hotten, John Camden. 1859. A Dictionary of Modern Slang, Cant, and Vulgar Words. London: John Camden Hotten. Partridge, Eric. 1950. Slang Today and Yesterday. 3rd edn. London: Routledge & Kegan Paul. Shadwell, Thomas. 1688. The Squire of Alsatia. London: James Knapton. Smith, Alexander. 1719. The Thieves New Canting Dictionary. London: Sam Briscoe. Starnes, De Witt T. & Gertrude E. Noyes. 1946. The English Dictionary from Cawdrey to Johnson 1604–1755. Chapel Hill: University of North Carolina Press.

47



DOST A significant instance of historical lexicography Marace Dareau Scottish Language Dictionaries

The Dictionary of the Older Scottish Tongue (DOST) has been compiled over a period of eighty years according to the historical principles laid down in the Oxford English Dictionary (OED). No attempt will be made here to give an historical outline of the project; that is available in Volume XII of the dictionary. Rather, I want to concentrate here on how the editing of DOST has furthered the craft of historical lexicography. There are two main lines I want to pursue: first that which I have called encyclopaedist, where information not strictly necessary to the definition of a word is introduced. The second I have called the question of phonemic variants, by which I refer to the editorial policy of splitting words into two or more entries on the basis of their phonology. Between the beginnings with Sir William Craigie and the final publication in 2002, DOST changed in both the above respects. It is these changes and the reasons for them that I wish principally to talk about. To facilitate presentation I will divide this history into three major phases identified by their editorial styles.

1.

Phase 1

During the first phase, a methodology based on that of the OED was established. DOST is indeed the historical dictionary that followed most closely the methodology of Sir James Murray: the layout of headwords, list of variant spellings and etymology, followed by the analysis of the word into senses and sub-senses, each accompanied by its paragraph of illustrative quotations,

50

Marace Dareau

imitates closely the pattern of OED. DOST copies this format so closely, of course, because its instigator and first editor was Sir William Craigie, who was Joint Editor of OED from 1901–1933. He looked on the DOST scheme, as he did the other dictionary schemes he proposed at this time, from an OED perspective. This is important in that the changes introduced by later DOST editors have subtly but very definitely altered that perspective. Craigie saw this historical dictionary of Scots as part of a process of filling out the history of English and thus allowing comparisons to be made between the various historical ‘periods’ of English. The seminal paper describing these views and plans for future dictionaries is “New Dictionary Schemes”, which Craigie read to the Philological Society on 4th April 1919. In this paper he put forward an ambitious proposal for a series of period dictionaries: Dealing as it [sc. OED] does with all periods of English, from the seventh century to the twentieth, it has been impossible for it (beyond certain limits) to devote special attention to any one of these. Yet each definite period of the language has its own characteristics, which can only be appreciated when it is studied by itself, and which are necessarily obscured when it merely comes in as one link in the long chain of the language as a whole. To deal adequately with each period it is necessary to take it by itself and compile for it a special dictionary, as full and complete as may be. When this process has been completely carried out for all periods, the task of comparison will be a fairly simple one. (Craigie 1925–32: 7)

He was aware that the Scottish material in OED is different from the other ‘periods’ he refers to. This is how he describes its inclusion in OED
: While the older Scottish tongue has thus received very generous treatment in the Dictionary, the appearances it makes there are necessarily scattered and to a great extent subject to accident. At the best, it is submerged in a great mass of earlier, contemporary, and later English with which it has little in common. Considered by itself it is a very definite thing, beginning with the fourteenth century, flourishing as a literary medium from about 1375 to 1600, and maintaining a precarious existence in writing till towards the close of the seventeenth century, when a new period definitely sets in and continues unbroken down to the present day. (idem: 9)

Here, Craigie describes Scots as “a tongue” and “a very definite thing”. It is treated in OED more or less as a dialect of English, though of course in some entries it goes off on its own track altogether. The question of ‘dialect’ or ‘language’ is not one into which I intend to enter here; I seek to do no more than try to explicate some of the rather mixed messages of Craigie’s approach.

DOST

By the time he came to write the Preface to Volume I of DOST, where he outlines his editorial policy, he accords this “very definite thing” the title of ‘language’ and in contradistinction to the later, post-1700 period, he clearly had come to the conclusion that Older Scots is a language not a dialect: This dictionary is intended to exhibit and illustrate the whole range of the Older Scottish vocabulary, as preserved in literary, documentary, and other records, down to the year 1600, and to continue the history of the language down to 1700, so far as it does not coincide with the ordinary English usage of that century. Words not found before 1600 are also included when they are not current, or are not used in the same sense, in English of the period, or when they have some special bearing on Scottish history or life… The dictionary, it will be obvious, is not merely a linguistic record. A large number of the words it contains are of historical or legal interest, are intimately connected with the older life of the Scottish nation or are descriptive of the special features of the country. The history of many of these has hitherto been imperfectly traced, and much light is thrown upon them by the fuller evidence here provided. (idem: I. vii)

This quotation illustrates also another aspect of Craigie’s attitude to compiling a dictionary of Scots. He was intensely interested in aspects of historical lexicography other than the purely linguistic. He was indeed interested in including information about the history of the society in which the language functioned. The possibility of insights into the history of societies in former times was one of the benefits to be gained from such historical lexicography. As he said in his proposal for the dictionary of Middle English in his “New Schemes” paper mentioned above: A complete dictionary of Middle English would be a work of marvellous richness and interest, not merely in respect of the language but for the light it would throw upon the manners and customs of the time. (idem: 7–8)

Such clearly was his interest; nonetheless his basic intention, in keeping with the philological theory of the time, was the analysis of language according to linguistic principles. The nature of a historical dictionary was defined by Murray in the Preface to Volume I of OED as providing the history of a language
: The Dictionary of the English Language is not a Cyclopaedia: the Cyclopaedia describes things; the Dictionary explains words, and deals with the description of things only so far as is necessary in order to fix the exact significations and uses of words. (1888: vi)

Craigie in his analysis of entries largely followed that methodology.

51

52

Marace Dareau

2. Phase 2 Phase two covers the period from the commencement of the Editorship of Adam J. Aitken till the point in the 1980s when I took over editorial responsibility for the output of the dictionary. These divisions indicate merely the sorts of changes in editorial style that I am interested in for the purposes of this paper. Aitken joined the Dictionary in 1948 and took over from Craigie as Editor in 1955. In his Preface to Volume III, Aitken endorsed Craigie’s editorial principles in the following terms: “The basic editorial methods remain, with only minor modifications, as he [sc. Craigie] devised them”. Aitken was nonetheless aware that the coverage of the language provided in Volumes I and II was inadequate. He was particularly conscious of gaps in the Dictionary’s corpus, and as well as adding many sources to the collection, he had others, already excerpted, re-read. A later paper published in Robert Burchfield’s Studies in Lexicography reads: In 1952, dissatisfied with the coverage of the existing collection, I launched a new reading programme, with more than 50 new voluntary excerptors, reading both printed editions and, mostly on microfilm, manuscripts. This greatly increased, not far short of doubling, the size of the collection for the letters still to be edited, and added also to the supplementary material for earlier letters, bringing the total collection to well over 1 million examples. (Aitken 1987: 98–99)

Aitken, in fact, more than doubled the list of sources. To c600 titles listed in Volumes I and II he added another c700.1 He wanted to produce a dictionary that was as complete a record of the language as was within his power. Aitken also widened the scope of the editing. Under his regime sense analysis was refined and the illustration of usage came more and more to be considered an important part of an entry. His attitude is set out in a handwritten note found among the Dictionary’s records: Since many readers consult the dictionary for precise definitions of archaic and Scots words and technical terms, the present editor, commencing with the letter J, has departed from Sir William Craigie’s cautious and conservative practice in this respect of providing brief, generalised, often portmanteau definitions, and has aimed at a more elaborate subdivision by usage, with fuller, more precise and more detailed definitions, sometimes accompanied by brief notes of an encyclopaedic nature when the material for this lay to hand.

1.See The DOST Corpus, DOST XII: clxiii–clxxiv.

DOST

Further, it has been a principle to supply as far as possible those quotations which are most helpful in this direction.

As regards coverage, Aitken aimed at exhaustiveness for the pre-1600 linguistic record. At the same time, he continued Craigie’s policy of including material belonging to the seventeenth century along the lines indicated by Craigie in the passage from the Preface to volume I quoted above, that is, words distinctively Scottish in some regard were included, words apparently the same as in English and which began after 1600 were omitted. As a summation of Aitken’s long Editorship of DOST what I have said here is wholly inadequate. However, it pulls together what are in my view his most critical influences: the expansion of the source material, an essential step if full coverage of the language was to be achieved, and the expansion of what he saw as relevant to a description of the language into all sorts of usages that under Craigie were not explicitly illustrated. The principles remained linguistic, the dictionary was a record of a language, but encyclopaedic material is much more in evidence than in Craigie’s volumes. Aitken’s successor, Dr. James A. C. Stevenson, further refined the illustration of meaning and semantic and grammatical usage. His philosophy echoed Aitken’s with respect to the scope of the Dictionary. He said in a talk given in the 1970s: But the aim of DOST is not simply or even chiefly the definition of unfamiliar or obsolete terms. It could rather be described without undue pretension as an attempt to provide a key to the whole range of Scottish culture from 1200–1700. There is an abundance of useful and curious information, much of it not available elsewhere, on every aspect of life in these five centuries, and the quotations are sometimes supplemented by references to authoritative treatments to be found elsewhere.2

The attitudes so far developed throughout Phases 1 and 2 might be summarised in the following terms: Specifically Scots material is thoroughly covered in the first two volumes, but a truly encylopaedic attitude, that is, where more information describing things is given than is strictly necessary to the definition of the word, is not exhibited in Craigie entries; for instance, Bordour n has the sense “the border or border districts between Scotland and England”, as does the OED entry Border n., but goes into no further detail. March n, on the other hand, edited of

2.Undated typescript in DOST ’s archives.

53

54

Marace Dareau

course by Aitken, is considerably more encyclopaedic. It has the same sense, “The Scottish border with England”, divided into a number of sub-senses, giving such phrases as Law of the Merchis, and Day of march. These appear in small type, indicating their encyclopaedic status and lesser relevance to the entry where a dictionary entry describes an item of language. More recent editing is exemplified in Wardan n sense 2b “The governor or military commander of the Scottish/English Border” where examples are distinguished, for instance, as the Scottish or English wardan and quotations are much lengthier. One of them is nine lines long (the last quotation in paragraph (3))3 and describes the situation with regard to the authority of the sheriff and the wardan in Teviotdale. The quotation serves to describe the nature of the wardan’s office and power, adding much encyclopaedic information to the basic meaning of the word and as such goes well beyond what Craigie or probably even Aitken would have thought appropriate. The full development of the aims of such a lengthy project takes time, of course, and this is what happened over the half-century post-Craigie. The developments marking out Aitken’s period were that he acquired much more material to work with and had a keenness to exemplify usage, though he was still intellectually thirled to Murray’s view that a dictionary explains words, despite his interest in all the other things it might do. If Aitken did not think of himself primarily as an encyclopaedist, Dr. Hans Meier, himself formerly a member of DOST’s editorial staff, clearly did. In a review article in 1962, he described the dictionary produced by Aitken in exactly those terms: “Under Aitken, DOST has become more fully than ever an encyclopaedia of older Scottish culture and a first class reference book for Scottish historiography” (1962: 445). However, Aitken still described what he was doing in terms primarily of linguistic analysis and where he included encyclopaedist information it remained secondary: These more or less linguistic considerations are our main business as editors of DOST. But since they are achieved by a display of quotations from the original

3.As for the government and customes of the country [sc. Teviotdale] they had a sheriff as other shires had but his power was much eclipsed by an office and jurisdiction which was here called the wardan office [?erron. for officer] who was a person of great command and power, being the king’s lieutenant and in time of war, exercised martial law; and the sheriff and all the shire..were bound to attend him & to be answerable to his court. He was the kings chief justice in criminals within his jurisdiction.. The Scots and the English wardens had upon the border their meetings [etc.]; 16.. Macfarlane’s Geog. Coll. III.159.

DOST

texts we cannot but also provide incidentally, through the quotations, information about the things or concepts denoted by the words treated…From the letter H we have tried to make our definitions of precise terms of whatever specialty as accurate as possible and have added a large number of descriptive notes to technical terms of this sort. Naturally this often requires research in secondary literature…Much of the results of this is, among dictionaries, unique to DOST. (1981: 41–42)

The Dictionary thus almost by stealth came to include more and more encyclopaedic information even although the structure of entries, like those of OED, was still largely determined by linguistic considerations.

3. Phase 3 During the later 1980s, after both Aitken and Stevenson had retired, I eventually became responsible for the editorial aspects of DOST. My own view with regard to the encyclopaedist issue in a sense merely went further down the road on which my predecessors were already travelling: that the Dictionary might be more useful to all its users if the centrality of the language in society was treated as one of the fundamentally important elements of the editing process. But further to the question of the language in society was the question of to whom the dictionary was addressed. Aitken had emphasised the need for creating a record of the language: and in those days, the 1960s and 1970s, we never really considered the user. It was my involvement in the Concise Scots Dictionary (CSD) that made me aware of the human element on the other side of the production process. It was Iseabail Macleod, senior editor of CSD, who had worked in publishing, who made me aware of how a dictionary could be improved by thinking through what might be helpful to the user; especially users who were not necessarily primarily interested in language per se. We had always edited according to the lexicographical standards set by OED, wherever possible following its entry structure. Comparison with the OED entry for the ‘same’ word was a usual and useful part of the process of devising an entry. However, the temptation was always to be too reliant on the OED entry, and if the process of constructing an entry relied too heavily at the outset on the parameters of the OED entry, the opportunity to reveal the natural development of the word by a consideration of the Scots material on its own terms might be lost. Now that the view that entries would benefit by being

55

56

Marace Dareau

‘Scots-centred’ was incorporated as part of our methodology, this drawback was avoided by discouraging editors from looking at the OED entry until their own entry was roughed out and the nature of the material assessed on its own merits, the opposite way round from the editorial process I learned when I joined the Dictionary in 1969. Thus, from around the beginning of S, while, of course, not neglecting the essentials of etymology and linguistic history, the functioning of the word in society and its cultural history have been given as much importance as its linguistic characteristics. It was intended that the Dictionary should make it a primary objective to display the history of medieval and early modern Scottish society as it is recorded in the language of that society, with a shift towards a greater emphasis on the sociolinguistic aspects of language analysis. In this way the material was organised to display whatever aspect of the word seemed most important to the history of Scots, comparison with English being treated as a secondary matter. Thus, Service n. is defined firstly in its relation to the social order whereby land is held in return for service to a superior rather than as in OED, where the major section division I. “The condition of being a servant; the fact of serving a master” is contrasted with II. “The work or duty of a servant; the action of serving a master”, which is a purely linguistic distinction. It is worth pointing out that this is a change in point of view that in no way inhibits comparison with English. Indeed, since these parameters are clearer, it may well act to facilitate it. What I intended to do was to redress an imbalance deriving from a more strongly centralist, imperialist view of the language than is now plausible. Further examples of entries where the arrangement of senses must differ widely from that in OED in order to reveal adequately the modalities of Scots usage are as follows. In the entry Set v., the first major sense grouping is “to cause to sit, to seat, passing into, to cause to take place, to appoint, to arrange”. (The equivalent senses in OED are arranged in two sections widely apart: I, “to cause to sit” and V, “to appoint, arrange, …establish”.) This arrangement in DOST allows the development of the commonest sense in Scots, “to let or lease out (property)” (located in Section V, sense 57 in OED) to be placed close to the beginning of the entry, as is appropriate for a major sense, and relates it to the other senses concerned with institutional and legal affairs: “to cause a deliberative or judicial body to sit”, “to arrange, organise” and “to appoint” in relation to various public affairs. These senses, all concerned with the organisation of society, develop in a satisfying way from the core concept of sitting or being seated to arrange such matters, and contrast cleanly with the other major elements of the word in Scots, which are concerned with relationships of

DOST

direction (descent) and location (placing in a position, physically or mentally). This arrangement allows the structure of society to be revealed, once again, as a core concern. In the entry Tik(k)at n. the first sense in Scots relates to formal notification of the business of a court, “A notice posted in a public place, esp. a summons, advertised by being posted on the Edinburgh Tolbooth door”. Its earliest example is 1515, making this the earliest example of the word as yet known in either English or Scots. The first sense in OED’s entry “Ticket n”. is “a short written document”. This dates from 1528, the earliest English example, and the only one before the last decade of the sixteenth century. Sense 2 in the DOST entry is the equivalent of OED’s sense 1. The relationship to official business runs throughout the Scots material, reinforcing the dating criterion to commend it as the first sense in the entry. The high level of record and legal material in the documentary sources undoubtedly directs our understanding of the development of the language in its early stages in particular ways. This is fundamental, I believe, to an understanding of Older Scots and an important point when we come to consider especially the apparently low level of influence from Gaelic. My last example reveals, I think, Gaelic influence which can only be detected by analysing the language according to these more sensitive methods. In Wat(t)ir n. the senses have been ordered according to the predominant usages in Scots, giving much more prominence to the word as a geographical term, and especially as the term for a river, flowing water. This sense seems to be at least influenced by Gaelic usage, a point suggested to me by Geoffrey Barrow, Emeritus Professor of Scottish History at the University of Edinburgh. These geographical senses make up a major section of four main senses with a number of sub-senses, in contrast to OED’s single sense. Indeed, from S onwards, this approach informs every entry. This approach was new, and yet it was part of a process whose origins can be discerned at the very beginning of DOST
: This slow process of severing ourselves from dependence on OED goes back in many ways to the very beginnings, and to Sir William Craigie’s original determination to compile a separate dictionary of Scots (1925–30: 6–9) and it is the fuller, one might say, the logical, outcome, of treating every instance and word as an example of a language (the Scots language) that has aspects in common with its larger neighbour, English, but must be treated, in the first instance at least, as separate and unique in order to maximise our knowledge and understanding not only of the language as such but, equally importantly, of its social function. (Dareau forthcoming 2004)

57

58

Marace Dareau

To sum up, the Dictionary’s having an encyclopaedic function has been acknowledged right from the beginning of DOST. However, my editorship foregrounded the view that the primary aim of historical lexicography is to provide the most useful tool possible for all those likely to be users of the Dictionary. This philosophical emphasis was the more easily put into practice in the 1980s in as much as modern publishing technology has removed the premium on short quotations and we no longer suffer under the problems of paper shortage and hot-metal printing. Both Craigie and Aitken suffered constantly from problems of scale, the need to minimise the quantity of published material, due both to the costs of printing and scarcity of paper, especially during and after the Second World War. Hot metal type lasted into the 1980s and made the use of long quotations expensive. This led to a preference for short quotations where the correctness of the analysis sometimes cannot be made clear from the material quoted. Computer technology, when it became available in the mid 1980s, reduced these difficulties substantially. By a different emphasis in the way we now analyse the material, as well as by the use of longer, more self-explanatory quotations, my intention was to develop the concept that a dictionary and an encyclopaedia need not be mutually exclusive and that a research tool such as DOST can be more than a record of the language without losing that necessary aspect of its function. The fact of my seeing the output volumes as a research tool brings me to the second of my themes and persuaded me also to make another change that Aitken had half made. That was to reject finally the concept of separate entries dependent upon spelling variation interpreted as suggestive of a separate etymological source, so-called ‘phonemic variants’. Craigie came from a background of comparative philology. The primacy of this aspect of linguistics in Craigie’s experience seems to have been the source of this characteristic of his style of editing. Thus there are five separate entries associated with the various phonemic strands of the verb ‘give’: Gef, etymologised as Variant, of Gefe v (representing a short mid vowel and unvoiced consonant), Gefe, as Variant. of Geve v. (representing a long mid vowel and unvoiced consonant), Geve itself (representing a long mid vowel and voiced consonant), Gif, (representing a short high vowel and unvoiced consonant) and Give, etymologised as Variant. of Geve v., Gif v. (representing a short high vowel and voiced consonant).4 The fact that these spellings may indicate the 4.
This particular word forms the subject of an essay “A Re-editing of GIF”, compiled by the staff of DOST and contributed to The Nuttis Schell (Macafee & Macleod 1987: 25–57),

DOST

several different phonological sources of the word seems to have weighed heavily with Craigie. Of course this is a procedure that is used to some extent in OED (though Craigie here too seemed to use it more often than other OED editors), but it is much less common there than in the Craigie volumes of DOST, for OED has the modern spelling to give it an entry headword. Aitken had reduced the number of variant entries enormously. Nonetheless he, again seeing the dictionary as a record rather than something people had to use, saw it as a valid reflection of the nature of the language. Most large words, where it was appropriate, and even many small words, were split into variants of this sort; for instance Mikill, Mekill and Mukill, where Mekill is described as a variant of Mikill with e for i, Mukill as a variant of Mekill or Mikill with unexplained vowel change. Such entries are difficult to consult. Indeed this policy had been criticized as early as 1962 in the review by Dr. Hans Meier mentioned above: One disconcerting problem is the separate entry of spelling variants whenever these are numerous and/or can be interpreted to represent phonematic differences, i.e. separable morphemes. Since the order is strictly alphabetical, one may have to go to many different places for “the same word”. (1962: 447–448)

Although time was not seen as a problem at this date, at least not the problem it was later to become, and the main criterion was excellence, it was timeconsuming to edit in this fashion. (The method was to edit all the material together as a single entry then split it up into the desired number of variant entries, defining the senses of the variant entry or entries by means of crossreferences to the main entry.) Above all, however, such divisions are liable to obscure semantic variation, wherein so much that is characteristic of Scots is to be found. The policy from S on was that all spelling forms that could be designated as variants would be included in one entry and that semantic issues which might incline the editor to one or more than one entry would be dealt with on their merits. Examples of variants now included in single entries are sall and schall (s.v. Sal(l aux. v.), speke and speche (s.v. Spech(e n.), ta and tak (s.v. Ta(k v.), to and till (s.v. To, prep., etc.), thesaurar, tresaurar and thresaurar (s.v. Thesaurar(e n.), treasurer, and waw, wall and waiff (s.v. Waw n.2), a wave, all of which fall into the category of simple variants with no semantic questions involved, and each of which would under previous editors have been two or three separate entries.

presented to A. J. Aitken in honour of his life’s work and, in the case of this contribution, especially of his outstanding achievements in the lexicography of Older Scots.

59

60

Marace Dareau

The etymologies of these entries contain, of course, all relevant comparative material. For instance the etymology of Thesaurar(e lists the ME and ENE evidence for all the spellings in chronological order, followed by Old French tresor(i)er and late L. thesaurarius. The relationships of the different forms to their sources is pretty much self-evident. There was seen to be no need to separate out the tre-/thre- forms from those in the-, deriving the former from Old French and the latter from late Latin as does OED. Whether these are separate ‘words’ or not does not concern me. Rather, this method produces entries where all relevant information for each semantically defined unit is found at a single site. As a safeguard for those users particularly interested in orthographic or phonological issues, spellings which would by the previous policy have been in separate entries are located in separate paragraphs wherever these differences occur throughout the combined entry. What was of value in the separate entries is therefore not lost. For instance, all of the equivalent Scots material that is to be found in the OED entries Sire and Sir is included in the single entry Sir(e n., encompassing all of the forms sire, sir and schir. This allows the user to see how the meaning of the word developed from the term for a monarch or magnate to the title of a monarch, etc., then spreading in both these senses to lower levels of society. On the other hand, since sir and sire occur as separate words in modern English, it makes sense for the OED to determine the word division by this shortening of the vowel before a proper name. It is clear that in some cases this will prove a contentious policy, especially, perhaps, in such cases as To, Till, prep., etc. where to and till clearly are not etymologically related, so, it might be argued, they should by no means be included in a single entry. However, in Scots they function as if they were the ‘same word’. Putting all the material together at a single site allows easy comparison of how the two forms relate to each other. By 1994, when the To, Till entry was edited, DOST had undergone a review of editorial method and management in relation to its costs to completion. The outcome of that process was that the editing of T-Z would have to be completed in six years to achieve completion in 2000. T-Z in published form makes up two and a half volumes of the series, not far short of a quarter of the whole dictionary. The six years in which these volumes have been completed compares with an average of twenty-three years for each of the other three quarters, which in total had taken almost seventy years. So the time taken to separate the To, Till material into two mutually cross-referred entries was better spent editing combined entries.

DOST

The general thrust of the changes towards simplification, especially the abandonment of splitting so-called phonemic variants into separate entries, certainly met the needs of the project after the 1994 Review.5 This is not the occasion on which to go into the minutiae of the process that led to successful completion a mere six months beyond the deadline, and completing the actual editing within the deadline of December 2000, as we had promised. A few remarks are necessary, however, to give due credit to all those involved on this occasion of celebration.6 I also want to give some idea of how we did actually achieve what seemed in 1994 to be the unachievable. First of all it has to be stressed that no diminution in quality would be acceptable. William Gillies, Professor of Celtic in the University of Edinburgh, who accepted the challenge of managing the project for its final years, made the point at our first meeting with him that he would only consider doing so if we could assure him that we could complete as qualitatively good a dictionary as that already in print. This too had been a point made by the Review panel, and we ourselves, the staff of DOST, would not have been prepared to work in any other way. All that was left to us was to find editorial practices that would allow us to do carry this out. Having determined that the method of editing in the traditional way, then keying only the slips that would actually be published, was the only feasible course of action, how could the traditional editing process be speeded up without loss of quality in the final product? The approach we found effective was to test stringently each and every process in the editing of every word with the questions “is this really necessary?” and “is there a simpler equally satisfactory way of doing this?” The simplification of sub-sense divisions was the most useful change to previous practice. Divisions were restricted to the obvious ones, allowing speedy sorting of subsenses with a minimum of soul-searching over the rights and wrongs of a particular case. This was applied especially to the minor grammatical or semantic divisions displayed under the sub-section labels (1), (2), etc. The effect of this was to reduce editing time by simplifying choices. Consider, for instance, Restor(e v. 1. The quotations section is divided into six sub-sections: (1) separates things; (2) persons; (3) non-material things; (4) examples with the

5.See The History of DOST, Phase IV, DOST XII: XVI–XIX. 6.A special day of celebration for the publication of the last volume of DOST was held during the 12ICEHL conference, and was attended by past and present members of staff. Dr Philip Durkin made a speech of congratulation at a reception sponsored by Oxford University Press.

61

62

Marace Dareau

construction to something; (5) with the construction restore bak; (6) absolute (absol.) examples, i.e., those having no dependent construction. To achieve this result, editing time would be taken deciding what sorts of divisions to make. Then decisions would have to be made as to whether particular quotations belonged in one division or another, for instance, whether the quotation “The naturale heit consumys the humedite naturale & radicale that is restorit be mete, drink and nurising” should be assigned to (1) a thing or to (3) a nonmaterial thing. Furthermore, though one of the rules of editing limited the number of quotations per paragraph, there was no limitation on the number of paragraphs. Having six divisions permitted one, in theory, to deploy seventytwo quotations, not allowing for separate spelling paragraphs which might also be required. Furthermore, the distinctions made were frequently not of any specific importance to Scots. So the principle was adopted that divisions be reduced to a minimum. In the case of Restor(e v. this would have reduced the divisions (1), (2), (3) and (4) to one paragraph with only the constructions evidenced in (5) and (6) treated separately. This could have reduced the number of quotations in the entry to twenty instead of thirty-six, making a saving of sixteen quotations in the cost of keying, proofreading and printing. This reduction in copy would benefit the whole process. We found it was indeed possible to save time and effort in this way without any real loss to the coverage of the Dictionary. The level of separation used previously had led to over-exemplification in many cases. Exceptions were always permitted as this in itself could help difficult, time-consuming decisions to be avoided. Some previously used procedures had to be dropped. Formerly editors revised their own copy six months after first editing. This method of revising had been found to be very effective, provided that the six-month gap was adhered to. After 1994, schedules became too tight to allow for a six-month gap, so this was abandoned and replaced by immediate review by another editor. As one means of counteracting the very definite loss of the second bite at the cherry that an editor had before 1994, it was made a principle that every difficulty was resolved if at all possible at the first editing. This gave the reviewing editors a possible solution to consider, which at least allowed them to focus on whether the proposed solution was feasible or not. It is all too easy to shuffle through the evidence for a particularly difficult example or examples, never coming up with any sort of entry. Yet at the end of the day that is what every editor must achieve — an entry of some sort. Even an entry which turns out to be wrong is better than no entry at all, for at least that puts the evidence, misconstrued though it may be, into the public domain for other, perhaps

DOST

better informed, scholars to encounter and correct. Of course some difficulties could not be resolved first time round, but it set a tone that helped things along rather than acting as a brake. Further time-saving procedures were used, such as the approach to problem solving. A body of supporters had been building up over the years: in preference to searching in secondary sources, as had been Aitken’s preferred method, later staff had built up a network of experts willing to answer questions by correspondence. Now advice required on particular problems could be sought by e-mail from scholars all over the world. The completion of DOST was something many scholars gave time and expertise to during this period, to the extent that difficulties, with manuscripts, for instance, which would have been unresolvable in the past, now could be solved. Dr. Keith Williamson,7 for instance, was always willing to check a quotation from the manuscripts he was working with and the staff of the National Archives of Scotland on quite a few occasions supplied photocopies of pages of Edinburgh Testaments. Expert opinion was not restricted to linguistic questions either; the whole user community was eager to see the completion of DOST and willing to speed it along. The quality and size of the team was also critical. It consisted of the three full-time editors and one full-time and two part-time editorial assistants, rising to five editorial assistants from 1997. The team combined size and experience to a greater degree than at any time in the past. The importance of experience had been emphasised by Aitken during an earlier crisis in 1980: Some remarks [were made]… that one could buy three or more junior editors for the price of two seniors and that younger people worked faster than older. The implication of the former and the fact of the latter are untrue in the context of DOST (or any other similar dictionary, such as MED), at least if any regard at all is given to quality. More experienced persons in fact produce acceptable results much faster, because they already know much the junior has to find out ad hoc because they are more skilled at analysis and at the devising of definitions and because they have confidence in their findings where a junior hesitates and vacillates.8

The solution of replacing more experienced, more expensive staff by cheaper juniors was also suggested in 1994. However, over the succeeding six years,

7.Research Fellow, Institute for Historical Dialectology, University of Edinburgh. 8.Private letter to Angus McIntosh, Professor of English Language, University of Edinburgh, in DOST’s archives.

63



64

Marace Dareau

Aitken’s view was confirmed. Over that period, the team of editors remained unchanged and it is undoubtedly in large part due to their commitment and experience that, in 2001, the project was drawn to a successful conclusion. The work of the editorial assistants changed dramatically. Instead of keying in-house then checking each quotation individually, it was hoped that the timesaving required on the production side would be made largely by employing a data-entry agency to key the edited copy from the slips. An extended trial conducted with SPI (Technologies) Ltd demonstrated the practicality of this approach. The effectiveness of SPI’s contribution to the completion of the Dictionary cannot be overstated. The job of the editorial assistants became one of meticulously proof-reading the dictionary entries, checking not only the actual text but layout and structure — an enormous task, quite different from anything clerical or editorial assistants had undertaken in the past. An important consequence of keying the material at a relatively early stage in the process was the capability that was gained to sort the quotations electronically. The extensive verification of the quotation material required prior to publication was greatly accelerated by this ability to check all the quotations from one text in a single pass. The editing of DOST was completed on the 11th of December 2000 and proof reading by the end of June 2001.

References Aitken, A. J. 1981. “DOST: How we make it and what’s in it”. Proceedings of the Third International Conference on Scottish Language and Literature (Medieval and Renaissance) ed. by R. J. Lyall & F. Riddy. 33–51. ———. (1987). “The Period Dictionaries”. Studies in Lexicography ed. by Robert Burchfield. Oxford: Clarendon Press. 94–116. Craigie, William. 1925–32 (1931). “New Dictionary Schemes”. Transactions of the Philological Society. 6–11. Dareau, M. Forthcoming 2004. “DOST and Older Scots Scholarship”. Older Scots Literature, vol.1 ed. by Sally Mapstone. East Linton, Scotland: Tuckwell Press. ——— et al. 1987. “A re-editing of GIF”. The Nuttis Schell: Essays on the Scots language ed. by C. Macafee & I. Macleod, 25–57. Aberdeen: Aberdeen University Press. Dictionary of the Older Scottish Tongue (DOST). 1937–2002, ed. by William A. Craigie, A J. Aitken, J. A. C. Stevenson, M. Dareau et al. Aberdeen: Aberdeen University Press; Oxford: Oxford University Press. Meier, Hans. 1962. Review of DOST. English Studies XLIII: 444–448. Murray, James A. H. 1888. Preface to A New English Dictionary on Historical Principles. Oxford: Oxford University Press. Oxford English Dictionary (OED). 1884–1933, ed. by Sir James A. H. Murray, Henry Bradley, Sir William A. Craigie & Charles T. Onions. Oxford: Oxford University Press.



Image schemata and light A study in diachronic lexical domains in English Javier E. Díaz Vera Universidad de Castilla-La Mancha

1.

Aims and method

In this paper I am going to analyse the lexical domain of light in Old English and its evolution into later English, as an example of how we can reconstruct an inventory of conceptual categories from a mainly diachronic perspective. Within our model of lexical analysis, the Functional-Lexematic Model (FLM), the term ‘lexical domain’ is used to refer to the set of lexemes which together lexicalise all or part of a conceptual domain (Faber and Mairal Usón 1999: 59 and their Appendix 1 for a complete list of examples). Lexical domains are divided into ‘subdomains’, i.e. meaningful sets of lexemes within a lexical domain, which form the basis of its internal structure and focus on a particular area of meaning. Moreover, each domain or subdomain has a superordinate term, the ‘genus’, by means of which the other lexemes are directly or indirectly defined. Meaning definitions include a second type of semantic information, referred to as ‘differentiae’, which distinguishes a lexeme from others in the same lexical domain. It is important to emphasize here that our method of analysis is bottom-up (or data-driven). This implies that our inventory of lexical domains is not based on extralinguistic criteria, but on the detailed analysis of the definitions of each individual lexical item. In this way the areas of meaning established have a more solid basis and are not the product of intuition. Our definitional method, known as ‘definitional analysis’ (Faber and Mairal Usón 1999: 79–110), is made up of a combination of the distinctive feature approach (Bendix 1966) with semantic role assignment (Fillmore 1982), which is directly extracted from the

66

Javier E. Díaz Vera

argument structure of the unit’s predicate frame (Dik 1989). Broadly speaking, verbs with definitions having the same genus belong to the same lexical domain or subdomain. Once we have grouped lexemes into lexical domains and subdomains, we can elaborate lexical hierarchies where lexemes are ordered according to their degree of semantic and morphosyntactic coverage.

2. Lexical diachronic principles In the FLM, defining a verb implies locating it in semantic space. The initial step in our reconstruction of the internal hierarchical structure of the verbal domain of light consists in determining which Old English predicates had a greater semantic coverage. According to the FLM, the hierarchical semantic ordering of lexemes is correlated with the systematic distribution of their syntactic configuration. This principle has been formulated in terms of the ‘Lexical Iconicity Principle’, which implies that the greater the semantic coverage of a lexeme, the greater its syntactic variation (Faber and Mairal Usón 1994: 210–211). Verbs of light do not normally show a very high degree of syntactic variation. In fact, most of these verbs are intransitive and simply require a first argument in the semantic role of Force. In order to distinguish different degrees of variation, we must take a closer look at the nature of the entities that can occupy subject position with each Old English verb of light. Through the semantic analysis of a representative number of instances of each Old English predicate, we have found that OE scı¯nan can be used with up to four different types of subject, corresponding to the highest degree of variation within this domain (all the examples used in this paper have been extracted from Healey et al. 2000). Besides the ‘Lexical Iconic Principle’ described above, our decision to assign archilexematic position to OE scı¯nan is justified on the grounds of two diachronic principles, which we have found useful to specify the hyperonymic status of many units in previous lexical analyses: (1) Lexical Etymological Principle: the more central a member is within a conceptual category, the more likely to have been lexicalised in a former stage of the history of the language concerned. (Faber and Fernández Sánchez 1996) (2) Lexical Derivational Principle: The greater the semantic coverage of a lexeme, the greater its number of derivational formations. (Díaz Vera 2002: 55–56)

Image schemata and light

Table 1.Argument restrictions for OE scı¯nan Force

Example

Seo sunne on hadrum heofone beorhtost scineð (1) SV (a) S = heavenly body (natural light) [Bo: 019200 (9.21.1)] “The sun shines in the clear sky.” (2) SV (a) S = metal, polished, smooth, or glossy surface (reflected light)

Scean scir werod, scyldas lixton [Ex: 003700 (125)] “Shone the glittering army, the shields were bright.”

(3) SV (a) S = colour (beauty)

Seo hwitnes þære lilian scineþ on þe [HomU 18: 003500 (82)] “The whiteness of the lilies shines on you.”

(4) SV (a) S = human (positive qualities)

Þonne scinað ða rihtwisan swa swa sunne on hyra fæder rice [Mt(WSCp): 045100 (13.43)] “Then shine the righteous as the sun on the reign of their father.”

The ‘Lexical Etymological Principle’ reflects the general tendency of Old English strong verbs (such as scı¯nan) to cover wider areas of semantic space than weak verbs, which lexicalised at a later historical stage. The ‘Lexical Derivational Principle’ indicates that the higher positions within a given semantic hierarchy are normally occupied by underived verbs that frequently serve as a basis for the creation of new derived words. This is clearly the case with OE scı¯nan, from which such verbs as a¯scı¯nan, bescı¯nan, gescı¯nan, geondscı¯nan, oferscı¯nan and ymbscı¯nan are derived. This being so, we can confidently give OE scı¯nan archilexematic status, and define it as ‘to give off light, to shine’.

3. The semantic parameter stability of light In order to divide the semantic space corresponding to Old English verbs of light, we must be able to ascertain the semantic nuances between the remaining predicates within this dimension, and their semantic relation with the domain hyponym OE scı¯nan. We will take into account two basic semantic parameters: stability and degree of intensity of light. Regarding the semantic parameter ‘stability of light’, verbs of light can be divided into two main groups: verbs conveying a stable light and verbs conveying an unstable light (Faber and Pérez 1993). The unmarked pole corresponds

67

68

Javier E. Díaz Vera

to stable light, as represented by OE scı¯nan, which is the type of light we normally find in nature. Verbs referring to unstable light represent thus a subdomain within the general domain of light. Through the analysis of a representative number of occurrences of each predicate, and with the help of the definitions given by the Oxford English Dictionary (OED) and the onomasiological classification proposed in the Thesaurus of Old English (TOE
; Roberts and Kay 1995: 140–142), Old English verbs of light have been grouped as shown in Table 2. Table 2.The semantic parameter stability of light TO SHINE .BEORHTIAN ¿ .BLICAN ¿ .GLOWAN .L¿I HTAN .SC¿I MIAN .SC¿I NAN .OFERSC¿I NAN .TWINCLIAN

TO SHINE UNSTEADILY ..BLICCETAN ..FYRCLIAN ..GLITENIAN ..GLISNIAN ..RÆSCETTAN ..SCIMERIAN ..SPEARCIAN ..SPIRCAN

4. The semantic parameter intensity of light Our second semantic parameter has to do with the degree of intensity of the beam of light perceived by the speaker. The unmarked degree of light intensity corresponds to ‘bright light’, as expressed by OE scı¯nan. The remaining verbal predicates of the field occupy different positions on the scale of intensity, from ‘dim light’ to ‘extremely bright light’. In order to reconstruct the approximate position each lexeme occupied within this scale, we have used the translations given by Old English dictionaries (Bosworth Toller 1898, Hall 1960 and, where available, Cameron et al. 1986–1996). However, these lexicographic definitions must necessarily be validated through the detailed analysis of such different factors as etymology, collocations and Latin glosses. Our analysis indicates that the intensity parameter is relevant within the definition of the following Old English verbs of stable light (Table 3).

Image schemata and light

Table 3.The semantic parameter ‘intensity of light’ in verbs of stable light to shine dimly

TO SHINE

to shine fairly brightly

to shine very brightly

to shine extremely brightly

TWINCLIAN

¿ SCINAN

BLEORTHIAN ¿ BLICAN

¿ GLOWAN

OFERSC¿I AN

Our arrangement is borne out by such quotations as the following (the adjectives and adverbs used to interpret the semantics of each predicate are underlined): (1) OE twinclian: to shine dimly SV a. S = something that gives off light (Fo) e.g. Ic þæt lytle leoht geseah twinclian [Bo: 123800 (35.97.16)] “I saw that little light twinkle.” (2) OE blı¯can: to shine fairly brightly SV a. S = a surface that reflects light (Fo) e.g. ðæt ðu ðære gyldnan gesiehst Hierusalem weallas blican [MSol: 006600 (231)] “That you see the golden walls of Jerusalem glitter.” (3) OE glo¯wan: to shine very brightly SV a. S = something that gives off light and heat (Fo) e.g. Lecgað ða isenan clutas hate glowende to his sidan [ÆCHom I,29: 423.144] “Lay the iron plates glowing hot to his side.”

As for Old English verbs of unstable light, we have found that the parameter of intensity is not as relevant as in the case of verbs of stable light. Rather, Old English verbs of unstable light are organised in terms of manner, depending on the different ways the emission of light takes place. This is the case with OE glisnian, a verb used to indicate that an unstable light is reflected by a wet surface, or OE spearcian, which refers to the action of light sparkling in a continuous way.

69

70

Javier E. Díaz Vera

(4) OE glisnian: to shine unsteadily SV a. S = a wet surface that reflects light (Fo) e.g. glisnaþ glæshluttur [MRune: 001200 (29)] “glistens bright as glass.” (5) OE spearcian: to sparkle continuously SV a. S = something that gives off light (Fo) e.g. He spearcade, ðonne he spreocan ongan fyre and atre [Sat:002900 (78)] “He sparkled, then he started to speak fire and poison.”

The intensity parameter is relevant within the definition of the following Old English verbs of unstable light (Table 4). Table 4.Semantic parameter ‘intensity of light’ in verbs of unstable light to sparkle dimly

FYRCLIAN

TO SPARKLE

to sparkle brightly

SPIRCAN

SCIMERIAN RÆSCETTAN GLITENIAN BLICCETAN

Once our analysis of these two semantic parameters has been completed, we can propose the following reconstruction of the internal structure of this part of the domain, along with the definitions proposed for each individual predicate (Table 5).

5. The causative domain: To cause something to give off light Another subdomain within the general domain of light includes a small set of transitive verbs whose general meaning corresponds to the action ‘to cause something to give off light’. Most of these verbs are derived from OE scı¯nan and lı¯htan through the use of the prefixes a¯-, be-, ge-, geond-, in-, on- and to¯-. Whereas the last four prefixes are normally used to indicate the direction of the

Image schemata and light

Table 5.OE verbs of light Verbs of light 1. To give off light, to shine OE scı¯nan: to give off light, to shine .OE oferscı¯nan: to shine extremely brightly .OE scı¯mian: to shine .OE lı¯htan: to begin to shine .OE twinclian: to shine dimly .OE beorhtian: to shine brightly ..OE blı¯can: to shine brightly by reflecting light .OE glo¯wan: to shine with a very bright light and heat 1.1. To shine unsteadily .OE spircan: to shine unsteadily giving off sparks of light, to sparkle ..OE scimerian: to sparkle brightly …OE glitenian: to sparkle brightly (of stars) …OE bliccetan: to sparkle brightly (of lightning) …OE ræscettan: to sparkle brightly making a crackling sound (of fire) ..OE fyrclian: to sparkle dimly ..OE spearcian: to sparkle continuously changing from bright to faint .OE glisnian: to shine unsteadily by reflecting light on a wet surface

beam of light in relation to the place illuminated (e.g. inside, on, towards) the first three affixes act as transitivity markers. The highest frequency in our corpus corresponds to the predicates gescı¯nan and gelı¯htan; following the definitions proposed above for their non-causative counterparts, these two verbs will be defined here as: (6) OE gescı¯nan: to cause something to give off light SVO S = human (Ag) O = something that gives off light (Go) e.g. forðæm hit sio sunne ne mot gescinan [CP: 165500 (45.337.10)] “Since the sun couldn’t make it shine.” (7) OE gelı¯htan: to cause something to begin to give off light SVO S = human (Ag) O = something that gives off light (Go) e.g. þætte seolfa god wolde helwarum ham gelihtan [Sat: 015200 (428)] “That God would make hell shine.”

71

72

Javier E. Díaz Vera

Other Old English predicates included in this causative subdomain are: (8) OE breohtnian: to cause something to be bright SVO S = somebody or something (Ag/Fo) O = somebody or something (Go) e.g. god geberhtnade hine on hine seolfne [JnGl (Ru): 061700 (13.32)] “God made him bright on himself.” (9) OE a¯lı¯htan: to cause something to be bright by shining light on it SVO S = somebody or something (Ag/Fo) O = somebody or something (Go) e.g. hi scinon on ðære heofenan fæstnysse & alihton ða eorðan [Gen: 001500 (1.15)] “They shone on heaven’s firmament and lit up the earth.” (10) OE a¯tendan: to light a fire SVO S = human (Ag) O = fire (Go) e.g. and geseah ðær feower ormæte fyr atende [ÆCHom II,22: 005500 (193.94)] “And saw there four people lighting an immense fire.” (11) OE feormian: to cause something to shine by friction SVO S = human (Ag) O = surface (Go) e.g. he feormað his bernes flore [Lk(WSCp): [011800 (3.17)] “He will cleanse the floor of his barn.” (12) OE b¯ywan: to cause something (of metal) to shine by friction SVO S = human (Ag) O = surface (Go) e.g. þa ðe beadogriman bywan sceoldon [Beo: 062100 (2255)] “Those who should polish the war helmets.”

The resulting internal structure of this subdomain is shown in Table 6.

Image schemata and light

Table 6.OE verbs of light: causative subdomain To cause to give off light .OE gescı¯nan: to cause something to give off light ..OE gelı¯htan: to cause something to begin to give off light …OE a¯tendan: to cause something (a fire) to begin to give off light ..OE breohtnan: to cause something to be bright …OE a¯lı¯htan: to cause something to be bright by shining light on it ..OE hwı¯tan: to cause something to shine by cleaning it ..OE feormian: to cause something to shine by friction …OE b¯ywan: to cause something (of metal) to shine by friction

6. Paths of semantic change I will now analyse some of the semantic changes that have affected this lexical domain in the later history of the English language. By comparing our Old English definitions with the Middle and Modern English definitions given by the OED, we have noticed that the semantic changes undergone by these predicates have followed very similar diachronic paths, regarding both the stability parameter and the intensity parameter. Regarding the stability parameter, it has been noticed that some verbs expressing stable light in Old English have come to convey unstable light in later English. This is the case with Old English twinclian and blı¯can: (13) OE twinclian: to shine dimly > ME twynkle
: to sparkle dimly e.g. Hise eyen twynkled..As doon the sterres in the frosty nyght [c1386 CHAUCER Prol. 267] (14) OE blı¯can: to shine brightly by reflecting light > ME blike
: to sparkle brightly by reflecting light e.g. His lockkes ant his longe berd blikede al o gold [a1225 St. Marher. 9]

As for the intensity parameter, it has been observed that some Old English predicates expressing bright stable light and bright unstable light have come to express, respectively, soft stable light and soft unstable light in later English. This is the case with Old English glo¯wan and scimerian: (15) OE glo¯wan: to shine with a very bright light and heat > ME glowe
: to shine with a soft and warm light e.g. Glas þat glowed & glyŠt [13.. E. E. Allit. P. A. 114]

73

74

Javier E. Díaz Vera

(16) OE scimerian: to sparkle brightly > ME shimmer
: to sparkle softly e.g. Hit schemered & schon þurŠ þe schyre okez [13.. Gaw. & Gr. Knt. 772]

A second type of semantic change implies the extension from the domain of light into other lexical domains. Following Ritzke-Rutherford (1979:11), I will argue here that the semantic evolution of the Old English vocabulary of light indicates the existence of a tripartite human response to this natural phenomenon: 1. a physiological response, which responds to the pure sense of physical perception of light and warmth; 2. a psychological response, corresponding to the emotional overtones of light; 3. a metaphysical response, corresponding to an interpretation of light and darkness in terms of moral good and evil. As has been seen in the first section, OE scı¯nan could be used in order to refer not only to the perception on the speaker’s side of a physical light, but also to his perception of a quality, emotion or perception. When used with a human participant in subject position, OE scı¯nan “to shine” conveys different positive qualities related to ability, character, achievement and position at the figurative level. In most cases, these positive features are of a religious nature, insofar as light is considered one of the physical qualities of God and other divine beings (metaphysical response). (17) a.

OE scı¯nan (physical light) SV a. S = something that gives off light e.g. Þonne seo sunne on hadrum heofone beorhtost scineð [Bo: 019200 (9.21.1)] “When the sun shines most brightly in the clear sky.” b. OE scı¯nan (abstract light) SV a. S = someone that is brilliant in character e.g. Þonne scinað ða rihtwisan swa swa sunne on hyra fæder rice [Mt(WSCp): 045100 (13.43)] “Then shine the righteous ones as the sun in the kingdom of their father.”

A similar semantic extension has affected OE lı¯htan “to begin to shine”, a verb used with a human participant in the role of subject in order to refer to the acquisition of a new, positive quality. Whereas the earliest instances of this

Image schemata and light

change indicate spiritual illumination, a progressive extension towards intellectual illumination is found after the fifteenth century: (18) a.

OE lı¯htan: to begin to shine > lOE to illumine spiritually e.g. Þet he..mid his halie gast us lithe [c1175 Lamb. Hom. 63] b. ME light
: to illumine spiritually > to illumine intellectually e.g. God…light Šoure resoun, and make cleer Šoure vnderstondynge [1422 tr. Secreta Secret., Priv. Priv. 133]

Verbs of light are also used in order to express positive and negative emotions (psychological response). For instance, the causative verb OE beorhtnian “to cause something to be bright” is metaphorically used in early Middle English texts in order to refer to love, insofar as love is conceived as brightness of the heart. (19) ME brihten: to cause someone to be bright SVO S = emotion (love) O = human (heart) e.g. Luue, þet schireð & brihteð þe heorte [a1225 Ancr. R. 384]

Similarly, verbs used to express very bright light will be used in order to describe facial expressions related to very strong emotions, such as shame and anger: (20) OE beornan: to give off light, heat and gases > to burn with anger SV S = passion e.g. Ðonne beorneð in scortnisse eorre his [PsGlA: 001000 (2.13)] “Then burns shortly with wrath.” (21) OE glo¯wan: to shine with a very bright light and heat > to shine with shame SV S = human (body part) e.g. For shame of him my chekes wexen rede; Algates, they biginnen for to glowe [c1386 CHAUCER Can. Yeom. Prol. & T. 1096]

In order to offer an interpretation of these diachronic tendencies, we must reconsider the fact that the light we normally find in nature is bright and continuous, and is emitted either directly by the sun (our most prototypical source of light), the stars or a fire, or indirectly by a surface that reflects it. This preconceptual schema is used by speakers in order to conceptualise in the same fashion less prototypical examples of emission of light, such as artificial light and, on the metaphorical level, human virtues and emotions that are normally associated with light.

75

76

Javier E. Díaz Vera

As previous studies on semantic change (Johnson 1987; Sweetser 1990) have shown, word meanings change from more concrete to less concrete, so that any deviation from the prototypical model of light described above implies an increase in the degree of abstractness of the predicate. The semantic changes described here respond thus to the well-known tendency of old words to acquire new meanings that are less tangible and more abstract, extending a schema from the physical to the non-physical world, a tendency that applies to both the intradomain and the extradomain level.

References Bendix, E. H. 1966. Componential Analysis of General Vocabulary: The Semantic Structure of a Set of Verbs in English, Hindi and Japanese. The Hague: Mouton. Bosworth, Joseph & T. Northcote Toller, eds. 1898. An Anglo-Saxon Dictionary. Oxford: Oxford University Press. Cameron, Angus et al., eds. 1986–1996. Dictionary of Old English (letters A-E). Toronto: Pontifical Institute of Mediaeval Studies. Díaz Vera, Javier E. 2002. “The Semantic Architecture of the Old English Verbal Lexicon: A Historical-lexicographical Proposal”. A Changing World of Words: Studies in English Historical Lexicography, Lexicology and Semantics ed. by Javier E. Díaz Vera, 47–77. New York & Amsterdam: Rodopi. Dik, S. C. 1989. The Theory of Functional Grammar. Part 1: the structure of the clause. Dordrecht: Foris. Faber, Pamela & Chantal Pérez. 1993. “Image Schemata and Light: A Study in Constrastive Domains in English and Spanish”. Atlantis 15.117–134. ——— & Ricardo Mairal Usón. 1994. “Methodological Underpinning for the Construction of a Functional Lexicological Model”. Miscelánea: A Journal of English and American Studies 15.193–217. ——— & Eulalio Fernández Sánchez. 1996. “The Lexical Field of POSSESSION as a Construction of Conceptual Primitives”. Proceedings of the XIXth International Conference of AEDEAN ed. by J. Pérez Guerra et al., 263–268. Vigo: Universidade de Vigo. ——— & Ricardo Mairal Usón. 1999. Constructing a Lexicon of English Verbs. Berlin: Mouton. Fillmore, C. 1982. Frame Semantics. Seoul: Hanshin. Hall, J. R. Clark, ed. 1960. A Concise Anglo-Saxon Dictionary. 4th edn. Cambridge: Cambridge University Press. Healey, Antonette di Paolo et al., eds. 2000. The Dictionary of Old English Corpus in Electronic Form. Toronto: Pontifical Institute of Mediaeval Studies. Johnson, M. 1987. The Body in the Mind. Chicago & London: University of Chicago Press. Oxford English Dictionary. 1989. 2nd. edn on CD-ROM ed. by John A. Simpson & Edmund S. C. Weiner. Oxford: Oxford University Press.



Image schemata and light

Ritzke-Rutherford, Jean. 1979. Light and Darkness in Anglo-Saxon Thought and Writing. Frankfurt: Lang. Roberts, Jane & Christian Kay with Lynne Grundy. 1995. A Thesaurus of Old English. London: King’s College. Sweetser, Eve. 1990. From Etymology to Pragmatics: The Mind as Body. Metaphor in Semantic Change. Cambridge: Cambridge University Press.

77



Loanword etymologies in the third edition of the OED Some questions of classification Philip Durkin Oxford University Press, Academic Division

The OED was first published in fascicles between 1884 and 1928, with a onevolume supplement being added in 1933 and a further four-volume supplement, chiefly containing new words and senses from the nineteenth and twentieth centuries, between 1972 and 1986. All of these parts were brought together in the second edition of 1989, which has additionally been available online since March 2000. It constitutes the most comprehensive dictionary of the English language yet published, and is the only historical dictionary to attempt to cover virtually the full historical span of the language, omitting only words obsolete before 1150 and vanishingly rare items attested later. In particular, its etymologies are extremely full and detailed, providing, together with its daughter volume the Oxford Dictionary of English Etymology, the fullest etymologies available for the vast majority of English words. Unsurprisingly, data provided by the OED, and particularly its dated quotations and detailed etymologies, has frequently been used as the basis of attempts to estimate the numbers of loanwords of various origins entering the English language in different historical periods. Unfortunately, however, this has often been done using tools like the Chronological English Dictionary which hugely simplify OED’s data, or by using fairly arbitrary simplifications of OED’s data imposed by researchers keen to limit the size of their task. (See further Durkin (2002a) for discussion of this question.) For some years a third edition of the OED has been in preparation, in the course of which the whole of the existing text of the dictionary is being reassessed and where necessary revised for the first time. The results of this work

80

Philip Durkin

have been appearing online since early in 2000, now covering the alphabetical range from M, where revision work began, to monnisher; editorial work itself, which is necessarily somewhat ahead of publication, is rapidly approaching the end of the letter M, a significant milestone in the progress of the new edition of the dictionary. The new edition is seeing extensive changes to the dictionary text in various areas: new data is being accommodated in all parts of the text, for instance antedatings and postdatings of existing words and senses (with something like one in three of all words and senses being antedated); additionally, editorial conventions are being refined, in response to the new possibilities provided by online preparation and publication of the text, and in response to the new possibilities presented by the data available to today’s lexicographers. Etymologies are no exception to this. The etymologist’s toolkit has expanded greatly in recent decades. Foreign-language dictionaries have seen tremendous growth, particularly historical dictionaries of major European languages, just as have dictionaries devoted to particular periods or to world varieties of English. Additionally, major corpora are now coming to be available for a number of languages, which can be invaluable in filling gaps in other dictionaries’ coverage, for example in providing data on generic uses of proper names or phrasal constructions. (For further illustration see Durkin (1999), Durkin et al. (2002).) Using these tools, today’s OED editors are able to provide much fuller etymological information, in particular paying attention to the dating of particular forms and senses in many of the most significant donor languages, making it possible to supplement existing etymologies and in many cases to progress somewhat beyond the more traditional tests of morphological identity and ultimate etymological descent. In the process, some of the first edition’s conventions have been revised. In particular, the first edition’s fairly rigid division of loans into adoptions (abbreviated as “a.”), adaptations (abbreviated as “ad.”), and words “formed on or from native or foreign elements” (abbreviated as “f.”) has not been retained in the new edition, having been found unworkable in light of the fuller documentation now available on the immediate circumstances of loans. (See further Durkin 1999: 7–8.) Instead, a more flexible set of styles has been developed in which the available data can be presented more fully, and it is these that will be outlined here, with particular attention being given to those cases where the boundaries between loanword categories are somewhat opaque or uncertain, in order to test how useful such categories are.

Loanword etymologies in the third edition of the OED

To begin with perhaps the simplest category, a loan from an “exotic” language, mikoshi n., appears at first sight to be a very simple loan directly from Japanese, and is presented as such in OED2. However, if one examines the earliest quotation more closely, it becomes evident that not only does the lemma appear in a rather distinctive spelling, but the quotation itself is from a translation of a German text (which was in fact only printed considerably later, but was translated from manuscript by Scheuchzer): OED3, mikoshi n., etymology and first sense: [< Japanese mikoshi (1603 as micoxi in Vocabulario da Lingoa de Iapam) < mihonorific prefix (cf. mikado n.) + koshi palanquin, litter (carried by pole on the shoulders of two or more people). In quot. 1727 via German Mikosi (1779 in printed version of text translated from manuscript in quot.).] I. Simple uses. 1. In Japan: a portable Shinto or Buddhist shrine which is carried through the local community on ceremonial occasions to symbolize the presence of the deity. 1727 J. G. SCHEUCHZER tr. E. Kæmpfer Hist. Japan I. IV. iv. 296 The Mikosi.. eightangular shrines, almost too big to be carried by one man. 1877 W. SIMPSON in Trans. Soc. Biblical Archaeol. 5 550 (heading) They are also called Mikoshi; mi, is ‘precious’ or ‘honourable’, Koshi, is ‘seat’. 1911 Japan Mag. July 152/1 The Kami, or god, of the temple are paraded through the streets, under a mikoshi, or sacred canopy highly decorated. 1952 Monumenta Nipponica 8 27 The book is profusely.. illustrated… A view of Lord Elgin’s entry into Edo in a norimono exhibits him seated in a mikoshi with a phoenix on the top. 1989 Japan Times 19 May 5/7 Many hands make light work of carrying a mikoshi (portable shrine). 2000 Daily Yomiuri 5 Dec. 11/4 A favorite Japanese nickname [of Thunberg’s cranesbill], mikoshigusa, refers to the appearance of the spent capsules, which remind people of the fancy gabled roof of the portable mikoshi shrines used during Shinto festivals.

Here one may perhaps decide to regard the 1727 use as something of an early false start, and regard the word only from its second example, of 1877, as a direct Japanese loan, although if one does so one cannot therefore use 1727 as the date of this as it were ‘pure’, direct Japanese loanword. The dictionary reader who is keen to use the OED as a database of loans from a particular source must therefore be prepared to examine even the apparently simplest cases quite carefully; the dictionary entry is intended to be taken as a whole, and if it is not the reader may be in danger of not extracting exactly the desired data. The point at issue at mikoshi is essentially a greater sensitivity to the word’s immediate etymology, the mode of its direct transmission into English, and it

81

82

Philip Durkin

is one of the aims of the etymologies in OED3 to bring this aspect of the etymology into sharper focus. This does not apply only to the famous cases, such as ginger in A. S. C. Ross’s classic Philological Society study (Ross 1952), but also to a great many apparently much simpler words. A subtle shift of focus in this regard may be seen by comparing OED2’s etymology for maritime a. with OED3’s: OED2, maritime a., variant forms and etymology: Forms: α. 6 myrytayne, maritayne, 7 maritan(e, maritin, marratine, maretine, 7–8 maritine; β. 7 marittime, 7–8 maritim, 6– maritime. [ad. L. maritim-us, f. mari-, mare sea + suffix -timus (occurring in fı¯nitimus neighbouring, le¯gitimus lawful; also forming superlatives as intimus inmost, ultimus last). The β forms are from the L. directly or through mod.F. maritime (cf. Sp. marítimo, Pg. maritimo, It. marittimo). An OF. form, maritim, was corrupted, partly owing to confusion of suffixes, into maritin, maritaim, maritain (latinized marita¯nus), whence the α forms above.] OED3, maritime a., variant forms and first part of etymology: Forms: α. 15 maritayne, myrytayne, 16 maretine, maritan, maritane, maritin, marratine, 16–17 maritine. β. 15– maritime, 16 marittim, marittime, 16–17 maritim; Sc. pre-17 maretym. [< Middle French, French maritime (1336) or its etymon classical Latin maritimus relating to the sea, situated near the sea, nautical < mari-, mare sea (see mare n.4) + -timus, suffix (occurring in fı¯nitimus neighbouring, le¯gitimus legitime a., and forming superlatives such as intimus intime a., ultimus ultime a.); in α forms via Middle French maritain (c1470), maritin (c1500) marine, maritime, alterations (with substitution of suffix: cf. -ain -an, -in -ine1) of Middle French maritime. Cf. Italian marittimo (a1484), Spanish marítimo (1493), Portuguese marítimo (16th cent.).]

Here the basic analysis has not changed: some of the forms (by a small margin the earliest, and hence marked alpha) are via French forms which show suffix substitution; the other forms, including the modern output form, could be from either Latin or French, with neither the form of the word, nor its date in French, nor its meaning in either language helping to resolve this question. However, in the OED3 entry one may observe two major differences in the presentation of this material. Firstly, the reader is given more of the supporting documentation, present-day dictionaries of French making it possible to date the various French forms, and to ascertain that a single series of glosses apply equally to the French and Latin words. Secondly, and perhaps more fundamentally, the uncertain immediate etymology is now given the first position in the

Loanword etymologies in the third edition of the OED

revised etymology. OED2, by contrast, chose to highlight first the deeper etymology, stating at the outset that the word was, fundamentally, an adaptation of Latin, and only mentioning in the following note that the alpha forms are via French and the beta forms could equally well be from either Latin or French. For OED3 it is felt that with generally much richer documentation for constructing etymologies comes a responsibility to present as clearly as possible what can be ascertained of the word’s immediate transmission into English — although it should be noted that it is not always possible to compress all facets of this into a simple opening formula, and it is assumed that all of the information presented in the etymology will be taken into consideration by the reader. In other cases, the evidence will be such as to suggest a truly “mixed” transmission, with an English word showing the result of parallel borrowing of forms from more than one language. Such etymologies are particularly common for Middle English and Early Modern English, where the early evidence often points particularly strongly to parallel borrowing of both a French word and its Latin etymon. (See further Durkin (2002b) for full investigation of this phenomenon.) However, such mixed transmissions are hardly limited to the learned discourse of the late medieval or early modern periods. maidan n. provides an example from a quite different geographical and historical setting. In OED2 this was given as a loan from Persian, the etymology saying simply “< Persian maida¯n” and no more, but close inspection of the quotations reveals a mixture of milieux for this loan, some examples clearly showing a Persian locale, while others are from South Asia, suggesting parallel borrowing of Urdu maida¯n (itself in turn borrowed from Arabic mayda¯n, just like the Persian word). Furthermore, the earliest example is again via another European language, here Italian, although unlike the mikoshi example this is not accompanied by any peculiarity of form. Looking at the dictionary record for other European languages yields several other early parallels which could perhaps have had a hand in the word’s transmission; French almost certainly has done so through the 1662 and 1687 translated sources: OED3, maidan n., etymology, definition, and quotations: [Partly < Persian maida¯n and partly < Urdu maida¯n, both < Arabic mayda¯n. In quot. c1551 via Italian maidan (late 15th cent.). Cf. also post-classical Latin maydanum (14th cent.), Portuguese maidão (1608), French meydan (1653; later maidan (1721)).] Esp. in South Asia: a market place, park, or other open space in or near a town; an esplanade or parade ground.

83

84

Philip Durkin

?1551 W. THOMAS tr. G. Barbaro & A. Contarini Trav. Tana & Persia (1873) II. 53 Sending me word wthall that I shulde go to Maidan, that is, to witt, to the markett place. 1625 W. FINCH in S. Purchas Pilgrimes I. IV. 423 The Medon, which is a pleasant greene, in the middest whereof is a May-pole to hang a light on. 1662 J. DAVIES tr. A. Olearius Voy. Ambassadors V. (1669) 172 The Meydan, that is the great Market-place. 1662 J. DAVIES tr. A. Olearius Voy. Ambassadors V. (1669) 178 The Market-place, or Maydan, is large and noble. 1687 A. LOVELL tr. J. de Thevenot Trav. II. 79 There are many squares in Ispahan, but..the Meidan..is the greatest and finest place in the world. 1698 J. FRYER New Acct. E.-India & Persia 249 The Midan, or open space before the Caun’s Palace. 1743 R. POCOCKE Descr. East I. IV. v. 184 Public places out of town, commonly call’d Meidans, or Places, where they have..summer-houses. 1827 J. S. BUCKINGHAM Trav. Mesopotamia I. xi. 375 A Maidan, or open space, where the horses are kept in the air. 1844 J. H. STOCQUELER Hand-bk. India 480 Dum Dum..is a spacious cantonment, with an extensive maidaun, or esplanade. 1879 A. FORBES Camps, Quarters, & Casual Places (1896) 283 Before me on the maidan is the plain monument to Sir Mountstuart Jackson. 1882 B. M. CROKER Proper Pride II. i. 14 He.. was galloping away over the moonlit midan. 1924 E. M. FORSTER Passage to India i. 8 There is an oval Maidan, and a long sallow hospital. 1964 R. PERRY World of Tiger i. 5 After the savannahs have been burnt..the resulting park-like grassy maidans become the favourite grazings of.. deer. 1976 Listener 17 June 772/2 When freedom came to Ghana.. the excited crowd carried the whole cabinet of the new leaders out into the darkness of the maiden. 1989 G. MEHTA Raj xxxiii. 236 The younger girl plays polo.. in the public maidan.

Examples such as this, very much the norm among loanwords with any significant time-depth in OED3, raise many questions about how far a clear immediate etymology can be identified. Instead, one may suspect the interplay of a number of different models. Perhaps there may be an early borrowing from a single source which becomes well established, subsequently showing semantic or formal influence from related words in other languages, but ultimately showing only a single “point of entry” into the language. Or perhaps there may be a series of separate points of contact, with maybe an initial isolated borrowing via an intermediary language which gains no firm footing, followed by a second, more resilient loan direct from a more “exotic” language as contacts become more fully established. Or perhaps there may be a series of ad hoc borrowings, sometimes with one language as the immediate stimulus and sometimes with another, giving rise to a certain currency in written or spoken use which leads to the word becoming firmly established in English without there being any single identifiable “parent” or any single identifiable instance of “borrowing”.

Loanword etymologies in the third edition of the OED

Sometimes the evidence may lead the lexicographer to incline more to one of these explanations than to another, but in a great many cases this is simply unknowable, and one should not expect always to be able to identify a single, indisputable account of a word’s immediate origin. In some cases, such as mikoshi earlier, one may be tempted to split the material, and identify two separate words, an early loan via German which is represented by only a single use, and a subsequent loan directly from Japanese. However, practical lexicographical experience will reveal how dangerous this road can be. Dictionaries, especially historical dictionaries, depend for their coherence on a fairly “baggy” concept of what constitutes a word, at least so far as establishing a dictionary entry is concerned. The very full documentation provided by a large historical dictionary such as the OED can bring this “bagginess” all too clearly to light, but the temptation as a consequence to fragment the material on etymological grounds is one that is to be resisted in all but the most clear-cut and well documented instances: on a purely pragmatic level, further evidence that leads to a revision of one’s assumptions has an unfortunate habit of turning up just when a firm decision has been made to split. As a practical demonstration one may take a further Japanese loanword, mikan n., for which OED2’s earliest quotation is dated 1947, but which it has been possible to antedate several times during the OED revision process, first to 1891, then to a 1727 Scheuchzer example (which does not this time show any morphological oddity, but does appear to show some semantic confusion, explained in the etymology note). However, in this instance it has been possible to antedate still further to 1618 in Cocks’s Diary
: OED3, mikan n., etymology, definition, and earliest quotations: [< Japanese mikan (1603 as miccan in Vocabulario da Lingoa de Iapam) < mi(in mitsu honey < Middle Chinese, perh. ult. < the Indo-European base of mead n.1) + -kan citrus (< Middle Chinese). Cf. Chinese mìgn mandarin, tangerine. In quot. 1727 via German, where the 1777 printed version of the text (translated in the quot. from manuscript) in fact reads ‘Limonen und Pomeranzen wachsen hier häufig, und von verschiedener Art. Die edelste Art nent man Mican’.] The fruit of any of several species of citrus (family Rutaceae) originally cultivated in Japan; esp. a variety of mandarin orange, a satsuma. Cf. satsuma n. 2. 1618 R. COCKS Diary 12 Feb. (1883) II. 14 Pasquall Benita came from Langasaque to Firando and brought me a present of coiebos, micanas, and

85

86

Philip Durkin

peares. 1727 J. G. SCHEUCHZER tr. E. Kæmpfer Hist. Japan I. I. ix. 117 Oranges and Lemons grow very plentifully, and of different sorts. That sort of Lemons, which is reckon’d the best, is call’d Mican. 1891 Scribner’s Mag. 9 165/2 The Musmee wears a wondrous dress Kimono, obi, imogi..! Her girdle holds her silver pipe, And heavy swing her long silk sleeves With cakes, loveletters, mikans ripe, Small change, musk-box, and writing-leaves. 1947 J. BERTRAM Shadow of War 193 Mikans the sweet, juicy mandarin oranges.

This 1618 example is rather interesting, as it seems to show accommodation to the morphology of a Romance language, although no foreign-language intermediary exists for this source, and the 1603 Portuguese-Japanese glossary (João Rodrigues Vocabulario da lingoa de Iapam (1603)) in fact has miccan, with no final vowel, while Cocks elsewhere has a habit of dropping final vowels from Japanese words, for instance cattan for Japanese katana, a type of Samurai sword, so this may be no more than a type of linguistic eccentricity in this particular source, giving no reason to suspect that this is not in fact a direct loan from Japanese.1 Similar difficult questions are found when one turns one’s attention to indirect loans. The approach to naturalized words, where all or part of an etymon is remodelled to accommodate established English derivational morphemes, will typically differ according to the date at which words enter English. For the Middle English or Early Modern periods the process of naturalization will typically be understood rather than expressed explicitly, an example being mammillary a., where the English word (which is attested from the early seventeenth century) is given as a loan either from Middle French or from post-classical Latin, with a cross reference to the entry for the suffix -ary being supplied to explain the standard method of accommodation of words of this type: OED3, mammillary a., etymology only: [< Middle French mamillaire (early 16th cent.) or its etymon post-classical Latin mamillaris mammillar a.: see -ary2. Some of the collocations at sense 1 are prob. after collocations in post-classical Latin, e.g. corpus mamillare, processus mamillaris. Cf. earlier mammillar a.]

In the modern (or late modern) period, where the form and spelling of such word-forming elements is essentially fixed, and the phenomenon is well

1.We are very grateful to Mr. Nicholas Warren, the OED’s consultant on Japanese material, for supplying both the example from Cocks and the expert analysis of Cocks’s linguistic habits.

Loanword etymologies in the third edition of the OED

established of parallel naturalized loans alongside unnaturalized ones (such as for example mastersinger n. and meistersinger n.), this style will normally be avoided, and the morphological remodelling instead spelt out as a formation from word-forming elements by now well established in English, after a foreign model, as in metalloscopy n., where the English word (for which the first quotation is dated 1898) is presented as formed from the established English word-forming elements metallo- and -scopy on the model of the French word métalloscopie
: OED3, metalloscopy n., etymology only: [< metallo- + -scopy, after French métalloscopie (1877 or earlier).]

The entries for combining forms themselves, such as macro- below, give the supporting analysis, with early borrowings such as macrology n. being treated as straightforward loans with consequent morphological remodelling of final -ia to -y (neither macro- nor -logy being yet taken to give new formations within English), whereas later formations such as macrosomite n. are presented as formed from existing English elements after a German model, in order to best explain the (unvarying) accommodation to modern English spelling and morphology, with the preponderance by this date of entirely new formations from the same elements within English lending the crucial support for this analysis: OED3, macro-, first part of etymology: [< ancient Greek µακρο-, combining form of µακρς, large (see meagre a.), found in a large number of ancient and later Greek formations and in classical and post-classical Latin loans from these. Found in loans (as macrocosm n., macrology n.) from early 17th cent.; English formations are found from 1820s. Cf. French macro- (formations in which are found from beginning of 19th cent.), German makro- (formations in which are found from at least end of 18th cent.: see macrobiotic a.). Freq. forming words used contrastively with corresponding formations in micro- (see further spec. senses below)…] OED3, macrology n., first part of etymology: [< classical Latin macrologia < ancient Greek µακρολογ α < µακρς macro- + -λογ α (see -logy)…] OED3, macrosomite n., etymology: [< macro- + somite n., after German Makrosomit (V. Graber 1888, in Morphol. Jahrb. 14 350). Cf. microsomite n.]

However, as these examples show, the dividing line is hardly clear-cut, and I would suggest that it would be unwise to attempt to see a strong boundary

87

88

Philip Durkin

between the two types of formation, direct loans on the one hand as against English formations after a foreign-language model on the other. Rather, I would suggest that the differing modes of presentation should be seen as pointers to the underlying rather more complex process of change in accommodation of new technical vocabulary within English. The stylistic presentation of macrosomite or metalloscopy is identical to that of my final category, the calque, the difference being that in the case of the calque there may be no etymological relationship between the foreign-language word and the English word formed on its pattern, as for instance magic bullet n., where the English compound renders the German compound Zauberkugel without using elements which have any etymological relationship with those of the German word: OED3, magic bullet n., etymology: [< magic a. + bullet n.1, after German Zauberkugel, first used in this sense by Paul Ehrlich (1854–1915; cf. quot. 1907), German medical scientist, in an address given at the opening of the Georg Speyerhaus in Frankfurt, Germany, in Sept. 1906, but not published until much later (see Coll. Papers P. Ehrlich (1960) III. 42–52). German folklore includes a number of stories in which magic bullets of supernatural accuracy play a prominent role. The best-known is that which forms the basis for the opera Der Freischütz by Carl Maria von Weber (see weberian a.2), and which is alluded to in examples such as the following: 1936 G. B. SHAW Millionairess Pref. 127 War is like the seven magic bullets which the devil has ready to sell for a human soul. Six of them may hit the glorymonger’s mark very triumphantly; but the seventh plays some unexpected and unintended trick that upsets the gunman’s apple cart.]

Instances of calqueing may obviously also be found among the later senses of a word. For instance, at majority n. the earliest two senses are identified clearly as reflecting the senses of the parent Middle French and Latin words, while sense 3c is interestingly calqued on literary use of an unrelated Latin word, plures (itself probably calqued on ancient Greek hoi pleones in this specific use): OED3, majority n., extract from etymology and sense 3c: […In sense 3c after classical Latin plu¯res (see plural a.) in this sense (in Plautus and Petronius, prob. after ancient Greek ο πλ ονες)…] †c. the majority
: the dead. Chiefly in phrases to join the majority and to go (also pass over) to the majority
: to die. Obs. 1719 E. YOUNG Revenge IV. i, Life is the desart, life the solitude; Death joins

Loanword etymologies in the third edition of the OED

us to the great majority. 1764 London Mag. Nov. 581 ‘Oh, no, it is all over with me; I am going, as fast as possible, to join the majority.’ — Ad plures. 1837 T. CARLYLE Mirabeau in Crit. & Misc. Ess. (1840) V. 200 This Mirabeau’s work then is done… He has gone over to the majority: Abiit ad plures. 1883 Longman’s Mag. Dec. 179 He had passed over to the majority..we should see his face no more. 1889 T. A. TROLLOPE What I Remember III. 61 He, too, alas! has joined the majority.

The picture that emerges from this brief survey is thus one of a broad spectrum of immediate foreign-language influence on the development of English words, within which broad categories, such as direct loans, mixed etymologies, naturalized words, or calques may be identified. However, as the examples selected have shown, the boundaries between these categories are not hard and fast, and for the lexicographer the choice of one style of presentation over another will often be a finely balanced question where the available data has to be weighed carefully, and where it is impossible to ensure rigid consistency in the demarcation of one category from another. Such considerations are of particular importance in planning the future electronic tagging of the dictionary database. Where the boundaries between categories are not clear-cut, it is important to resist the temptation to encode into the electronic text fine distinctions which may give the user of the dictionary a false impression of simplicity in the underlying data. Rather, it is planned at present to encode into the text only a single identifier which would embrace all of the types of immediate influence discussed here, whether introduced by “<” or “after”, and whether occurring at the level of a headword, or of a subsidiary sense or subsidiary lemma. Such tagging is, however, a good way off, and it would be as well to conclude with a brief summary of how the data can best be used by readers today. Historical dictionaries remain the best available tool for anyone wanting to investigate how foreign-language influences have shaped the lexis of a language. As present work on OED3 shows, it is possible for a historical dictionary today to draw upon greatly enriched supporting data in this area of its work, to give a much more nuanced account of individual word histories. However, what historical dictionaries cannot do is to make an inherently complex story less complex, or to reduce the lexical history of a language to a series of lists of accessions from assorted languages with dates, especially if we wish to progress past the ultimately fairly meaningless compilation of statistics on what proportions of the vocabulary of English are ultimately Germanic, Latin, Greek, Slavonic, Indic, etc. Making use of the dictionary data requires a certain amount of patience and

89



90

Philip Durkin

careful consideration, and also a willingness to accept that the answers may not be clear-cut and simple, and that in a great many cases there will be some degree of genuine uncertainty as to the mode of a word’s immediate transmission. If approached from such a perspective, it is to be hoped that the fuller data presented in OED3 will provide the essential basis for an improved understanding of the history of lexical and semantic borrowing in English.

References Culpepper, Jonathan & Phoebe Clapham. 1996. “The Borrowing of Classical and Romance Words into English: A Study Based on the Electronic Oxford English Dictionary”. International Journal of Corpus Linguistics 1.199–218. Durkin, Philip. 1999. “Root and Branch: Revising the Etymological Component of the OED”. Transactions of the Philological Society 97. 1–50. ———. 2002a: “Changing documentation in the Third Edition of the Oxford English Dictionary
: Sixteenth-century vocabulary as a test case”. Sounds, Words, Texts and Change. Selected Papers from 11ICEHL, Santiago de Compostela, 7–11 September 2000 ed. by Teresa Fanego, B. Méndez-Naya & E. Seoane, 65–81. Amsterdam: John Benjamins. ———. 2002b: “‘Mixed’ etymologies of Middle English items in OED3: some questions of methodology and policy”. Dictionaries: the Journal of the Dictionary Society of North America 23. 142–155. ———, Jane McCauley, Tania Styles, Samantha Schad, Neil Fulton, Anthony Esposito & Joanna Tulloch. 2002: “From the etymology workshop: work in progress on the Oxford English Dictionary”. Oxford University Working Papers in Linguistics, Philology and Phonetics 7. 225–242. Finkenstaedt, T., E. Leisi & D. Wolff. 1970. A Chronological English Dictionary. Heidelberg: Winter. The Oxford Dictionary of English Etymology. 1966, ed. by Charles T. Onions, with the assistance of G. W. S. Friedrichsen and Robert W. Burchfield. Oxford: Oxford University Press. The Oxford English Dictionary. 1884–1933, ed. by Sir James A. H. Murray, Henry Bradley, Sir William A. Craigie & Charles T. Onions; Supplement, 1972–1986, ed. by Robert W. Burchfield; 2nd. edn, 1989, ed. by John A. Simpson & Edmund S. C. Weiner; Additions Series, 1993–1997, ed. by John A. Simpson, Edmund S. C. Weiner & Michael Proffitt; 3rd edn (in progress) OED Online, March 2000–, ed. by John A. Simpson. Oxford: Oxford University Press. Ross, A. S. C. 1952. Ginger: A Loan-word Study. Oxford: Blackwell. Wermser, Richard. 1976. Statistische Studien zur Entwicklung des englischen Wortschatzes. Bern: Francke.



“Non olet” Euphemisms we live by Andreas Fischer University of Zürich

1.

Introduction

The lexicon is a challenge for linguists. Like other levels of linguistic organisation it is held to be a system “où tout se tient”, but its staggering size and complexity defy attempts to understand and model it in its entirety. Linguists dealing with the lexicon thus tend to choose one of two options. They either work ‘top down’, i.e. they focus on general processes (regularities of semantic change, for example) and use words only selectively for purposes of illustration. Alternatively, in the ‘bottom up’ procedure, they study all words within one semantic domain and attempt to account for them from the point of view of a specific question. This paper follows the second model: it restricts itself to the vocabulary for ‘toilet’ in the history of English and investigates it in view of a problem relevant for the lexicon in general, namely why it is so much larger than appears to be necessary from a purely functional point of view. Technical innovations may make certain new terms necessary (an obvious point), but my main interest here is in multiple categorisation, i.e. in the fact that one and the same ‘thing’ may be — and very often is — characterised from different points of view. Multiple categorisation is a general feature of the lexicon, but it is especially striking in the case of ‘toilet’ because of the potentially ‘offensive’ character of the topic and the need to mitigate it through euphemisms.1

1.Readers likely to be offended by the topic should heed the moral of the anecdote which has given this paper its title (Bartels 1992:116–117). According to Suetonius’ Lives of the Caesars, the emperor Vespasian was well known for his imaginative ways of raising money. When his son Titus blamed him for having introduced a ‘urine tax’ (probably a tax imposed on fullers

92

Andreas Fischer

The paper contains five sections: a presentation of the Historical Thesaurus of English (HTE) as the source of the material I have used (2) is followed by a brief history of the toilet (3), the notional categorisation of the ‘toilet’ words attested throughout the history of English constitutes the central part (4), and the paper concludes with a discussion of the categories thus established from the point of view of metonymy (5) and a look at the role played by taboo, euphemisms and dysphemisms (6).

2. Words for toilet and the Historical Thesaurus of English The material on which this paper is based is the vocabulary for ‘toilet’ in the history of English, as collected by and for the Historical Thesaurus of English (HTE).2 When presented as a poster at the 5th ESSE Conference in Helsinki in 2000, this vocabulary comprised 157 (124 + 33, see below) words, to which I have added a small number which are attested in the OED, but had not, then, been included in the HTE. The OED and the Thesaurus of Old English do not, of course, record all words for ‘toilet’ ever used by speakers of English; in addition to words used but lost forever there exist items of restricted currency that are only listed in specialised collections (glossaries, dictionaries, thesauruses) of (historical) slang or of regional and national varieties of English. In the rest of this section I will summarize the HTE’s classification of the material and look at its chronological distribution. However, it is not my intention here to discuss the HTE’s — preliminary — ordering of the words in question, and in the rest of this paper I will use them for a classification of my own.3 The HTE’s classification of the lexical field ‘toilet’ is as follows. The main heading ‘privy/latrine’ is divided into six sub-sections labelled ‘the king’s’, ‘for

and tanners for using urine from public latrines), he produced a handful of coins and asked whether the smell of money offended him. Titus replied that it did not and Vespasian then said “Atquin, e lotio est” (‘And yet, it comes from urine’). In popular lore, Titus’s negative answer was given the form “Non olet” (‘It does not smell’). 2.When completed, the HTE will present the vocabulary of English from Old English to Modern English in thesaurus form. It is based on the Oxford English Dictionary (OED, Simpson & Weiner 1989), with supplementary material from the Thesaurus of Old English (Roberts & Kay 1995); for details see Coleman & Kay (2000: 229–239) and Kay, Wotherspoon & Sylvester (2001). 3.I should like to thank Christian Kay for allowing me to use the HTE material for this paper. Any mistakes or misinterpretations, of course, are mine. Note that the numbers given in Section 2. refer to the HTE material as presented in Helsinki in 2000.

“Non olet”

two persons’, ‘spec. outside’, ‘water-closet/lavatory’, ‘chamber-pot etc.’ and ‘close-stool/commode’. The section ‘water-closet/lavatory’ has two subsections: one entitled ‘public’, which is further subdivided into ‘men’s’ and ‘women’s’, and one called ‘W. C. appliances’. The figure below provides an overview of this classification, together with the number of words listed under each heading: privy/latrine (73)

the king’s (1) for two persons (1) spec. outside (4) water-closet/lavatory (27)

public (8)

men’s (4) women’s (6)

W. C. appliances (4) chamber-pot etc. (18) close-stool/commode (11)

For my own study I have defined the object in question as ‘a place, room or booth for defecation and urination’, thus excluding the more specialised categories ‘W. C. appliances’, ‘chamber-pot etc.’ and ‘close-stool/commode’ with altogether 33 words. When one groups the remaining 124 words according to the date of their first attestation (in centuries), the following picture emerges: OE, to 12th c. 13th c. 14th c. 15th c. 16th c. 17th c. 18th c. 19th c. 20th c.

13 3 2 6 17 13 12 15 43

total

124

There is, clearly, an overall tendency for more and more new words to become available: in the 13th to l5th centuries we find between 2 and 6 per century, in the 16th to 19th between 12 and 17, while in the 20th there are 43. These figures do not take into account obsolescence and word loss (which is documented in the HTE), but it is clear that the overall number of words available has increased substantially: according to the OED (which is not a perfect guide in these matters, however), no fewer than 61, i.e. about half, of the 124 words listed in the HTE are still current.

93

94

Andreas Fischer

3. A brief factual history The purpose of this brief section is simply to provide the factual knowledge necessary to understand the words that will be treated in the main part of this paper. Comprehensive information on (the history of) hygiene, sanitation and the toilet in particular is to be found in the works this section is based on: Guerrand (1985), Illi (1987), Palmer (1973) and Wright (1960). The toilet of the Middle Ages, if there was such a facility at all, was an earth toilet, consisting of a ditch or a hole in the ground covered by a wooden board with a hole or holes in it, which enabled people to sit rather than to crouch or squat. If not in the open, such toilets were either separate little buildings (i.e. outhouses) or they were attached to the outside wall of a building, both solutions guaranteeing a certain amount of privacy, protection and comfort. More privacy and even greater comfort was made possible by the (portable) chamber-pot in one’s own bedroom or by a so-called close-stool or commode, a sort of chair with a built-in chamber-pot. (As mentioned above I have disregarded the words for these portable items, which the HTE has listed in separate sections.) Earth toilets of any kind were smelly and potentially unhygienic, a problem for which water provided an ideal solution. Latrinae, communal toilets where excrement was flushed away by running water continually or at regular intervals, were a common feature of the Roman world. In the Middle Ages, however, they seem to have been confined to monasteries and it was only in the late 16th and then again in the 18th century that the flush-toilet was (re-)invented. The first known inventor was Sir John Harington (1561–1612), Elizabeth I’s godson, who designed what was, in fact, a water-closet, although he himself did not use that word.4 The device did not catch on and it took another two hundred years before the water-closet was re-invented (and patented) by a certain Alexander Cummings in 1775.5 The early water-closets were similar to traditional railway toilets, a valve releasing a certain amount of water after the act. They were improved in the 19th century, the most effective and popular improvement being the flush-out closet, the “forerunner of the modern WC” (Palmer 1973: 39), where a certain amount of water was left in the bowl after the flush,

4.Harington published his invention in a treatise entitled A New Discourse of a Stale Subject Called the Metamorphosis of Ajax (1596). The text is accessible in Donno’s (1962) edition; see also Palmer (1973: 26–31) and Wright (1960: 71–75). 5.See Illi (1987: 202) and Palmer (1973: 22).

“Non olet”

thus effectively sealing the toilet from the drains and their smells. Once water came to be a feature in toilets, they were linked to or at least mentally connected with wash- or bathrooms and generally with care of the body and hygiene. Also to be taken into account is the development of public lavatories. The quotation from the OED below indicates that the idea goes back to the 18th century, but in London public lavatories were actually only introduced in the middle of the 19th century. 1727 (title) The Grand Mystery.. proposals for erecting 500 Publick Offices of Ease in London and Westminster. (OED s.v. office, first quotation under sense 9.b.)

Public lavatories were separated according to the sexes, and so, from the late 19th and 20th century onwards, one also finds words for separate women’s and men’s rooms. It is clear from this overview (and from the arrangement of the material in the HTE) that changes in living and developments in sanitation account for a substantial number of new words. In addition to the 73 general words for ‘privy/ latrine’ and to the 4 for special words for an ‘outside toilet’, for example, the HTE lists 27 for ‘water-closet/lavatory’, 18 words for ‘chamber-pot etc.’ and 11 for ‘close-stool/commode’. There was and is, then, a need for new words because of developments in sanitation. However, the attested lexicon is far bigger than what would be needed to accommodate the innovations just described. In the next section I will make an attempt to account for this abundance of terms by looking at how they can be categorised, notionally or cognitively.

4. Notional categorisation As the words making up the list of the HTE show, the notion ‘toilet’ may be conceptualized very differently. My own data-driven attempt to classify as many of the attested words as possible has resulted in no fewer than 12 categories, which I will present in the following.6 I have categorized most, but not absolutely all of the 124 words listed in the HTE and have added a few additional

6.The most detailed classification before the one presented here was made by Allan & Burridge (1991: 78–81). They write that “[t]he lexicon of urination and defecation divides into three categories: terms for the place where it is done, vocabulary for the acts performed, and names for the effluvia”. They have four categories for the “terms for the place where it is done”: “room and receptacle”, “a place where there is privacy”, “a place of easement” and “a place where there is water”.

95

96

Andreas Fischer

ones.7 I have disregarded, here and throughout the paper, any dialect or status labels given in the OED and the HTE. Note, finally, that I have classified certain words, especially compounds, more than once because their elements belong to different notional categories; in such cases the salient element is underlined.8 4.1 ‘Going to a place’ The dominant concept in the Anglo-Saxon period appears to have been that of ‘going to a certain place’, as evidenced by the simplexes gang/gong, genge and a series of compounds with gang as first or second element: earsgang, forþgang, gangern, gangpytt, gangtun, utgang. Gang/gong lived on unti1 1576, but the concept ‘going to a certain place’ seems otherwise not to have survived the Anglo-Saxon period.9 4.2 ‘The place itself, location, container’10 In one of the most frequent conceptualisations the place to which one goes to do one’s business is simply called the place, or more concretely, the house or something similar. The words may be simplexes (vault 1617–a1700, office 1727/ 1871–,11 closet 1869–, chalet 1882–, place 1901–, cottage 1909–, karzy 1961–),12 but more often they are compounds or complex noun phrases

7.A few of the recorded words for ‘toilet’ are completely mysterious as to their etymology and cannot, consequently, be placed in any kind of notional category. Examples are kocay c1440 (attested once, in the Promptorium Parvulorum as a gloss of Latin cloaca, together with privy) and dunny a1790– (“origin unknown”, “[f]req. in Austral. and N. Z. slang”). 8.Within each category examples will be listed in chronological order according to their first attestation. The dates come from the OED, but are here quoted from the HTE. 9.Dieter Kastovsky (oral communication) has suggested that Old English gang etc. may be resultative, in which case it would mean ‘the place one goes to’ rather than ‘going to a place’. In this case category 4.1 would merge with 4.2. 10.This is Allan & Burridge’s “room and receptacle”. 11.OED, sense 9.b, usually in the singular. This is a variant of sense 9.a. ‘pl. The parts of a house, or buildings attached to a house, specially devoted to household work or service; the kitchen and rooms connected with it, as pantry, scullery, cellars, laundry, and the like; also, often including stables and other outhouses, the barns and cowhouses of a farm, etc.’ 12.According to the OED the word is a “[c]orruption of It. casa house”. Allan & Burridge (1991: 78), however, speculate that “kazi originated in East Africa, probably among soldiers: it is the Kiswahili word for ‘work’ and, we think, referred to the work in digging latrines”. Neither etymology is supported by quotations.

“Non olet”

(gangern OE, feltun OE, room-house c1250 (room here meaning ‘spacious’, see 4.5), chamber foreign 1297 (foreign here meaning ‘out of doors’), house of easement 1513/1712, draught-house 1594/1611/1884, necessary vault 1609, house of office 1613/1652/1823, gingerbread-office a1643, necessary place 1697, passagehouse 1727, outhouse 1819–, small room 1858–, etc.). Some of the words in the category may be called strong euphemisms (ironic ones?), because the place mentioned hides its purpose: woodshed 1844–, the geography (of the house) 1927–, throne-room 1941, House of Lords 1961–.13 Reference is usually to some sort of building, but occasionally also simply to a hole in the ground or a ditch (gangpytt OE, grepe OE, grof OE, siege-hole 1477-79, bench-hole 1555–a1656, dike 1923–) or to the container (can 1900–).14 Alternatively, the word may indicate that the place in question is outside of or behind one’s dwelling (foreign 1303–1570, backside 1704, back-house 1847/1939, rear(s) 1902–; heolstor OE means ‘hiding place, retreat’). A rather special case in this category is constituted by head(s) 1748/1905– ‘a ship’s latrine (in the bows)’, metaphorically named for its place in the bows of a ship.15 The concept ‘place’ is attested throughout the history of English. 4.3 ‘A place for men or women’ A distinction between toilets for men and women is only usual in public toilets. The words in question are: ladies’ room 1880–, ladies’ cloakroom 1918–, ladies’/ Ladies’ 1918–, men’s (room) 1929, gentlemen(s’) 1929–, gents 1938–, powder-room 1941– (also classed in 4.10), little girls’ room 1949–, little boys’ room 1957–, women’s room 1961–. Words of this type are only attested from the late 19th century onwards; in present-day usage they are among the most frequent, however. For want of a better place, two isolated late 16th and early 17th century instances of common house and commons may also be classed here (see quotations below). The toilet here is seen as ‘a place for all’.

13.See 6. for a discussion of strong versus weak euphemisms. 14.The 19th-century Hopper Closet (Wright 1960: 201–202) is not listed in the OED, but hopper is obviously an extension of hopper n.1, sense 3 ‘in a corn or other grinding mill, a receiver like an inverted pyramid or cone, through which grain or anything to be ground passes into the mill; so called because it had originally a hopping’ and sense 4 ‘applied to similar contrivances for feeding any material to a machine, and, generally, to articles resembling a mill hopper in shape or use’. — Many of the words in categories I have excluded (‘chamber-pot etc.’, ‘close-stool/commode’ etc., see above) denote a container. 15.I owe this example to Robert Stockwell (oral communication).

97

98

Andreas Fischer

1596 Harington Metam. Ajax (1814) 51 By digging pits in the earth or placing the common houses over rivers. (OED s.v. common house, sense 4.b (a)) 1624 Heywood Captives iii. in Bullen O. PI. IV, Heare’s a place Though neather of the secretest nor the best, To unlade myself of this Iniquity [a dead body]. Hee’s where hee is in Comons. (OED s.v. commons, sense 5)

4.4 ‘Privacy’16 An important feature of toilets is the privacy they provide (and which most people find desirable), and so it is not surprising to find the concept ‘privacy’. The best example is one of the oldest words still used, namely privy a1225– (from French privé), but there is also privy-house c1460–1679, withdraught 1493-94–, draught c1500–1681, draught-house 1594/1611/1884, private 1600, retreat 1653, closet 1869– (also listed in 4.2). 4.5 ‘Smallness’ The concept ‘smallness’ is represented in only a few words. It may well be implied in closet 1869– (see 4.2 and 4.4), but is explicit in little house 1720–, petty 1848/1961 (“widely prevalent in familiar use” according to the OED, from French petit), small room 1858–, smallest room 1930–. This appears to be a fairly recent concept, evidenced only since the 18th and, above all, 19th centuries. Very occasionally, and only in earlier periods, do we find the opposite, as in room-house c1250 (room here meaning ‘spacious’) or long-house 1622 (possibly a lavatory for several people, in which case it should be classed with common house and commons in 4.3). 4.6 ‘Sitting, seat’ ‘Sitting’ or ‘seat’ are obvious conceptualisations of a toilet, and they are in fact found throughout the post-Anglo-Saxon history of English, with words like siege, stool, chair or seat
: siege c1400–1555, siege-house 1440–1647, siege-hole 1477-79, stool of ease 1501/1561, stool 1516-17–1768-74, the stool 1542–1676, stool of easement 1580, necessary stool 1761, easing chair 1771, seat of ease c1850, throne-room 1941.

16.This is Allan & Burridge’s “a place where there is privacy”.

“Non olet”

4.7 ‘Physical comfort, ease’17 Physical comfort is conceptualized in words such as stool of ease 1501/1561, stool of easement 1580, closet of ease 1662, house of ease 1731, easing chair 1771, seat of ease c1850, comfort station 1923– (also listed in 4.10), rest room 1955–. Note that the key word in this category from the 16th to the 19th centuries was ease (lost today in this meaning), and that the two 20th century examples are Americanisms. Somewhat tentatively, I also place the term convenience here: convenience 1841–, public convenience 1938–. 4.8 ‘Necessity’ The feeling of need to relieve oneself is conceptualized in expressions such as necessary vault 1609, necessary house 1611–1828, necessary place 1697, necessary 1756–1844/1877, necessary stool 1761. As can be seen, this concept was lexicalized only in the 17th to 19th centuries, and the word used was always necessary. 4.9 ‘Water, washing, cleaning’18 As was pointed out in the previous section, the water closet was an invention of the 16th and above all of the 18th century, and it is thus only from that century onwards that we can expect conceptualizations focusing on water, washing and cleaning in general: water-closet 1755–, lavatory 1845 (first attested in 1656, from Lat. lavatorium ‘a place for washing’),19 washroom 1892–, W. C. (short for water-closet) 1892–, lav (short for lavatory) 1913–, lavabo (Lat. ‘I shall wash’) 1930–, bathroom 1934–, W. (short for W. C.) 1953–, lavvy 1961–, washroom1962.20 It should be noted at this point that the words water-closet and lavatory predate Alexander Cummings’ invention of 1775 (see 3. above); the early quotations simply refer to a room in which water was available. Note also that closet (in the sense of ‘toilet’) and earth-closet are later than water-closet, the OED attesting them only from 1869 and 1870 onwards, respectively.21 It is

17.This is Allan & Burridge’s “a place of easement”. 18.This is Allan & Burridge’s “a place where there is water”. 19.This is the dating according to the OED
; the date given in the HTE is 1924–. 20.In the sense ‘a room equipped for washing’ washroom is attested much earlier, namely from 1806 onwards. 21.Earth-closet, then, appears to be what Leisi (1997) calls a “new word ex contrario”, i.e. a

99

100 Andreas Fischer

interesting that in all the conceptualizations above the purpose of the water is washing or cleaning and not flushing; in fact the word flush, rather surprisingly, is not found in the corpus at all. From the point of view of etymology, latrine 1642–1808/1867 (from French latrine, from Latin latrina, a contraction of lavatrina) and the shortening lat(s) 1927– belong in this group as well, but the connection with washing was already lost when latrine was borrowed from French in the 17th century. The word loo 1940– may also belong here, since both Allan & Burridge (1991: 79) and Noble (1982: 70) mention French l’eau as a possible etymology.22

4.10

‘Care of the body (other than cleaning, washing)’

Washing is a part of bodycare, and so it is not surprising that the transfer from ‘using a lavatory’ to ‘washing’ is further extended to ‘caring for one’s body’. The prime example is the word toilet
: originating in French as a diminutive form of toile, it first meant ‘a piece of stuff used as a wrapper for clothes’ (1540) and ‘a towel or cloth thrown over the shoulders during hair-dressing’ (1648), then ‘a cloth cover for a dressing-table’ (1682), ‘the articles required or used in dressing’ (1662), ‘the table on which these articles are placed; a toilet-table’ (c1695), ‘the action or process of dressing’ (1681), ‘the reception of visitors by a lady during the concluding stages of her toilet’ (1703), ‘manner of style of dressing; dress, costume, “get-up”’ (1752) and — finally — ‘a dressing-room […]. Hence, a bath-room, a lavatory; (contextually), a lavatory bowl or pedestal; a room or cubicle containing a lavatory’. This last sense is attested quite late, from 1819 onwards, and the first quotations which clearly document the sense ‘lavatory’ rather than ‘bathroom’ are found only in the 20th century. In presentday British English toilet is a fairly neutral word,23 while in American English

new word for an old, well known item which has become necessary because of an innovation (in this case, the water-closet). Some of Leisi’s examples: longbow (coined because of crossbow), tonal music (coined because of atonal music), natural language (coined because of artificial language), etc. 22.According to other suggestions (Allan & Burridge 1991: 79–80 and Noble 1982: 70) loo may be derived from bourdalou(e) ‘a type of oval slipper-shaped urinal or chamber-pot intended mainly for the use of ladies when travelling’, from lieu ‘place’ or from Waterloo. In the OED the derivation from Waterloo is rejected (since it cannot be demonstrated) and the etymology of the word is said to be “obscure”. 23.One of the anonymous reviewers of this paper notes that the word toilet “is still not neutral for all speakers of [British English], although generally so in public contexts. It

“Non olet”

it has come to mean almost exclusively ‘toilet bowl’ and is regarded as a vulgarism (see 6. below). Expressions like wardrobe 1382–c1450/1847, comfort station 1923– (also listed in 4.7), powder-room 1941– (also listed in 4.3), cloak-room 1953–, cloaks 1957– may also be seen as belonging in this category. 4.11

Names

Two first names have been and are still being used for the toilet: jakes c1530– (jakes-house 1577, ajax 1588–1611) and john 1932– (johnny 1932–). In both cases the OED simply gives the names in question (Jaques/Jakes or Jakkes ‘Jack’s’, and John) as etymologies without venturing any further explanation. Noble (1982: 71) suggests that jakes commemorates Sir John Harington, but this is unlikely since Harington himself, like Shakespeare, puns on the names Ajax, Jacques and (a) jakes.24 Another term probably derived from a personal name is jerry n.3 (from the Biblical Jeroboam), which the OED records with the senses ‘water closet’ 1850 and ‘chamber-pot’ 1859–.25 The modern toilet is an industrial product and it comes as no surprise that trade and product names are occasionally used as common nouns. The oldest example is vespasienne 1922– (from French, after the Roman emperor Vespasian),26 anglicised as vespasian 1938/1980. Elsan is the trade name of the Elsan (from Ephraim Louis Jackson, and sanitation) Manufacturing Co. for ‘a type of lavatory in which the sludge is rendered inoffensive by chemical means’. The OED attests it from 1939-40 onwards. Superloo 1969–, finally, seems to have been a specially named facility in railway stations in the late 1960s and early 1970s: 1969 Daily Tel. 17 Dec. 1/4 The charge for using the ‘*superloo’ at Euston and Victoria will go up from 6d to 1s. (OED s.v. super-, sense 6.c.) 1972 Travelling Autumn 43/3 Edinburgh’s Waverley Station..will provide superloos, catering facilities. (OED s.v. super-, sense 6.c.)

possibly still has the stigma of being lower middle class, as in A. S. C. Ross and Nancy Mitford’s work, where it’s one of the criterial pairs. That was in the more sensitive 1950’s, but I think some people still avoid the word”. According to Ross (1954: 46/1956: 32), toilet-paper is non-U, lavatory-paper U. 24.He does so in the title (A New Discourse of a Stale Subject Called the Metamorphosis of Ajax; see note 4) and in the text of his treatise (Barber 1997: 253). 25.According to Allan & Burridge (1991: 78) crapper may be a last name; see note 29 below. 26.See note 1.

101

102 Andreas Fischer

Miss White is also a name of sorts. In the OED it is only attested once, from Joyce’s Ulysses, and may well be an instance of private language:27 1922 Joyce Ulysses 347 When she wanted to go where you know she said she wanted to run and pay a visit to the Miss White (OED s.v. miss n.2, sense 2.h.)

4.12

‘Direct terms’, vulgarisms

Having established these eleven different conceptual categories which all denote an aspect of the ‘thing’, but not the ‘thing’ itself, one may well ask whether it is not possible to call a spade a spade. Apparently it is possible, but only rarely, the reason being that there are no neutral words for the acts of defecating and/or urinating. To defecate and to urinate are Latinisms used mostly by doctors and in technical language; ordinary people who want to or have to talk about these matters must resort either to baby language (to do a poo, to do a wee) or to ‘direct’ words which are felt to be vulgarisms, be they relatively old like shit and piss,28 or more recent like bog and crap. We cannot be sure about the status of the two Old English nouns in this category (adel(a) ‘mud, dirt, filthy place’, earsgang ‘privy; excrement, anus’) and there are some isolated words that may be called neutral (purging place 1577, issue 1588/1616, cloaca 1840–). All others, however, must be regarded as vulgarisms: boggard 1552–1647, bog-house 1666–1761, bog a1789–, shit-house 1795–, piss-house 1931, crapper 1932–, shouse (a contraction of shit-house) 1941–, shitter 1969–. A note on bog (boggard, bog-house) and crapper. One might suspect bog ‘toilet’ to be a transferred use of bog ‘a piece of wet spongy ground’, but the OED sees them as two entirely different words and connects boggard, bog-house and bog ‘toilet’ with a verb to bog (v.3) meaning ‘to exonerate the bowels; also trans. to defile with excrement’. In unusually strong language it is called “[a] low word, scarcely found in literature, however common in coarse colloquial language”. No etymology is suggested and no quotations are given. Crapper is derived from the verb to crap ‘to defecate’, documented from 1846 onwards; two 19th century quotations in the OED (from Hotten’s Slang Dictionary 1859, 1874) contain further words for ‘toilet’ which do not appear as separate

27.Private language is probably rich in idiosyncratic words for ‘toilet’, which, however, will only rarely find their way into writing. 28.The verbs to shit and to piss are attested from early Middle English onwards; shit/shite is Germanic, while piss is a borrowing from Old French.

“Non olet” 103

headwords: crapping case, crapping ken and crapping-castle. Just as in the case of bog, the OED does not suggest an etymology.29

5. Categorisation through metonymy In the previous section I have classed the documented words for ‘toilet’ into altogether twelve categories. Ten (or perhaps eleven)30 of these are instances of indirect classification in that they name neither ‘the thing’ itself nor its central purpose, highlighting instead another aspect of it (privacy, comfort, cleanliness, etc.). In other words, they are all metonymic expressions, at least in origin, and following Lakoff & Johnson (1980) one might say that we live not only by metaphors but also by metonymies. While metaphors “allow us to comprehend one aspect of a concept in terms of another” (Lakoff & Johnson 1980: 10), metonymies allow us to comprehend a concept through one of its aspects. Being characterised by contiguity rather than (perceived) similarity, metonymies may be somewhat more constrained and thus less subjective and less ‘creative’ than metaphors, but like metaphors they give insight into how a speech community conceptualises its world.31 Metonymies are restricted by inherent properties of the objects or ideas they represent (when you say that you have drunk a glass of water you do this because one usually drinks water from a container made of glass), but they are not rigidly fixed: metonymies, like metaphors, thus may differ from speech community to speech community or may change over time. The material presented here, being restricted to English across its documented history, does not enable us to make cross-linguistic and cross-cultural comparisons, but it allows some insight into change over time. Some categories,

29.According to Allan & Burridge (1991: 78), however, crapper might be a name and would thus belong in my Section 4.11: “Crapper is curious: it ought to be neutral if it is named after Thomas Crapper who patented a valveless ‘water waste preventer’ which guaranteed ‘a certain flush with an easy pull’ in 1882 […]; but crapper is dysphemistic, presumably because of the verb and noun crap ‘shit’, which we say later (following the OED) derives from a Dutch verb. Either Crapper’s name was remarkably coincidental, or the stressed vowel in the old word croppen […] got changed to ‘a’ as a result of Crapper’s name”. 30.The names discussed in 4.11 may also be seen as metonymic. 31.In the wake of Lakoff & Johnson’s pioneering book, metaphors were studied more intensely than metonymies for a long time, but this has changed in recent years; see, for example, Barcelona (2000), Dirven & Pörings (2002), Goossens et al. (1995) and Panther & Radden (1999).

104 Andreas Fischer

like ‘the place itself, location, container’ (4.2) are in evidence throughout the history of English, while others such as ‘sitting, seat’ (4.6) or ‘care of the body (other than cleaning, washing)’ (4.10) are attested from Middle English onwards.32 New categories may result from external, i.e. technical or social developments (see 3.): this is the case with ‘water, washing, cleaning’ (4.9; in operation since the 18th century, resulting from the introduction of the water closet) and ‘a place for men or women’ (4.3; in operation since the late 19th century, resulting from the introduction of ‘modern’ public toilets). More interesting and more difficult to explain, however, is the appearance (or disappearance) of categories which cannot be explained in this way. Examples are ‘privacy’ (4.4; not prominent today: the word privy is used, but probably only few speakers associate it with ‘privacy’) and ‘necessity’ (4.8; operative only in the 17th to 19th centuries).

6. Taboo, euphemisms and dysphemisms In the previous section I have called the classification of ‘toilet’ words indirect or, more specifically, metonymic, since the words in categories 4.1 to 4.11 emphasize an aspect of the ‘thing’ rather than ‘the thing’ itself. Indirectness here has a very specific, euphemistic function in that it provides speakers with the means to deal with the taboo topics of defecation and urination. When confronted with taboo areas speakers often (or rather, ideally) have three types of words at their disposal: neutral words, euphemisms as a means of paraphrasing and thereby avoiding the tabooed topic, and dysphemisms as a conscious way of naming the ‘thing’ directly and of emphasizing its negative associations. Euphemisms are intended to spare the hearer; they are a means for speakers to save their own and their hearers’ face. Dysphemisms, by contrast, are meant to shock or offend hearers; like swear-words they are face-threatening acts par excellence. If hearers are ‘shockproof’, however, dysphemisms may also function as in-group markers that signal closeness and linguistic intimacy. To die is a neutral word, to pass away is a euphemism, to croak a dysphemism.33 What is the situation in the area under investigation?

32.Since the vocabulary of Old English is documented less well than that of later periods, these (as well as other) concepts may have been lexicalised already in Old English. 33.Linguistic taboo, euphemisms and dysphemisms in general are discussed in Allan & Burridge (1991), Enright (1986), Noble (1982) and Pei (1970).

“Non olet” 105

As discussed in the previous section, the words in the first eleven categories are all metonymic expressions, at least in origin. Metonymies, of course, are not euphemisms per se. In this case, however, they are, since the words in question express an aspect (the place, privacy, comfort, cleanliness, etc.) which is not the central one. By not mentioning the central aspect they avoid it and are thus euphemisms.34 They are only what might be called ‘weak’ euphemisms, however, by contrast with those words which do more than just name one aspect of the thing at the expense of others, but which actually hide its purpose. These may be called ‘strong’ or full euphemisms: in category 4.2 outhouse is a weak euphemism, woodshed a strong one, in category 4.9 water-closet is a weak euphemism, bathroom a strong one. Note that some of these strong euphemisms are so ‘over the top’ that, in fact, they reveal more than they hide: the geography of the house, throne room, House of Lords (all in 4.2). From the point of etymology, there are practically no words in the whole corpus that may be classed as neutral.35 A substantial number of the ‘original’ euphemisms, however, have lost their euphemistic qualities and have become neutral words. Examples are lavatory, water-closet, W.C. or toilet (which is the prototypical example; note that in British English it is a fairly neutral word, while in American English it is felt to be a vulgarism [denoting the container]). This ‘lowering’ from euphemism to neutral word (and, occasionally, to dysphemism) is the fate that befalls euphemisms generally. In a never-ending process the loss of euphemisms calls for the creation of new ones, which may, in turn, lose their euphemistic value. It is in this way that the ‘tabooed’ lexicon of a language may grow. It might be objected that urination and defecation were less taboo in some periods than in others. In his book Chaucer’s Bawdy, for example, Ross (1972: 18) states: Excretion was an accepted and semipublic event that Chaucer rarely uses for comedy. In the last [i.e. 19th] century, these body functions have become rites perfomed in the shamefast privacy of a closed room, the excreta being immediately laved away by sparkling rivulets, to be seen and smelled no more.

34.For discussions of ‘toilet’-words from the point of view of euphemism see Allan & Burridge (1991: 78–80), Barber (1997: 252–254), Burchfield (1986: 24–26) and Noble (1982: 70–72). 35.For possible exceptions see 4.12.

106 Andreas Fischer

However, even though people may have been less squeamish in earlier periods than in the hygiene- and privacy-oriented present, the words and categories discussed in this paper show that euphemisation has been at work throughout: it must be regarded as the principal reason why the ‘toilet’ vocabulary of English is so rich. Words “that do not smell”, it appears, have always been and still are, popular currency.

References Allan, Keith & Kate Burridge. 1991. Euphemism & Dysphemism: Language Used as Shield and Weapon. New York & Oxford: Oxford University Press. Barber, Charles. 1997. Early Modern English. Edinburgh: Edinburgh University Press. Barcelona, Antonio, ed. 2000. Metaphor and Metonymy at the Crossroads: A Cognitive Perspective. (= Topics in English Linguistics, 30.) Berlin & New York: Mouton de Gruyter. Bartels, Klaus. 1992. Veni, vidi, vici: Geflügelte Worte aus dem Griechischen und Lateinischen. 9th edn. Darmstadt: Wissenschaftliche Buchgesellschaft. Burchfield, Robert. 1986. “An Outline History of Euphemism in English.” Enright 1986. 13–31. Coleman, Julie & Christian J. Kay, eds. 2000. Lexicology, Semantics and Lexicography: Selected Papers from the Fourth G. L. Brook Symposium, Manchester, August 1998. (= Amsterdam Studies in the Theory and History of Linguistic Science, Series IV: Current Issues in Linguistic Theory, 194.) Amsterdam & Philadelphia: John Benjamins. Dirven, René & Ralf Pörings, eds. 2002. Metaphor and Metonymy in Comparison and Contrast. (= Cognitive Linguistics Research, 20.) Berlin & New York: Mouton de Gruyter. Donno, Elizabeth Story, ed. 1962. Sir John Harington’s A New Discourse of a Stale Subject, Called the Metamorphosis of Ajax: A Critical Annotated Edition. London: Routledge & Kegan Paul. Enright, D. J., ed. 1986. Fair of Speech: The Uses of Euphemism. Oxford: Oxford University Press. Goossens, Louis, Paul Pauwels, Brygida Rudzka-Ostyn, Anne-Marie Simon-Vandenbergen & Johan Vanparys. 1995. By Word of Mouth: Metaphor, Metonymy and Linguistic Action in a Cognitive Perspective. (= Pragmatics & Beyond, New Series, 33.) Amsterdam & Philadelphia: John Benjamins. Guerrand, Roger-Henri. 1985. Les lieux: histoire des commodités. Paris: Editions La Découverte. Illi, Martin. 1987. Von der Schîssgruob zur modernen Stadtentwässerung. Herausgegeben von der Stadtentwässerung Zürich, Abteilung des Bauamtes I. Unter Verwendung eines unveröffentlichten Manuskriptes von Hansruedi Steiner. Zürich: Verlag Neue Zürcher Zeitung. Kay, Christian J., Irené A. W. Wotherspoon & Louise M. Sylvester. 2001. “One Thesaurus Leads to Another: TOE, HTE, TME”. Lexis and Texts in Early English: Studies presented



“Non olet” 107

to Jane Roberts ed. by Christian J. Kay & Louise M. Sylvester. (= Costerus, New Series, 133.) Amsterdam & Atlanta, GA: Rodopi. 173–186. Lakoff, George & Mark Johnson. 1980. Metaphors We Live By. Chicago & London: The University of Chicago Press. Leisi, Ernst. 1997. “Neuwörter ex contrario — ein oft übersehener Worttypus.” Lebende Sprachen 42.3. 106–108. Noble, Vernon. 1982. Speak Softly: Euphemisms and Such. [Sheffield:] The Centre for English Cultural Tradition and Language, University of Sheffield. Palmer, Roy. 1973. The Water Closet: A New History. Newton Abbot: David & Charles. Panther, Klaus-Uwe & Günter Radden, eds. 1999. Metonymy in Language and Thought. (= Human Cognitive Processing, 4.) Amsterdam & Philadelphia: John Benjamins. Pei, Mario. 1970. Words in Sheep’s Clothing. London: George Allen & Unwin. Roberts, Jane & Christian J. Kay with Lynne Grundy. 1995. A Thesaurus of Old English. 2 vols. (= King’s College London Medieval Studies, 11.) London: King’s College. Ross, Alan S. C. 1954. “Linguistic Class-Indicators in Present-Day English.” Neuphilologische Mitteilungen 55. 20–56. ———. 1956. “U and Non-U: An Essay in Sociological Linguistics.” Noblesse Oblige: An Inquiry into the Identifiable Characteristics of the English Aristocracy. By Alan S. C. Ross, Nancy Mitford, Evelyn Waugh, ‘Strix’, Christopher Sykes & John Betjeman. Illustrated by Osbert Lancaster. London: Hamish Hamilton. 9–36. Ross, Thomas W. 1972. Chaucer’s Bawdy. New York: E. P. Dutton. Simpson, J[ohn] A. & E[dmund] S. C. Weiner, eds. 1989. The Oxford English Dictionary. 2nd edn. Oxford: Clarendon Press. Wright, Lawrence. 1960. Clean and Decent: The Fascinating History of the Bathroom & the Water Closet. London: Routledge & Kegan Paul; repr. London: Penguin, 2000.



Intrusive [h] in present-day English accents and ·hÒ-insertion in medieval manuscripts Hypercorrection or functionally-motivated language use? Martina Häcker University of Konstanz

In his entry on the letter H, the first editor of The Oxford English Dictionary (originally called A New English Dictionary), James A. Murray (1901) states: “During the Middle-English period, and down to the 17th c., we find numerous instances of the non-etymological absence or (more often) presence of initial h in native words.” Unlike Murray, many of his contemporaries considered [h]dropping as a comparatively recent phenomenon, using the first occurrence of contemporary criticism of [h]-less pronunciations as their primary criterion for dating it, while they largely ignored [h]-insertion, interpreting it as hypercorrection. The question was reopened by James Milroy (1983), who claims on the basis of spellings that h-dropping was widespread in the Middle English period: By about 1300 there appears to have been loss of initial /h/ in the dialects of Eastern England from Kent and Surrey to Lincolnshire. … Early ME [Middle English] /h/ loss is from the regions that were amongst the most important commercially and administratively, and it comes from texts many of which are quite formal in style and learned in content. (Milroy 1983: 47–48)

Milroy argues that the omission of initial /h/ must have carried a certain amount of prestige, which would be in line with the findings of sociolinguistic research that changes typically start from the (upper) middle class. The /h/-less prestige variety, according to Milroy (1983: 50), most probably arose from the French-English contact situation, as h-loss “does not seem a ‘natural’ change in Germanic”. He concludes that

110

Martina Häcker

even if there were sporadic tendencies to /h/-loss in OE [Old English], the French-English contact situation was the single most important influence on its rapid progress in ME.

While Milroy and earlier scholars disagree in their interpretation of manuscript evidence, they all appear to take it for granted that some form of language contact with a Romance language was responsible for h-loss: language contact with a spoken /h/-less language in Milroy’s view (French), and with either a spoken /h/-less language (Anglo-Norman) or a written language that was characterised by ·hÒ-variation (Medieval Latin), according to earlier scholarship.1 Thus Luick (1964: 871) states that: Initial h before vowels was always retained in the early period. If occasionally an initial h is missing in Old English texts or is inorganically inserted in a word beginning with a vowel, this is most likely to be explained by the fact that some scribes were influenced by Latin manuscripts written by Norman scribes in which a certain amount of uncertainty existed with respect to initial h.2 [author’s translation]

Wyld (1920: 295–296) argues along the same lines: Norman scribes are very erratic in their use of h- in copying English manuscripts, and we therefore cannot attach much importance to thirteenth- or even to fourteenth-century omissions of the letter which occur here and there.

Despite the wide-spread agreement on the idea that Normans are directly or indirectly responsible for unetymological spellings in English manuscripts,

1.Beeson (1925: 15) describes the use of h in Medieval Latin manuscripts as follows: “h was freely added or omitted, frequently causing confusion (ac, hostium, hortus, and especially in forms like hii, hiis, where it is sometimes impossible to tell whether the form is from hic or is).” The loss of initial h is already attested in Late Latin, where both addition and omission of ·hÒ is frequent, as are comments on h-variation by Latin authors. Thus, for example, Augustinus (conf. 1,18, 29) states that an [h]-less pronunciation of homo is criticised more severely by his contemporaries than disobeying God’s laws by hating a fellow citizen, while Catullus (84) in his famous epigram pokes fun at someone called Arrius, who apparently was in the habit of inserting [h] before words beginning with a vowel, thus pronouncing insidias (ambush) as hinsidias and Ionios (Ionians) as Hionios (Kramer 1976: 48–57). 2.Luick’s German original is: “Anlautendes h vor Vokal hat sich dagegen in alter Zeit durchwegs gehalten. Wenn gelegentlich in altenglischen Texten h im Anlaut fehlt oder unorganisch einem vokalisch anlautenden Wort vorausgesetzt wird, wie aefde für haefde “hatte” und andererseits hierre für ierre “Zorn”, so rührt dies wohl daher, daß manche Schreiber durch das Vorbild lateinischer Manuskripte, die von Romanen herstammten und in denen eine gewisse Unsicherheit bezüglich der Setzung des h herrschte, beeinflußt waren.”

Intrusive [h] in present-day English accents

there is no conclusive evidence to prove this.3 The assumption appears to have been that the mother tongue of Normans was /h/-less and the variation in manuscripts would then be a transfer phenomenon. French was, however, not an /h/-less language at the time of the Norman Conquest. Not until 1673 did the Académie Française comment on the sloppy speech of the people of “d’outre Loire” (the other side of the Loire) and of Paris, who no longer pronounced ·hÒ in words that do not derive from Latin, and regretted the spreading of that bad habit (Thurot, vol, 1881: 396, cited from Pope 1934: 94): Dans tous les autres mots qui ne viennent point du latin, l’h aspire fort, quoyque le mauvais usage introduit par les gens du province d’outre Loire, et mesme par le peuple de Paris s’efforce de l’abolir tout à fait.

According to Thomason and Kaufman (1988: 127), in Normandy ache aspiré is pronounced up to this day,4 and reflexes of ache aspiré are still found in present-day standard French in the elision and liaison rules that inter alia require the elision of the vowel of the masculine singular definite article and the pronunciation of the ·nÒ of the masculine singular possessive pronoun before words beginning with ache muet (e.g. l’ami [lami] “the friend”, mon ami [mõnami] “my friend”), but not before those beginning with ache aspiré (le héros [lœero] “the hero”, mon héros [mõero] “my hero”). This means that language contact with Romance would only explain ·hÒ-omission in words of Latin etymology or Greek words which were borrowed through Latin, but not ·hÒ-omission or ·hÒ-insertion in native English words. Moreover, [h]-loss is not unnatural for a Germanic language as Milroy (1983: 50) claims, but is also found in some German and Swedish varieties (Minugh 1985, Häcker 1998). While the omission of the letter ·hÒ in initial position strongly suggests that a word was pronounced without [h], the function of ·hÒ-insertion is not clear. Most scholars assume that ·hÒ-insertion and ·hÒ-omission are reflections of the same phenomenon. They interpret ·hÒ-additions as hypercorrections, the underlying assumption being that the scribe is a speaker of an /h/-less variety and therefore uncertain where ·hÒ is required in the standard language. Another

3.The myth of the Anglo-Norman scribe goes back to Skeat (1897). For a critical review of this term see Clark (1992). 4.Thomason and Kaufman (1988: 127) state that “Romance languages generally lack both /h/ and (at least word-initially) /w/, since Latin /h/ > Ø (…). But medieval French had — and modern French dialects of Picardy, Wallonia, and Normandy still have — word-initial /h/ in Germanic loanwords”.

111

112

Martina Häcker

function of ·hÒ-insertion that has been suggested in the literature is that it is a diacritic marking a hiatus. This hypothesis was advanced by Scragg (1970) in his analysis of initial ·hÒ in Old English, while a third possibility, namely that ·hÒ in words with ·hÒ-insertion represents the same sound as etymological ·hÒ without being the result of hypercorrection, is not considered in the literature. It is the purpose of this paper to investigate the contexts in which ·hÒ-insertions are found and to identify possible functions of ·hÒ-insertion. Scragg’s study suggests that ·hÒ-omissions and ·hÒ-insertions occur at different times at different places as well as in different phonological contexts. If a distinction between function words and lexical words is made, we find a north-south division in Old English manuscripts. In West Saxon texts, there is evidence of ·hÒ-variation in both Early West Saxon and in the tenth- and eleventh-century texts, with the highest number of instances in the Exeter Book. Here initial ·hÒ is omitted nine times, in words such as elpe for “helpe”, eortan for “heortan”, but also inserted in words, such as hus for “us”, his for “is”. Northumbrian texts show a different pattern: no examples of ·hÒ-insertion or ·hÒ-omission in the early period (seventh and eighth centuries), but a considerable number of h-insertions in the later period (tenth century), in particular in the Lindisfarne Gospels. ·hÒ-omission occurs predominantly in positions which may be unstressed, such as is instead of his, and the prefix und instead of hund “hundred” in medial positions of numbers. The Durham Ritual, glossed by the same scribe, shows fewer instances of ·hÒ-insertion and ·hÒ-omission and the Northumbrian part of the Rushworth Gospel has only one instance of ·hÒ-omission (it for hit) and a small number of ·hÒ-insertions such as hengla for engla. This means that despite differences between individual scribes and also individual texts by the same scribe there is evidence in late Northumbrian texts of a considerable amount of ·hÒ-insertion, but relatively little ·hÒ-omission, except in frequent unstressed words, i.e. the function words which are pronounced without initial [h] in present-day English unless they are emphasised. Scragg’s (1970: 192) comments on the phonological context of ·hÒinsertions are perceptive: More than half the unexplained examples occur between two vowels. This is most noticeable when the inorganic h appears after a prefix ending in a vowel, e.g. Kentish Proverbs gehydlect, Parker Chronicle ahebbad, Benedictine Rule gehendod, and various MSS of Ælfric’s Latin Grammar of-ahefesige, gehyppe, gehiht
; but many of the other examples occur initially when the preceding word ends in a vowel: Tanner Bede helde and one example of his, Kentish Proverbs his and hup, WS Gospels B MS hyfelan, Paris Psalter prose hanweald.

Intrusive [h] in present-day English accents

The occurrences seem too frequent to be coincidental, especially since they parallel so closely the use in Late Latin of h as a diacritic signifying a hiatus.

Does the preferred position of ·hÒ-insertions allow us to draw any conclusions about the nature of the sound represented by the symbol ·hÒ? Scragg (1970: 195) argues for a glottal stop: For unhistoric insertion of h, the explanation of use of h as a diacritic is offered for the majority of West Saxon examples, at least. Here, reflection of speech is possible, in that the h may represent a hiatus between the vowels perhaps in the form of a glottal stop. But the important point is that inorganic h seems not to represent an inorganic aspirate, and as with omission of h, we may say that few examples point directly to an instability of the aspirate. (Scragg 1970: 195)

Scragg’s interpretation raises the question of why scribes would use the letter ·hÒ to represent a glottal stop that marks a hiatus. A closer look at the relationship between sound and symbol seems therefore appropriate. Classical Latin has a one-to-one correspondence between sound and symbol. When the Latin alphabet was adopted for languages that had sounds which were absent from Latin, either additional symbols had to be introduced or combinations of symbols had to be used to represent such sounds. The former strategy is represented in the use of the thorn, which was borrowed from the runic alphabet, while the latter can be seen in the combination ·thÒ, which gradually replaced the thorn in Middle English (cf. Scragg 1974: 2). Language change presented no problem for the system as long as it was regular. In cases of phonetically conditioned allophonic variation the same symbol could stand for two sounds. This is, for example, the case with ·fÒ in Old English, which corresponded to either [f] or [v] depending on the phonetic context. Irregularities in the correspondence of sound and symbol are caused by language contact and irregular changes.5 In pre-Conquest English the relationship between sound and symbol was much more regular than in later times. The symbol ·hÒ is found in Old English on its own and in combinations. On its own it represents the allophones of the phoneme /h/, that is [h] [ç], and [x]. In combinations it

5.Thus the letter sequence ·ghÒ which originally corresponded to the allophones [x] or [ç] has several realisations in present-day English. The different realisations in enough, [f], Keighley [θ] and though, where the only reflex of former [x] is the diphthongization of the vowel, are the result of irregular or incomplete change, while the realisation in spaghetti [g] is due to the fact that the word was borrowed without changing the spelling, which is based on the Italian system where the addition of ·hÒ indicates that the symbol ·gÒ that normally would represent an affricate before front vowels here represents a plosive.

113

114

Martina Häcker

is restricted to initial position in the combinations ·hnÒ, ·hlÒ, ·hrÒ, and ·hwÒ. These are held by some scholars to be originally sound sequences, [xn], [xl], [xr], [xw], which later developed into voiceless counterparts of [n], [l], [r], and [w]. Other scholars claim that the combinations already represent [n], ® [®l], [®r], and [w] ® in the earliest Old English texts, which implies that initial ·hÒ in these combinations functions as a marker of voicelessness.6 From Middle English onwards ·hÒ is used in a number of combinations, in which its function appears to be a modification of the sound associated with the symbol preceding ·hÒ. This modification may refer to the manner of articulation, the place of articulation or voice: 1. symbol of a plosive + ·hÒ = fricative or affricate (·chÒ, ·thÒ, ·ghÒ) 2. symbol of an alveolar fricative + ·hÒ = palato-alveolar (·shÒ) 3. symbol of voiced approximant + ·hÒ = voiceless approximant (·whÒ) The letter ·hÒ used as a symbol to represent a glottal stop does not appear to fit particularly well into the English orthographic system, as it does not represent a plosive in any of the combinations, nor does a glottal stop between vowels fit in the Old English phonological system. Old English fricatives had phonologically conditioned voiced and voiceless allophones, the voiced allophone being used between vowels, which is exactly where ·hÒ appears in Scragg’s southern examples. It therefore seems more likely that the unetymological ·hÒ-insertions represent a fricative, which may well have been voiced. Its function would then be that of a linking sound similar to that of [j] and [w] in present-day RP (cf. Cruttenden/Gimson 1994: 264). The evidence of alliterative poetry, on which Scragg’s conclusion that ·hÒ does not seem to be the aspirate is based, is not conclusive. Scragg (1970: 186–187) adduces alliteration as evidence, pointing out that “Hebreos” in various poems alliterates with words beginning with vowels, as does “Heliseus” in Juliana and “Holofernes” in Judith. Scragg’s view that such an alliteration would not occur if ·hÒ corresponded to [h] depends on his interpretation of [h] as a consonant. Old English scribes may, however, have perceived [h] as a breathy vowel onset and may therefore have permitted alliteration with vowels.7

6.See Lutz (1991: 21–37) for a detailed discussion. 7.Cooper’s 1687 description of h suggests that h may have been considered a breathy vowel onset at least by some people in the seventeenth century: ‘H (…) hath no particular formation, neither does it ought to be call’d a letter or not (…) let everyone enjoy his own opinion’ (quoted from Mugglestone 1995: 112).

Intrusive [h] in present-day English accents

Insertion of unetymological ·hÒ in medial position is not particularly frequent in Middle English texts. The Linguistic Atlas of Late Mediaeval English, henceforth LALME (McIntosh, Samuels and Benskin 1986, Vol. 4: 320), lists as the only counties and regions where medial ·hÒ-insertion is found the Isle of Ely, Norfolk, and Suffolk. The examples given in LALME are kyngdham, waxhen (see also Vol. 1, dot map 1171). This suggests that the instances of ·hÒ-insertion in these counties and regions may have the function of providing a consonantal syllable onset. It does not suggest that ·hÒ is inserted to provide a linking sound. A search in edited texts revealed some more likely cases for a linking function. Several of these are found in Furnivall’s (1882) edition of the Fifty Earliest Wills in the Court of Probate, London. The will of John Olney from Weston, Buckinghamshire (1420) has three instances of prahyng for praying (Furnivall 1882: p.47, ll. 12 and 13, and p.48, l. 20), as well as one instance of the variant spelling prehyng (Furnivall 1882: p.48, l. 10). At word boundaries ·hÒ-insertion is more frequent. John Olney’s will has his moder, hour lady (Furnivall 1882: p.47, l. 14), to hech of þes houses (Furnivall 1882: p.47 l. 20), besides yn the chapele off owr lady (Furnivall 1882: p.47, l. 17) to ech off hem (Furnivall 1882: p.48, l. 16) and to eche off hem (Furnivall 1882: p.48, l. 14). The will of Stephen Thomas from Lee, Essex (1417–1418) has six instances of howe for owe, which are preceded by I (Furnivall 1882: pp.39–40), and one instance of es Howynge (Furnivall 1882: p.41). The will of Robert Schapman from Hornsey, Middlesex (1428) has two instances of to hevery (Furnivall 1882: p.80, ll. 9 and 11) and one each of to yowre Lady for “to our Lady” (Furnivall 1882: p.80, l. 2) and of houre La[dy] for “of our Lady” (Furnivall 1882: p.80, l. 4). In the majority of these instances ·hÒ-insertion occurs at a word boundary between vowels. In particular the spelling with ·yÒ in Schapman’s will suggests that ·hÒ is inserted in the spelling where a voiced [B] was used in speech.8 William Chamberlayn’s will (Cusack 1998: 334–335), which was written in 1542 “with myne hovnd hand” (with my own hand) contains many more instances of ·hÒ-insertions, but no word-internal ones. Chamberlayn was a citizen of London at the time of his death and appears to have lived and worked

8.According to LALME (vol. 4, p.320), initial ·hÒ-insertion is widespread in the south of England, occurring in Bedfordshire, Berkshire, Cambridgeshire, Devon, Gloucestershire, Hampshire, Kent, Norfolk, Oxfordshire, Somerset, Suffolk, Surrey, Sussex, Warwickshire, Wiltshire, Worcestershire, Monmouthshire (see also vol. 1, dot Map 1172). Unfortunately, LALME does not give us any information on the phonological context of the ·hÒ-insertions. It is therefore not possible to draw any conclusion about the possible functions of the insertions.

115

116

Martina Häcker

there for most of his life, but we do not know whether he was also born there.9 The following ·hÒ-insertions occur in the text: (1) to the heye halter “to the high altar” (lines 10–11) (2) I be qveythe to hevery on “I bequeath to everyone” (line 18) (3) sent oloves in the holde Iury “Saint Olave’s [church] in Old Jewry [street]” (line 32) (4) I be qvethe to halle “I bequeath to all” [the h of halle is crossed out] (line 35) (5) to hether off them “to either of them” (line 37) (6) I have wreton thys wt myne hovnd han·dÒ10 “I have written this with my own hand” (lines 42–43) (7) I be qvethe master dobes dovtars heche off them “I bequeath to each of master Dobby’s daughters” (lines 43–44)

All of these ·hÒ-insertions occur between voiced sounds, and five out of seven between vowels. They are, however, restricted to word boundaries. In contrast, word-internal ·hÒ-insertion is one of the most striking features of a text which is almost contemporary with William Chamberlayn’s will, The diary of Henry Machyn, citizen and merchant-taylor of London, from A. D. 1550 to A. D. 1563.11 Table 1 is based on Wijk’s (1937: 219–221) list of ·hÒ-insertions supplemented by the alternative spellings listed in his index. Table 1 shows that ·hÒ-insertion occurs most frequently before -ing forms of verbs, where forms with ·hÒ by far outnumber forms without ·hÒ. Wijk (1937: 220) interprets them as hypercorrect spellings used by a speaker of an /h/-less variety of English: A further, and perhaps even more decisive proof of the loss of h in M.’s dialect is afforded by his frequent, in certain words almost regular employment of h in hiatus.

9.A note of caution should be sounded here, as Henry Machyn, who lived and worked in London, has been proved to be a Yorkshireman. For details see note 11. 10.Here the manuscript is damaged and the d is “conjecturally supplied by the editor”. 11.The geographical origin of Henry Machyn has been the subject of considerable scholarly debate. The editor of the text believed him to be of London origin, and this was the accepted view until Wijk (1937) argued for a South-Eastern Yorkshire origin on the basis of Machyn’s spelling. Britton (2000), using the fit technique, places Machyn in South West Yorkshire. The fact that the Cockney origin of Machyn was a rather persistent myth may well reflect the extent to which [h]-dropping and [h]-insertion were associated with that variety.

Intrusive [h] in present-day English accents

Table 1.·hÒ-insertion in medial position compared to alternative spellings in The diary of Henry Machyn ·hÒ-insertion at syllable boundary behyng blowhyng blohyng berehyng

23 3 26 35

carehyng conveyhyng dohyng drahyns folowhyng folohyng gohyng gahyng lyhyng

13 1 5 1 2 8 22 1 2

marehyng playhyng plahyng prahyng rohyng sayhyng

1 16 10 1 2 4

showhyng strowhyng tarehyng tornayhyng turnahyng bowhelles halohyd heher (higher) tawhear (tawer) tryhumpe wyhalles (viols) Hehonger Hehnger (High Ongar) Nuwhyngtun Nuhyngtun

1 1 1 1 1 1 1 1 1 3

hiatus at syllable boundary

bereyng

1

kareyng

1

folouyd

1

lyeng lyung

3 5

playng*

20

sayng* seyng*

1 2

halowyd

2

tryum

1

Nuwyngtun

1

elision at syllable boundary

beryng bereng caryng careng

3 1 11

gaune-wyke (going-week)

1

halod

1

3 1 4 1

*In the case of the asterisked items playng, sayng, and seyng it cannot be excluded that these spellings reflect elision in speech.

117

118

Martina Häcker

Jones (1989) argues that ·hÒ may have been inserted by Machyn for phonological reasons, to create ambisyllabicity. He states that these items show a syllable interface between verb “stem” and suffix which is characterized by the contiguity of two syllable peaks; the syllable to the left of the suffix being coda empty, the suffix itself having no initiating consonantal element. …the segment [h] is possibly best interpreted as a devocalised version of any segment with which it is contiguous and could therefore be seen … as the relatively less vocalic “filling out” element for both syllables. (Jones 1989: 270)

The insertion of such a filling-out element has not been described as a feature of any present-day British English accent to my knowledge, but it has been noted by Lass as a feature of South African English. Lass (1996: 138) gives the following description of the phenomenon: What I will call “linking [h]” is typical of the English of many Afrikaans firstlanguage speakers (and a good number of coordinate bilinguals as well); it occurs in both educated und uneducated speakers, even in some who show no trace of an Afrikaans accent. Linking [h] is mainly restricted to word-internal hiatus (“murmured”) [B] (the norm for Afrikaans /h/, and that of many mother-tongue SAE varieties).12 Some common examples: cre[h]ate, re[h]action, li[h]aison, co[h]operate, pro[h]active, pre[h]empt, bi[h]ology, pi[h]ano

Lass interprets the “murmured [B]” as Afrikaans influence. Such a murmured [B] is indeed a characteristic feature of Afrikaans (Donaldson 1993: 14), but here it does not appear to be used as a hiatus filler. Moreover, intervocalic [x] is frequently elided in Afrikaans if the preceding vowel is long or a diphthong, but neither [B] nor [‘] stop is inserted to avoid a sequence of vowels. The elision is also reflected in the spelling, e.g. Afrikaans spieël versus Dutch spiegel (Donaldson 1993: 14). This raises the question of why linking [B] is particularly frequent with Afrikaans speakers. Irrespective of the question of its origin, the existence of linking [B] in South African English shows that inserted ·hÒ in earlier English may indeed reflect [h] or [B] in speech.

12.Lanham (1967: 99) describes the realisation of /h/ in South African English as follows: “Allophones of /h/ in SAE are similar to those of RP. There is, however, virtually no glottal friction in allophones in word-initial position which are pure voiceless vocoids. In the position before /y/ (e.g. in hue /hyuu/ or /hyu™/) a palatal fricative may be heard. Between vowels (e.g. in rèhéarse /rih66s/ or /rih6™s/ with no preceding /+/) a voiced ‘h’ is often heard.” See also Wells 1982: 619.

Intrusive [h] in present-day English accents

According to Lass, linking [h] is not found across word boundaries in South African English. It does, however, occur in British English varieties, although it is a feature that goes widely unnoticed. Stanley Ellis, the main fieldworker for the Survey of English Dialects, is a native of Bradford in West Yorkshire. The Survey of English Dialects recordings are prefaced with information on the locality and the speaker. In the preface to each recording Stanley Ellis gives the name of the informant, his occupation or former occupation, and his age. In the frequently recurring sequence “farm labourer, aged” he does not use a linking [r], but regularly inserts a linking [h]. Isolated examples of [h]-insertion at word boundaries are also found in Cockney (Sivertsen 1960). One of Sivertsen’s informants, AP, who was born in 1876 and recorded between 1949 and 1956, pronounces their arm with a linking [h] instead of the linking [p] that would be used in RP (Sivertsen 1960: 249). This is the only instance that occurs in the transcripts in Sivertsen’s appendix (1960: 233–262), but the following additional examples of [h]-insertions are given in her section on /h/ (1960: 141–142): (8) Áno I Á[h]aint (9) Átime to get Á[h]up (10) Áte Á[h]aitch (the letter sequence)13 (11) (h)er Á[h]errands (12) fo(r)Á her Á[h]anyway (13) I’ll Á[h]en(d) that story (14) I Álike Á[h]eels

Unfortunately Sivertsen gives, with the exception of the last example, only phonemic transcriptions. So we know that /k/ in I like [h]eels is realised as a glottal stop, but we do not know whether the /t/ in get up is affricated, realised as a tap or as a glottal stop, all of which are frequent realisations of Cockney /t/, nor do we know whether /l/ is vocalised in I’ll [h]end that story. Despite these

13.The pronunciation [heItw] of the letter ·hÒ is extremely frequent throughout England. The reason may be that it seems illogical to many people that the sound symbolised by the letter should be absent in the pronunciation of its name. This appears to be a comparatively recent pronunciation, as Sylvester Douglas (Jones 1991: 129) states in 1779 that “The Scotch name of the h is to the ear, the same word with each. But the English form the name, of the thin slender a as in slate prefixed to the sound of ch, in each. This name (…) rhymes exactly with the first syllable of the legal word laches”.

119

120 Martina Häcker

uncertainties we can state that in the majority of these examples [h] is inserted in hiatus position at a word boundary. In addition, Sivertsen lists in a footnote ·hÒ-omissions that she noted in letters from one of her female informants: (15) freds bike as ad its chips “Fred’s bike has had its chips” (16) thats me usband “that’s my husband” (17) thats ow it tis I was gitten em “That’s how (it is) I was getting them” (18) so you ad a nice oliday with yer dad “so you had a nice holiday with your dad” (19) I wonder our Eva’s getting on “I wonder how Eva’s getting on” (20) What a terrible thing as appened in the Swiss Alps “What a terrible thing has happened in the Swiss Alps” (21) they came ome and Mrs B … had the doctor home to him “They came home and Mrs B … had the doctor (come) home to him”

The letters have ·hÒ-omission in unstressed forms of have, in the phrase me usband, and after words ending in [s,z] (that’s ow, nice oliday, as appened), and after [m] (came ome), but not after vowels (doctor home and to him). It is noteworthy that the informant retains the ·hÒ in the last sentence in doctor home in (14), but appears to drop it and use an intrusive [p] in wonder our (“how”) in (12). The different treatment may depend on stress, as both doctor and home would receive stress in natural speech, while how would be unstressed. Sivertsen’s written examples strongly suggest that Cockneys who received little education put ·hÒ in their spelling where they pronounced it and omitted ·hÒ where they did not pronounce it, as the distribution of ·hÒ in the written examples is consistent with that of [h] in the spoken ones. Unfortunately, Sivertsen does not say whether her informants inserted ·hÒ in their letters as well. The absence of comments on ·hÒ-insertion in the letters may either be due to the fact that it did not occur in the letters or to the fact that Sivertsen failed to mention it in her footnote, as she was mainly concerned with [h]-dropping and interpreted [h]-insertion as hypercorrection. According to Sweet (1907: 56) the phoneme /h/ had a voiced allophone in nineteenth- and early twentieth-century educated British English:

Intrusive [h] in present-day English accents

In most languages, when an aspirate comes between voiced sounds, it is formed with “half-voice” or imperfect vocality. Thus in English behold! compared with hold! the chords vibrate throughout the whole word, but their vibration is so feeble during the h that the contrast of this weak vocality with the full vocality of the other sounds is enough to produce the effect of aspiration. In the emphatic aha!, on the contrary, the glottis is opened enough to let out a distinct puff of air, instead of merely relaxing its closure, as in halfvoice.14 [Sweet’s emphasis]

It would seem that Cockneys have extended the use of the ‘half-voice’ allophone to hiatus position in a similar way that speakers of non-rhotic varieties of English have extended the use of linking [p] to hiatus position, thus producing intrusive [p]. The comparison with linking [p] is interesting in a further respect. According to Cruttenden/Gimson (1994: 264) intrusive [p] “before a suffix is strongly resisted”, while it is more frequent at word boundaries. The same appears to hold for linking [h] in Cockney. A further extension of the use of linking [B], comparable to the insertion of [p] in words such as drawing [drf˜pI]] or withdrawal [wIðdrf˜p6l], would lead to word-internal [h]. Such an extension could be an alternative explanation for South African linking [B], if Cockney linking [B] was exported to South Africa. Further research into eighteenth- and early nineteenth-century Cockney and South African English is required to support or refute such a hypothesis. To sum up: several twentieth-century accents of English attest the occurrence of intrusive [h, B]. The voiced variant is predominantly found between voiced sounds, i.e. in the same position where unetymological ·hÒ-insertions are found in West Saxon texts and which is also the preferred position for ·hÒinsertion in fifteenth and sixteenth-century wills that contain a high number of ·hÒ-insertions. The spelling of Sivertsen’s Cockney informants suggests that people with little education do not spell phonemically, but phonetically. It may be argued that the same holds for earlier writers. This suggests that ·hÒ-insertion may be taken at face value and interpreted as [h] or [B]. A close analysis of the contexts shows that ·hÒ-insertion does not occur randomly, but is phonetically motivated. The insertion between voiced sounds, in particular vowels, suggests

14.This pronunciation is no longer the norm in British English: “Although /h/ functions in English essentially as a voiceless syllable-initial phoneme (…) a few speakers use a voiced (or slightly voiced) allophone medially between voiced sounds, e.g. in such words as anyhow, perhaps, behind. In such pronunciations, the strong airstream of /h/ is accompanied by vocal fold vibration, the result being a kind of breathy vowel or voiced glottal fricative [B]”. (Cruttenden/Gimson 1994: 174) [author’s emphasis].

121

122 Martina Häcker

that the murmured [B] which was the norm for /h/ between vowels at the turn of the last century may have been used as a linking sound to facilitate a smooth transition between vowels not only in present-day South African English but also at different stages in the earlier history of the English language.

References Beeson, Charles Henry. 1925. A primer of medieval Latin: An anthology of prose and poetry. Chicago: Scott and Foresman. Britton, Derek. 2000. “Henry Machyn, Axel Wijk and the case of the wrong Riding: The south-west Yorkshire character of the language of Machyn’s diary.” Neuphilologische Mitteilungen 101. 571–596. Clark, Cecily. 1992. “The myth of ‘the Anglo-Norman scribe’.” History of Englishes: New methods and interpretations in Historical Linguistics ed. by Matti Rissanen et al., 117–129. Berlin & New York: Mouton de Gruyter. Cruttenden, Alan, rev., A.C. Gimson. 1994. Gimson’s pronunciation of English. London: Arnold. Cussack, Bridget. 1998. Everyday English 1500–1700. Edinburgh: Edinburgh University Press. Donaldson, Bruce C. 1993. A grammar of Afrikaans. Berlin, New York: de Gruyter. Furnivall, F. J., ed. 1882. The fifty earliest wills in the Court of Probate. (= EETS, o.s. 78) London: Oxford University Press. Häcker, Martina. 1998. “Why is there no /h/-dropping in Scots? Loss and insertion of /h/ as a contact phenomenon.” Tendenzen europäischer Linguistik: Akten des 31. Linguistischen Kolloquiums, Bern 1996 ed. by Jürgen Strässler, 71–76. Tübingen: Niemeyer. Jones, Charles. 1989. A history of English phonology by Sylvester Douglas. London, New York: Longman. ———. ed. 1991. A treatise on the provincial dialect of Scotland. Edinburgh: Edinburgh University Press. Kramer, Johannes, ed. 1976. Literarische Quellen zur Aussprache des Vulgärlateins. Meisenheim: Hain. Lanham, L. W. 1967. The pronunciation of South African English. Cape Town, Amsterdam: Balkema. Lass, Roger 1996. “Glottal stop and linking [h] in South African English: with a note on two antique connections.” Speech past and present: Studies in English dialectology in memory of Ossi Ihalainen ed. by Juhani Klemola et al., 130–151. Frankfurt am Main: Lang. Luick, Karl. 1964. [reprint of 1919–40]. Historische Grammatik der Englischen Sprache. Stuttgart: Tauchnitz. Lutz, Angelika. 1991. Phonotaktisch gesteuerte Konsonantenveränderungen in der Geschichte des Englischen. Tübingen: Niemeyer. McIntosh, Angus, M. L. Samuels & Michael Benskin, eds. 1986. A Linguistic Atlas of Late Mediaeval English. 4 vols. Aberdeen: Aberdeen University Press.



Intrusive [h] in present-day English accents

Milroy, James. 1983. “On the sociolinguistic history of /h/-dropping in English.” Current topics in historical linguistics ed. by Michael Davenport et al., 37–53. Odense: Odense University Press. Minugh, David. 1985. “As old as the ’ills.” NOWELE 6. 23–43. Mugglestone, Lynda. 1995. “Talking proper”: The rise of accent as social symbol. Oxford: Clarendon Press. Nichols, John Gough, ed. 1847. The diary of Henry Machyn, citizen and merchant-taylor of London, from A. D. 1550 to A. D. 1563. London: Camden Society, o.s. 42. Pope, Mildred K. 1934. From Latin to modern French with especial considerations of AngloNorman. Manchester: Manchester University Press. Scragg, D. G. 1970. “Initial h in Old English.” Anglia 88. 165–196. ———. 1974. A History of English spelling. Manchester: Manchester University Press; New York: Barnes and Noble. Sivertsen, Eva. 1960. Cockney phonology. Oslo: Oslo University Press. Skeat, Walter W. 1897. “The proverbs of Alfred.” Transactions of the Philological Society. 399–418. Sweet, Henry. 1907. The sounds of English: an introduction to phonetics. Oxford: Clarendon. Thomason, Sarah Grey & Terrence Kaufman. 1988. Language contact, creolization, and genetic linguistics. Berkeley, Los Angeles, London: University of California Press, Wells, J. C. 1982. Accents of English, vol. 3: Beyond the British Isles. Cambridge: Cambridge University Press. Wijk, Axel. 1937. The pronunciation of Henry Machyn, the London diarist: a study of the south-east Yorkshire dialect in the early 16th century. Uppsala: Appelberg. Wyld, H. C. 1920. A History of modern colloquial English. London: Fisher Unwin.

123



Mergers, near-mergers and phonological interpretation Raymond Hickey Essen University

1.

Introduction

The issue treated in this paper is a phenomenon which has caused linguists of very different theoretical persuasions considerable concern. The matter at hand is that of phonemic mergers and all that they entail. A merger is taken to exist when two systemic sounds which were distinguished by speakers come not to be so, i.e. they become a single sound. Mergers always involve the collapse of a phonemic distinction by one sound becoming identical with another. Later shifts will mean that the merged sounds move together. If one views mergers from the point of view of lexical sets this becomes obvious. A famous merger from the history of English is the collapse of a phonemic distinction between the vowels in the elements of the meet and meat lexical sets. The long mid front vowel which resulted from this in turn was raised to /i˜/ at a later stage of the Great Vowel Shift, i.e. the members of the meat and meet lexical sets had become indistinguishable. Change in the value of the resulting vowel meant that both lexical sets were affected. The mention of these lexical sets also illustrates a further aspect of mergers, namely that they frequently involve a residue which remains as an indication of the distinctiveness which obtained before the merger. In the case of the meat class, there are a few words, like break and great, which did not partake in the merger as their lack of the /i˜/ vowel in present-day English shows. Another general point concerning mergers, which is of central concern in this paper, is that a merger, once it has occurred, renders the recovery of the former distinctiveness impossible for later generations of language learners. This assumption requires (i) that the merger is not just phonemically, but also phonetically, complete and (ii) that language learners are not exposed to varieties of their language in which the merger has not taken place.

126 Raymond Hickey

2. Well-known examples of mergers Varieties of English provide many examples of mergers, only some of which are present in either the British or the American standard. A selection of these are presented here to indicate the range of available mergers. In this context it is essential to distinguish between those mergers which are independent of the phonetic context in which they occur and those which are not. Context independent 1. The merger of ME /7˜/ and /e˜/ to /i˜/ (meat/meet) (general southern British English) 2. The merger of ME /ai/ and /a˜/ to /ei/ (tail/tale) (general English) 3. The merger of /#/ and /f˜/ to /"(˜)/ (cot/caught) (forms of American English) 4. The merger of the square and nurse lexical sets to the nurse value (fair/ fur) (forms of Ulster English, recent Dublin English) 5. The merger of /~6/ and /f˜/ in words like poor and pour (forms of RP) 6. The merger of /v/ and /w/ to [ö] (vet/wet) (18c/early 19c southern British English) Context sensitive 1. The merger of /f˜r/ and /o˜r/ to /o˜(r)/ (morning/mourning) (most varieties of English) 2. The merger of short vowels before historic /r/ to /8˜/ or /˜/ (tern/turn) (most varieties except perhaps Scottish and Irish English) 3. The merger of /7/ and /I/ to /I/ before nasals (pen/pin) (south-west Irish English, southern American English) 4. The merger of /ei/, /e/ and /æ/, often to /7/, before /r/ (Mary/merry/marry) (to varying degrees in various forms of American English) 5. The merger of /7/ and /%/ before /r/ (merry/Murray) (Philadelphia English) With the exception of (6) above (Trudgill, Schreier, Long and Williams 2004), the mergers are all vocalic. It should be mentioned that the application of a general phonological process may lead to mergers, above all consonantal ones, but these phenomena are regarded as outside the remit of the current paper. For instance, final devoicing in German and most Slavic languages leads to a merger of voiced and voiceless obstruents in word-final position, resulting in voiceless segments. There is a difference in kind here. Final devoicing is an active process in the phonology of certain languages and applies to any segment meeting its structural description. The mergers being considered in this paper are, on the other

Mergers, near-mergers and phonological interpretation 127

hand, historical changes which took place due to shifts in the sound structure of varieties of English and which have, in the main, ceased to be active.

3. The environment for context-sensitive mergers Context-sensitive mergers are frequently determined by the presence of a following sonorant, i.e. /n, l, r/. It is interesting to consider why this subset of sounds involve mergers, as vowel mergers do not seem to be determined by obstruents, i.e. a merger of, say, /e˜/ and /i˜/ to /i˜/ before /t/ would be unusual. Before attempting an answer to the question as to why sonorants so often determine mergers, one should mention that the determining element must follow the vowel affected, i.e. it must be in the coda of the syllable of which the vowel represents the nucleus. Sonorants are most likely to have the observed determining effect in mergers because of their vowel-like quality. It is furthermore known that coda sonorants tend to become absorbed into the nucleus of the syllable they occupy, something which is widely attested in the history of English; consider the loss of consonantal /l/ in talk, walk, etc. or the more recent south-east English loss of vocalised /l/ in pronunciations like [mI~k] for milk or the development of words with historic syllable-final /-r/ in non-rhotic varieties. Another characteristic of the vocalic environment before sonorants is that it tends not to show the entire range of contrasts available in a language, e.g. /fI/ in English does not occur before /r/ and /au/ does not occur before /m/.

4. Quality and quantity of mergers Vocalic mergers can be of two types, those which involve a shift of one of the vowels in phonological space and those which involve a change in the quantity of one of the two vowels. Both kinds are attested with context-sensitive mergers in the environment of sonorants. A typical quality change before sonorants would be the raising of vowels before syllable-coda nasals, seen historically in English in think (cf. German denken) and of course in the name of the language and country itself. The merger of /7/ and /I/ to /I/ before nasals is attested synchronically in American English of the Lower South and in south-west Irish English, leading to the well-known pen – pin merger. The acoustic motivation for this is the anti-resonance between 800 and 2000 Hz (Fry 1979: 118f.), found

128 Raymond Hickey

with nasals, which has the effect of depressing the first formant of the flanking vowel, hence raising it acoustically. Quantity mergers are common before /r/. In the history of English varieties have progressively lost the length distinction for vowels before /-r/ with an attendant merger of quality, e.g. err and earn; tern and turn. Indeed a distinct short vowel quality is only maintained if the /r/ is simultaneously the onset of a further syllable, cf. very /-7.r-/ but verve /-8˜-/, /-˜-/. Other examples of merger before tautosyllabic /r/ are found with back vowels, cf. those varieties which do not distinguish words of the type horse and hoarse, i.e. show /ho˜(r)s/ rather than /hf˜rs/ and /ho˜rs/ respectively. Note that non-rhotic varieties of English never maintain this historical distinction, i.e. the absorption of /r/ into the syllable nucleus would seem to have precluded the maintenance of the distinction.

5. The reporting problem Before coming to discuss further cases of mergers, one should point out a difficulty when looking at possible cases of historical mergers. This is what I dub the reporting problem. By this is meant that the orthography is inadequate for representing many distinctions which may be crucial in distinguishing segments suspected of having merged. An illustration of this can be made by citing data for two varieties which in the orthography would look identical. For instance, one of the major distinctions between varieties of English in the north of Ireland and in the south lies in the realisation of the vowel in the face lexical set. In the south this set has a monophthong, i.e. [fe˜s], while in the north the vowel has an off-glide, i.e. [fe6s] (Milroy and Harris (1980: 203ff.) comment on a similar type of glide for Belfast in their discussion of mergers.). The orthography of English cannot make this distinction but this is crucial in keeping apart speakers from the north and the south of Ireland. To quote an example from a much better known variety of English, take the RP realisations of words like poor and paw. For some speakers who do not have the merger of these two syllable nuclei, the difference lies in the presence of a schwa at the end of first word, as a reflex of an historic /r/, i.e. [pf6] versus [pf˜]. For historical stages of English many linguists have expressed doubts about the accuracy of reports, e.g. Milroy and Harris (1980: 206–9) doubt many of the orthoepists’ comments on mergers in sixteenth-century London English. Another aspect of the reporting problem is seen when linguists are dealing with a variety of which they have had little if any direct experience. For instance,

Mergers, near-mergers and phonological interpretation 129

in supraregional varieties of English in the south of Ireland, there is a distinction between dental and alveolar points of articulation for stops in the thinker [¯t I]k] and tinker [tI]k] lexical sets. Not all non-Irish linguists are able to perceive this distinction, which all non-local southern Irish speakers are aware of. Yet another instance would be the apico-alveolar fricative which one finds, again in southern Irish English, intervocalically and word-finally in unchecked position, i.e. the sound in pity [pI§ti] and cat [kæ§t]. The literature on Irish English abounds with inaccurate descriptions of this /t/-realisation, e.g. as an /s/ sound, as an affricate, as a flap, to mention some examples. All one can say here about the reporting problem is that one should be careful about supposed mergers. In the following a particular case of reported merger from the history of English is scrutinised somewhat.

6. The intersection of trajectories in change: Is the result always a merger? The case in question is that of long low front vowels in sixteenth- and seventeenth-century London English. This issue is normally discussed with reference to the development of the vowels in words of the mate and meat lexical sets in London in the early sixteenth century (as documented in many orthoepic treatises and in verse rhyme, cf. Barber 1976: 292f.). Sets

ME outset

16th century

17th century

meet meat mate

/e˜/ /7˜/ /a˜/

[e˜] [7˜] [7˜]

[i˜] [i˜] [7˜]

(Labov 1994: 296)

In essence, the matter is as follows: the vowel in the mate set was rising from /a˜/ on its way to its later destination of /e˜/ (still later /ei/ in RP) and went through an intermediary stage /7˜/. The vowel in the meat lexical set still showed the open front vowel /7˜/ which it had since Middle English, i.e. it had not been raised to /e˜/ on the way to /i˜/ (the late nature of this shift is known from the retention of a mid vowel in many conservative varieties such as Irish English). The vowel in the mate lexical set also showed an /7˜/ by having evolved from Middle English /a˜/ through /æ˜/. If there was a merger at the point where the rising mate met the unshifted meat then after this the two should have continued on their way together,

130 Raymond Hickey

something which is known not to be the case. By the seventeenth century in London it is the words of the meat and meet classes which have merged, with the mate set distinct from both (Trask 1996: 282). There are basically two explanations of the unexpected situation of the seventeenth century. 1. In the sixteenth century mate and meat did indeed merge but were separated out by the adoption of pronunciations distinguishing between mate and meat from low prestige varieties of English in London in the course of the seventeenth century. This would make the situation an instance of the very common change from below (Trask 1996: 284). In support of this view linguists often point out additionally that such mergers are typical of looseknit networks where phonological distinctions do not have the vital function of reinforcing a closely bound community (Croft 2000: 193). 2. In the sixteenth century mate and meat did not merge entirely. The situation was one of near-merger and later this separated out to give the distinctly different pronunciations which appear in the seventeenth century and which have characterised southern British English since. For the purpose of the present discussion I have dubbed explanation (1) an external explanation, as it depends on an influence from outside the variety with the merger, and I call explanation (2) an internal explanation, as it relies completely on developments within the variety in question.

7. What is a near-merger? Turning for a moment to the internal explanation, one can consider the contentious question of near-mergers. A near merger is found where a speaker consistently makes a small articulatory difference between items of two lexical sets but cannot distinguish these auditively, specifically when the pronunciations are offered to the speaker for evaluation. It must be emphasised that the essential crux of the near-merger assumption is that speakers cannot hear the phonetic distinction which linguists tease out in a spectrographical analysis and by examining vowel formants through plotting F1 (Y-axis) against F2 (X-axis). What might such fine phonetic distinctions look like? For vowels, a slight difference in height, length, peripherality might be present, or the presence of an up-glide or in-glide after one of the vowels, something like [7I] or [76]. For consonants, a minimal distance in point of articulation, slight affrication for stops, differences of voice onset time which might slightly devoice or voice obstruents are some of the phonetic possibilities.

Mergers, near-mergers and phonological interpretation

Labov (1987: 319) states that near-mergers have important consequences for functional explanation. It appears that the communicative role of phonemic contrasts can be suspended for a considerable period of time without disrupting the integrity of the word classes and the system they participate in. There is no doubt that phonemes do function to distinguish words. But the historical development of the system of phonemes is not narrowly controlled by that communicative function.

One particular theoretical consequence of assuming the reality of near-mergers is the following: the articulatory and auditive sections of a language’s phonology must be acquired separately for near-mergers to be passed on from one generation to the next (Labov does not discuss this, except for mentioning “a considerable period of time” by which I assume that he means across several generations). This does not have to apply within a generation: after all one can imagine speakers producing closer and closer realisations of two segments and that at some stage they no longer hear the distinction although they still make one, however slight phonetically. But if the near-merger is transmitted across generations then children must acquire fine phonetic distinctions which they do not hear. However, the hearing must be the basis for picking up the distinction in the first place. The only way out of this impasse is to maintain that language learners unconsciously note fine phonetic distinctions which they are not consciously aware of and cannot distinguish later, should this be demanded of them. In the case of the mate/meat pair this would imply that throughout the sixteenth century a certain group of Londoners acquired two finely distinguished versions of /7˜/ which were then separated out in the seventeenth century without any external assistance.

8. The fate of near-mergers To be clear about Labov’s standpoint: near-mergers can separate out, i.e. an increase in phonetic distinctiveness of the two elements may occur and the two elements can become re-aligned to different phonemes in the sound inventory of speakers. But can they? There are no reported cases of this happening. However, this is what is assumed — going on the internal explanation — for the undoing of the putative mate/meat near-merger in the seventeenth century. One reason I propose for maintaining that near-mergers have never separated out into fully-fledged phonemic oppositions is that they are only

131

132

Raymond Hickey

short-lived; crucially, they seem only to occur within the lifetime of an individual. If someone makes a fine phonetic distinction between two sounds which he/she does not perceive, then this may well be the terminal stage in the loss of a distinction. One can imagine a scenario in which the distinctiveness of two sounds is being lost and that there is a differential rate for production and perception. When production of two sounds drops below a threshold of distinctiveness then human perception gives up on this and the speaker/hearer regards the two sounds as the same. To clinch the case for mate–meat being ‘just’ a near-merger in sixteenth-century London English, which was later separated out into the distinction between /e˜, ei/ and /i˜/, one would have to be able to point to at least a few other cases where it is known definitively that this has happened. But, unfortunately for the proposal of near-mergers, such cases do not seem to exist.

9. How could mergers be ‘de-merged’? 9.1 Internal motivation (i): Underlying representations In the early days of generative phonology, linguists such as Morris Halle apparently believed that sounds could have different underlying representations but be phonetically identical. Applied to the historical case considered above, this interpretation would suggest the mate class had a different underlying representation (mentally for speakers) than did the meat class, this then allowing them to unravel the merger in the following century. Halle (1962) and later Chomsky and Halle (1968) justify the assumption of deriving surface contrasts from differing underlying forms by reference to the well-known vowel alternations as in divine : divinity, serene : serenity, sane : sanity, etc., a somewhat different but related phenomenon. Apart from the ontological status of underlying representations, there is a serious flaw in this interpretation: there is no principled manner to account for what segments speakers would have retained for possible later ‘unmerging’. There does not seem to have ever been a classical generative phonologist who suggested that speakers of English retain different underlying representations of, say, the nose and knows lexical sets, just in case they might want to distinguish in a later generation between vowels deriving from ME /o˜/ and ME /ou/. In fact the assumption that speakers maintain distinct underlying representations after mergers would mean that languages become increasingly more

Mergers, near-mergers and phonological interpretation

complex by retaining historical distinctions. If one continues this line of thought, it becomes absurd: how long do speakers maintain former distinctions in their set of underlying representations? Do speakers of English maintain a distinction between /k/ and /x/ underlyingly although the latter element has long since disappeared from standard English? The answer is obviously ‘no’, but the question still needs to be addressed of how speakers decide on what distinctions to maintain for future ‘unmerging’ of mergers. For the serene : serenity cases, one could postulate that speakers maintain a different underlying form, here a long vowel in the second syllable of serenity, which is shortened by the trisyllabic shortening rule, as long as a morphological alternation links two forms synchronically. But nothing linked the mate and meat lexical sets morphologically in the history of English so there would never have been a synchronic motivation for speakers to maintain a vowel different from that on the surface. Furthermore, even assuming that one accepts the scenario of different underlying representations and surface forms, there would be no means for language learners to determine what vowel they should postulate for the underlying representation, as there was, and is, no morphological alternation to suggest this. 9.2 Internal motivation (ii): Lexical diffusion of changed pronunciations The essence of this argument is as follows: there was a merger of mate and meat in early sixteenth-century London English, but not for all tokens of each class. The residue of non-merged tokens later provided the means for undoing the merged tokens by reinstating the non-merged vowel values for the entire set. It is true that there are changes which must have progressed by lexical diffusion because residues remain, frequently in phonetic environments which are inherently resistant to the change. A classic example is the lowering and unrounding of early modern /~/ which often did not take place where the following segment was inherently rounded and/or back in articulation, hence the current pronunciations of bush and bull where a rounded alveolo-palatal fricative and a velarised lateral follow the vowel respectively. The concern here is with a scenario which might have obtained and which could have led to total merger being avoided. But for this to be credible one should be able to point to cases where the residue of lexical diffusion was turned around at some later point and became the default value in a lexical set. However, the residue scenario does not help in interpreting the mate–meat merger as the input of meat was [7˜] and the raised value of mate was also [7˜].

133

134 Raymond Hickey

Instead one might postulate a subset of the meat lexical class which had already become higher than [7˜], i.e. [7˜›]. In this situation such items would have exerted a pull effect on the other members of the lexical set, those which had [7˜] and which were represented similarly to the members of the mate lexical set in sixteenth-century London English and which are the attestational basis for the assumed merger (see detailed discussion in Labov 1994: 298ff.). Applied to the case of the mate–meat merger the scenario would look something like the following: Long low and mid front vowels in Early Modern London English Sets

ME outset

16th century

17th century

meat

/7˜/

[e˜] (later [i˜])

mate

/a˜/

majority: [7˜] subset: [7˜›] [7˜]

[7˜]

This scenario is entirely speculative, but if it could be shown, or if parallel situations could be demonstrated, then it would provide a principled manner in which an apparent merger could have been undone internally. 9.3 External motivation: Adoption of non-merged segments from another variety The standard wisdom on this view since the seminal article by Weinreich, Labov and Herzog (1968), and which can be traced to a suggestion by E. J. Dobson (1968 [1957]), is that the apparent reversal of mergers stems from the adoption of a different phonological subsystem which was available in the social environment of speakers with the merger in question. Specifically, in the case of sixteenth-century English, those varieties which did not have the mate–meat merger provided a pronunciation model later in the seventeenth century so that the merger was undone. In his discussion of the matter, Labov (1994: 302) stresses that the group in society which is most likely to have had the raising of /7˜/ and /a˜/ to the merged value /7˜/ was in the interior of the social hierarchy, neither the lower, underprivileged class nor the higher, aristocratic group. This mid-range section of London society was represented by people such as merchants, clerks, teachers, etc. This group would also have been that which would have written most and hence provided attestations for the merger under consideration here. The survival of lower prestige varieties of London English which did not have the merger would have offered a model to Londoners in the

Mergers, near-mergers and phonological interpretation

following century for the unravelling of a merger which previous generations had undergone. When one looks around for parallel cases of such unmerging then one can find them. To conclude this paper, two such instances are listed to show how such externally motivated ‘unmerging’ is attested in different contexts. 9.4 Two attested cases of ‘unmerging’ consonantal mergers 1. During the first half of the nineteenth century those speakers of southern British English who had the merger of /v/ and /w/ to [ö] in the vet and wet lexical sets undid this merger as a result of exposure to varieties which did not have it. The merger is no longer attested for British English though remnants are found in extraterritorial varieties (Trudgill et al. 2002). This merger was also attested in earlier forms of Irish English. 2. Pre-nineteenth-century Dublin English had alveolar stops in the thinker and tinker lexical sets. The high influx of migrants from the west of Ireland meant that varieties appeared in Dublin which did not have this merger (western speakers used the dental stops of Irish in the thinker lexical set). The dental stop realisation was adopted in the thinker lexical set by the middle classes in Dublin (probably as a means of dissociation from speakers of popular Dublin English which had the dental–alveolar merger, Hickey 2000). This then became part of the supraregional form of southern Irish English which stems from late 19th century/early 20th century middle-class Dublin usage.

10. Conclusion When postulating a scenario for language change in the past, maximum plausibility of the case one is presenting must be strived for. Such plausibility can be increased by pointing to other instances of the type of change under consideration and the explanation being offered. Preferably, parallels should come from the present to ensure that one can check on these. Synchronic evidence for the existence of near-mergers is slight and its assumption across generations is very speculative as it throws up questions about the transfer of near-mergers from one generation to the next which have not been addressed by those linguists who have put forward this scenario for consideration. Furthermore, no instances can be demonstrated to have existed in any concrete case.

135

136 Raymond Hickey

With reference to the question of ‘unmerging’ discussed in the latter sections of this paper, the internal explanations put forward both suffer from a lack of parallels which would strengthen their case. Neither differing underlying representation and surface forms nor the switchover from minority to majority case in a lexical diffusion scenario can be supported by other parallel instances. The external explanation on the other hand has two strong advantages. The first is that it has the support of contemporary sociolinguistic observation which often shows the adoption of pronunciations from one variety into another. The second is that the merger would appear to have held for only one section of sixteenth-century London society, a mobile middle group, a section known from contemporary sociolinguistic investigations to be particularly innovative. Given the lack of hard evidence here, the external explanation gains more credence: the body of evidence for transfer phenomena in contact situations is very considerable and the likelihood — among speakers of the same language — is increased if transfer leads to distinctions hitherto unavailable and/or if it serves an additional social function of either accommodation or dissociation among social groups, especially in urban settings.

References Barber, Charles. 1976. Early Modern English. London: André Deutsch. Croft, William. 2000. Explaining language change. An evolutionary approach. London: Longman. Chomsky, Noam & Morris Halle. 1968. The sound pattern of English. New York: Harper and Row. Dobson, E. J. 1968. English pronunciation 1500–1700. Vol.1 — Survey of the sources. Vol.2 — Phonology. 2nd edn. Oxford: Oxford University Press. Fry, Dennis. 1979. The physics of speech. Cambridge: Cambridge University Press. Halle, Morris. 1962. “Phonology in generative grammar.” Word 18: 54–72. Reprinted in The structure of language: Readings in the philosophy of language ed. by J. A. Fodor & J. Katz (1964), 344–352. Englewood-Cliffs, N. J.: Prentice-Hall. Hickey, Raymond. 2000. “Dissociation as a form of language change.” European Journal of English Studies 4:3. 303–315. Labov, William. 1987. “The overestimation of functionalism.” Functionalism in linguistics ed. by René Dirven & Vilém Fried, 311–332. Amsterdam: John Benjamins. Labov, William. 1994. Principles of linguistic change. Volume 1: Internal factors. (= Language in Society, 20). Oxford: Blackwell. Milroy, James & John Harris. 1980. “When is a merger not a merger? The MEAT/MATE problem in a present-day English vernacular.” English World-Wide 1.199–210. Newmeyer, Frederick J. 1998. Language form and language function. Cambridge, Mass.: MIT Press.



Mergers, near-mergers and phonological interpretation

———. 2003. “Formal and functional motivation for language change.” Motives for Language Change ed. by Raymond Hickey, 18–36. Cambridge: Cambridge University Press. Trask, R. L. 1996. Historical linguistics. London: Edward Arnold. Trudgill, Peter, Daniel Schreier, Daniel Long & Jeffrey P. Williams. 2004. “On the reversibility of mergers: /w/, /v/ and evidence from lesser-known Englishes.” Folia Linguistica Historica 26. Weinreich, Uriel, William Labov & Marvin I. Herzog. 1968. “Empirical foundations for a theory of language change.” Directions for historical linguistics ed. by Winfried Lehmann and Yakov Malkiel, 95–188. Austin: University of Texas Press.

137



New light on the verb “understand” Carole Hough University of Glasgow

The semantic development of the verb understand is generally considered to be obscure. An apparently transparent formation from the preposition under with the verb stand, its meaning is not obviously related to that of either of its parts. The present-day sense is already evidenced in Old English understandan, and the existence of cognates in other Germanic languages, including Old Frisian, Middle Danish, Middle Swedish, Middle Dutch, Old Icelandic, Middle High German and Middle Low German, suggests a common ancestor in the Germanic language group (Hoad 1986; OED). Previous discussions from Wood (1899) and Hempl (1899) onwards have tended to focus on the use of under as a verbal prefix, suggesting a sense “between” or “among” as the starting-point for a development to an abstract sense relating to perception.1 The course of this development remains unclear, however, and a further obstacle is that the verb forstandan also occurs in Old English with the same meaning, and again with cognates such as OS forstandan, OHG firstantan, MHG versta¯n and MDu firstantan (Hoad 1986). This appears to suggest that the semantic weight is borne not by the prefix but by the stem, an inference supported by the occurrence of other verbs whose prefixed and unprefixed forms have closely related or identical meanings, such as OE fo¯n and underfo¯n “receive”, OE gietan and undergietan “perceive”, OE niman and underniman “take” and OE stelan and forstelan “steal”.2 Attempts have also been made to relate the term to the stem OE standan “stand (upright)”, most recently by Newman (2001). Drawing attention to

1.A review of previous scholarship is provided by Newman (2001: 187–190). 2.There are contrasting instances where such prefixes substantially alter the meaning of the stem verb. A useful discussion is Ogura (1993).

140 Carole Hough

work on polysemy in the Cognitive Linguistics paradigm, and in particular to a study by Gibbs et al. (1994) which established balance, verticality and center–periphery as the major image schemas associated with the act of standing, he points out that all three give rise to metaphorical uses relating to mental acuity, such as “unbalanced”, “unstable”, “to be on top of something” and “get to the heart of the matter”. None actually corresponds to the sense of understand, however, and a potential problem with this line of argument is that the American undergraduates recruited by Gibbs et al. for their experiment may not share the same mental landscape as the Anglo-Saxons or early IndoEuropeans.3 Further down their list of image schemas associated with standing are resistance, linkage, container, enablement, attraction, blockage, full–empty, cycle and path, and it is possible that some or all of these may have been more relevant in former times. A complicating factor is the multiplicity of meanings adopted by the verb stand throughout its history. Gibbs et al. identify thirty-five different uses of stand in contemporary speech and writing, ranging from “stand at attention”, “united we stand” and “stand in someone else’s shoes” to “one-night stand”, “get stood up on a date” and “stand by your man”. Some of these meanings were already present in Old English; others have developed more recently. Conversely, some early senses have since dropped out of use. The purpose of this paper is to suggest that the compound understand is based on a meaning of the stem verb which is now obsolete but is evidenced in Old English. As is well known, many words in the semantic field of cognition have been transferred from a physical application, often reflecting well-established metaphors such as physical possession as mental possession (“he grasped the idea”), or physical vision as mental vision (“I see what you mean”). Discussed most recently by Kay (2000: 283–284), the metaphorical use of words from the field of vision to refer to cognition already appears in Old English, where for instance “clear understanding is described as beorht, hlutor (bright) or leoht”. It may therefore be relevant that one of the uses of the verb standan in Old English is in connection with light, apparently with the meaning “appear, shine”. This would offer a perfectly standard metaphorical development from physical light to mental perception, making it possible to explain the verb understand as an instance of the common metaphor “understanding is light”.

3.This is of course no criticism of the original study by Gibbs et al., which was concerned with psychological semantics rather than with historical linguistics.

New light on the verb “understand”

This sense of OE standan appears to have been overlooked in previous discussions of the verb understand, perhaps because the relevant definition is not included in standard dictionaries and reference works such as Toller (1898) and TOE.4 It is nonetheless evidenced in a number of texts, several of them poetic,5 and is sometimes though not always identified in glossaries to scholarly editions.6 Four instances occur in Beowulf. In lines 726b–727, the monster Grendel has just entered the hall of Heorot: him of eagum stod ligge gelicost leoht unfæger. (Beowulf lines 726b–727) [from his eyes shone a horrible light, most like fire.]

In line 1570b, when Beowulf has defeated Grendel’s mother in her underwater lair, standan again acts as the predicator of leoht “light”, varying lixan “glitter, gleam” in the a-verse and paralleled in the next line by scinan “shine”: Lixte se leoma, leoht inne stod, efne swa of hefene hadre scineð rodores candel. (Beowulf lines 1570–1572a) [Light gleamed, light shone within, just as the candle of the sky shines brightly from heaven.]

Line 2313b uses the verb in connection with Beowulf ’s third adversary, the firebreathing dragon: Ða se gæst ongan gledum spiwan, beorht hofu bærnan; bryneleoma stod eldum on andan. (Beowulf lines 2312–2314a) [Then the visitant began to spew forth flames, to burn the bright buildings; firelight shone, to the terror of men.]

4.Clark Hall (1960: 318) gives “appear, flash out”, but even this is incorrectly flagged as a poetic usage and cannot be taken as appropriate to a light shining constantly for a period of time, as in some of the examples cited below. 5.The corpus presented in this paper has been compiled primarily from the data in Healey and Venezky (1980), and comprises all instances which seem to me to be reasonably secure on contextual grounds. All quotations from Old English poetry are from Krapp and Dobbie (1931–1953). 6.For instance, whereas definitions of standan given by Klaeber (1950: 401) include “start, issue, arise, shine forth”, Wrenn (1973: 272) simply has “stand; issue; continue; start; go on”.

141

142 Carole Hough

Towards the end of the poem, Wiglaf is exploring the dragon’s hoard, and is able to examine the treasure by the light from a marvellous tapestry (line 2669b): Swylce he siomian geseah segn eallgylden heah ofer horde, hondwundra mæst, gelocen leoðcræftum; of ðam leoma stod, þæt he þone grundwong ongitan meahte, wræte giondwlitan. (Beowulf lines 2667–2671a) [Also he saw a golden banner hanging high over the hoard, the most marvellous handiwork, woven with skill; from it shone light, so that he could see the bottom, look through the works of art.]

Compounds of leoma again appear as the subject of standan in The Battle of Finnsburh line 35b and Christ and Satan line 127b. In the former, the verb is used of the light from flashing swords at the height of battle: Swurdleoma stod, swylce eal Finnsburuh fyrenu wære. (The Battle of Finnsburh lines 35b–36) [Sword-light gleamed, as though all Finnsburh were on fire.]

In the latter, the association with fire is literal rather than metaphorical: fyrleoma stod geond þæt atole scræf attre geblonden (Christ and Satan lines 127b–128) [Firelight shone throughout the terrible cave, mixed with venom]

A similar use occurs in Exodus line 111b in connection with the pillar of fire that guided the Israelites on their flight from Egypt: Heofonbeacen astah æfena gehwam, oðer wundor, syllic æfter sunne setlrade beheold, ofer leodwerum lige scinan, byrnende beam. Blace stodon ofer sceotendum scire leoman; (Exodus lines 107b–112) [A heavenly beacon rose each evening, a second marvel, glorious after the sun had set, to shine with fire over the people, a burning pillar. Bright rays shone clearly above the soldiers]

Another possible instance is in Dream of the Rood line 6b:

New light on the verb “understand” 143

Gimmas stodon fægere æt foldan sceatum, (Dream of the Rood lines 6b–7a) [Jewels shone/were fixed, beautiful at the surfaces of the earth]

Here the sense “stand, be fixed” is appropriate in the context of a setting for jewels, but Swanton (1970: 101) suggests that with the subject gimmas, the sense “to gleam or shine in a beam” may be the primary meaning. In view of the Rood-poet’s well known predilection for word-play, it is plausible that more than one meaning of the verb may be relevant.7 It is not quite axiomatic, however, since other texts testify to an established association with precious stones. Maxims II lines 22b–23a reads: Gim sceal on hringe standan steap and geap. (Maxims II lines 22b–23a) [A jewel must be fixed high and broad on a ring.]

Similarly the Old English Lapidary states: Oþær is saphyrus, se is sunnan gelic, & on him standað swilce gildene steorran. (Lapidary lines 5–6)8 [The second is sapphire, which is like the sun, and on it likewise golden stars are fixed.]

The occurrence in Dream of the Rood could represent another instance of this use, so it cannot be regarded as a secure example of the meaning “shine”. It would also be anomalous in that the subject is something that gives off light, rather than light itself. The verb standan otherwise seems to be used in this sense only with a subject meaning “light”. It seems to me that there are two possible ways in which a development of meaning from “stand, remain” to “shine, gleam” may have occurred. The first is suggested by the lines quoted above referring to jewels or to the sun. One of the uses of OE standan is to describe heavenly bodies fixed in orbit, as in the following extract from Ælfric’s Passion of St Sebastian: He cwæð þa ic hæbbe on minum hord-cleofan. an wundorlic weorc me to ge-wissunge. æfter steorrena gesetnyssum swa swa hi standað on heofenum. (Skeat 1881–1900: I,132)

7.Other instances of word-play in the poem are discussed in detail by Swanton (1969). 8.Quoted from Healey and Venezky (1980), s.v. standaþ.

144 Carole Hough

[Then he said: “I have in my treasury a marvellous instrument for my information, according to the positions of the stars as they are fixed in the heavens”.]

It may be possible that through its connection with heavenly bodies, the verb came to be associated with their other properties, and specifically with their light-giving attributes. Alternatively, the shift of meaning may derive from a use of the verb to denote something emanating from one place to another, identified by Toller (1898, s.v. standan) as sense XII “of direction (lit[eral] and fig[urative])”. As Swanton (1970: 101) notes, the verb is occasionally used of the issue of moisture, as in Natale Innocentium Infantum from Ælfric’s First Series of Catholic Homilies
: Him stod stincende steam of þam muþe. (Clemoes 1997: 221) [Vile-smelling breath came from his mouth.]

Similarly in the homily Alia Visio from the Second Series: of heora muðe and nosþyrlum stod stincende steam. (Godden 1979: 200–201) [Vile-smelling breath came from their mouths and nostrils.]

The Passio Sancti Bartholomei Apostoli from the First Series uses the same verb to refer to fire emanating from a source: him stod sweflan lig of þam muðe. (Clemoes 1997: 446) [Sulphurous fire came from his mouth.]

It may be significant that prose occurrences of OE standan in the sense “shine” often include a reference to direction, describing light emanating from one place to another. For instance, the annal for 1106 in the Peterborough manuscript of the Anglo-Saxon Chronicle documents the appearance of a strange star: Se steorra ætywde innon þ suð west. he wæs litel ge þuht. and deorc. ac se leoma þe him fram stod wæs swiðe beorht. 7 swilce ormæte beam ge þuht norð east scinende. (Plummer and Earle 1892–1899: I,240) [The star appeared in the south-west; it seemed to be small and dark, but the light which shone from it was very bright, and seemed like an intense beam shining north-east.]

The appearance of another star is recorded in the annal for 1110: Ðær æfter on Iunies monðe ætywde an steorra norðan eastan. 7 his leoma stod to foran him an þet suðwest. (Plummer and Earle 1892–1899: I,242) [Afterwards in the month of June a star appeared in the north-east, and its light shone in front of it to the south-west.]

New light on the verb “understand”

Both mention light shining in a particular direction, as does another reference to stars in Book 5,21 of the Old English translation of Bede’s Historia Ecclesiastica
: Stod se leoma him of, swilce fyrenþecele, óngean norðdæle middangeardes. (Miller 1890–1898: II,476) [Light shone from them, like a firebrand, towards the north of the earth.]

Other prose occurrences include the preposition (or separable prefix) up, indicating the direction of travel. Thus the prose account of St Mildred describes how: ðæs leohtes leoma stód up þurh þære healle hróf up to heofonum. (Cockayne 1864–1866: III,424)9 [a ray of light shone up through the roof of the hall up to the heavens]

A directly comparable use appears in the Liber Vitae of New Minster and Hyde Abbey: se leoma stod ymbe midde niht up þurh þære healle hrof swylce þær sunne scine10 [The light shone up through the roof of the hall at midnight, as though the sun were shining there]

Similarly, the Old English translation of Bede’s Historia Ecclesiastica, Book 3,9, records: Forðon ealle þa niht stód swylce beorht sunbeam from þæm wægne up oð heofen heah, þone mon meahte lytesne of ealre Lindesse stowum sweotole geseon. (Miller 1890–1898: I,182) [For all night a bright sunbeam shone from the wagon high up to heaven, which could clearly be seen from almost all parts of Lindsey.]

These references to light travelling from one place to another may illustrate a stage in the development of the verb standan from “emanate” through “emanate (of light)” to “shine”. This is fully plausible in the context of modern theories of semantic development. As mentioned above, one of the experiments described by Gibbs et al. (1994) attempted to relate the polysemous meanings of PDE stand to image schemas associated with the physical experience of

9.Orthography has been modernised, as Cockayne retains Old English letter forms, including wynn. 10.Quoted from Healey and Venezky (1980), s.v. stod.

145

146 Carole Hough

standing, including balance, verticality, center–periphery, resistance, attraction, path and so on. The data from the study indicated that “balance was the most relevant image schema to people’s bodily experience of standing, while path was viewed as the least relevant” (1994: 236). Bearing in mind that mental attitudes vary across time and space, the image schema path may well have been more relevant to the Anglo-Saxons and their Indo-European predecessors. The description provided by Gibbs et al. is as follows: Path refers to the experience of moving. A path has a beginning, a course along which one proceeds, and an end. (1994: 237)

This corresponds quite closely to the passages quoted above describing light beginning in one place and proceeding along a course to another, and it therefore seems possible that this image schema underlies the sense development of OE standan from “stand” to “shine”. There remains a grammatical hurdle to overcome in postulating a further development to the cognitive sense of understand, since the latter is usually a transitive verb, whereas standan (in any of its attested senses) is usually intransitive.11 Perhaps relevant in this connection is the verbal prefix under-/for- discussed towards the beginning of this paper. As has already been established, its semantic weight appears to be negligible. It may therefore be appropriate to suggest that it fulfils a grammatical function by giving the verb a transitive form.12 Finally, an explanation of understand as a transferred use of terminology from the field of vision is of course only plausible if the sense “shine” had already developed for the stem verb in Primitive Germanic. Otherwise it would not be possible to account for the existence of cognate terms in other Germanic languages. Early evidence is hard to come by due to the paucity of source material. Apart from Old Saxon and Old High German, none of the languages in question has written records prior to the ninth century.13 Early senses could well have been lost by this time. In England itself, the association with light appears to be restricted to the Early Middle Ages, surviving into Middle English only in the negative construction stonden derked “of the firmament: grow dark,

11.Exceptions include “I understand” (unless the object is taken to be deleted) and “I can’t stand it”. 12.I am grateful to Andreas Fischer and Peter Kitson for discussion of this point. 13.Both Old Saxon and Old High German have cognates of forstandan but not of understandan, perhaps suggesting that the former was the earlier type of formation.

New light on the verb “understand” 147

be darkened”, and as a copula with cler “of sight” (MED, s.v. sto¯nden v.(1) senses 14 and 29(a)). Parallels with the later development of cognates in other languages, however, do appear to support the theory of a common startingpoint. Discussing the collocation ege(sa) stod in Old English, Watanabe (1993: 28) draws attention to parallels in Old Saxon and Old Icelandic in order to argue that such constructions “show exaltation of sentiments among Germanic languages, chiefly in poetry”.14 Equally striking are correspondences between the idioms cited above to illustrate the development of OE standan from “emanate” to “shine” and expressions in current use in European languages such as Danish, Norwegian and Swedish.15 For instance, the Danish cognate stå is used in connection with heavenly bodies in expressions such as solen står op “the run rises” (Nielsen 1989: 1037, s.v. sun) and solen stod højt på himlen “the sun was high in the sky” (Vinterberg and Bodelsen 1990: 1175–1177, s.v. II stå). It also occurs in connection with fire(-light) in expressions such as der stod gnisir “the sparks flew”, stå i brand “be on fire” and stå i flammer (luer) “be ablaze, be blazing, be in flames”, and in connection with the issue of moisture or vapour in expressions such as lugten af whisky stod ham ud af halsen “his breath reeked of whisky”, der stod røg og flammer ud af maskinen “smoke and flames poured out of the engine”, fråden stod ham ud af munden “he was foaming at the mouth” and tårerne stod ham i øjnene “there were tears in his eyes” (Vinterberg and Bodelsen 1990: 1175–1177, s.v. II stå). Even more significant is the metaphorical use of the related causative verb at stille “to make stand” in expressions such as at stille nogen i et dårligt lys “to put somebody in a bad light” and at stille noget i et nyt lys “to cast new light on something”. Here perhaps is the final piece of the puzzle: a direct link between the unprefixed verb and an abstract sense relating to mental perception. In conclusion, I suggest that the verb understand derives from OE standan “shine”, in a transferred sense referring to mental rather than physical illumination. This explanation holds good for cognates in other Germanic languages as well as English. Previous discussions have attempted to explain the cognitive meaning in terms of a metaphorical development, and I suggest that it is possible to identify the metaphor in question as “understanding is light”.

14.I owe this reference to Michiko Ogura. 15.I am grateful to Graham Caie, Sara Liyja, Jacob Thaisen and other participants at ICEHL12 for information on this point. My most particular thanks are due to Jacob Thaisen, who not only provided invaluable help and advice, but also kindly supplied all the Danish material cited in this paragraph.

148 Carole Hough

References Clark Hall, J. R., with a supplement by Herbert D. Meritt. 1960. A Concise Anglo-Saxon Dictionary. 4th edn. Toronto: University of Toronto Press. Clemoes, Peter, ed. 1997. Ælfric’s Catholic Homilies. The First Series. Text. EETS SS 17. Oxford: Oxford University Press for the Early English Text Society. Cockayne, Oswald, ed. 1864–1866. Leechdoms, Wortcunning and Starcraft of Early England. 3 vols. London: Longman. Gibbs, Raymond W., Dinara A. Beitel, Michael Harrington & Paul E. Sanders. 1994. “Taking a Stand on the Meanings of Stand: Bodily Experience as Motivation for Polysemy”. Journal of Semantics 11.231–251. Godden, Malcolm, ed. 1979. Ælfric’s Catholic Homilies. The Second Series. Text. EETS SS 5. Oxford: Oxford University Press for the Early English Text Society. Healey, Antonette di Paolo & Richard L. Venezky. 1980. A Microfiche Concordance to Old English. Toronto: Pontifical Institute of Mediaeval Studies. Hempl, George. 1899. “The Semasiology of ’επσταµαι, verstehen, understand, unterstehen, gestehen, unternehmen, undertake, etc.”. Modern Language Notes 14.233–234. Hoad, T. F. 1986. The Concise Oxford Dictionary of English Etymology. Oxford: Oxford University Press. Kay, Christian J. 2000. “Metaphors We Lived By: Pathways between Old and Modern English”. Essays on Anglo-Saxon and Related Themes in Memory of Lynne Grundy ed. by Jane Roberts and Janet Nelson, 273–285. London: King’s College. Klaeber, Fr., ed. 1950. Beowulf and the Fight at Finnsburg. 3rd edn. Lexington, Mass: D. C. Heath & Co. Krapp, George Philip & Elliott van Kirk Dobbie, eds. 1931–1953. The Anglo-Saxon Poetic Records. 6 vols. London and New York: Routledge and Columbia University Press. MED
: Kurath, Hans et al., eds. 1952–2001. Middle English Dictionary. Ann Arbor: University of Michigan Press. Miller, Thomas, ed. 1890–1898. The Old English Version of Bede’s Ecclesiastical History of the English People. 4 vols. EETS 95, 96, 110, 111. London: Early English Text Society. Newman, John. 2001. “How to Understand understand”. Neuphilologische Mitteilungen 102.185–199. Nielsen, B. Kjærulff. 1989. Engelsk-dansk Ordbog. 3rd edn. Copenhagen: Gyldendal. OED
: Oxford English Dictionary. 1989. 2nd edn. Oxford: Clarendon Press. Ogura, Michiko. 1993. “Verbs Prefixed with ofer- and under- in OE and ME”. Chiba Review 15.19–49. Plummer, Charles & John Earle. 1892–1899. Two of the Saxon Chronicles Parallel. 2 vols. Oxford: Clarendon Press. Skeat, Walter W., ed. 1881–1900. Ælfric’s Lives of Saints. 4 vols. EETS 76, 82, 94, 114. London: Early English Text Society. Swanton, M. J. 1969. “Ambiguity and Anticipation in ‘The Dream of the Rood’”. Neuphilologische Mitteilungen 70.407–425. Swanton, Michael, ed. 1970. The Dream of the Rood. Manchester: Manchester University Press.



New light on the verb “understand” 149

TOE
: Roberts, Jane & Christian Kay with Lynne Grundy. 1995. A Thesaurus of Old English. 2 vols. London: King’s College. Toller, T. Northcote. 1898. An Anglo-Saxon Dictionary Based on the Manuscript Collections of the Late Joseph Bosworth. Oxford: Oxford University Press. Supplement by T. Northcote Toller (1921); Addenda by Alistair Campbell (1972). Vinterberg, Hermann & C. A. Bodelsen. 1990. Dansk-engelsk Ordbog. 3rd edn, ed. Viggo Hjørnager Pedersen. Copenhagen: Gyldendal. Watanabe, Hideki. 1993. “Some Neglected Aspects of Meaning of the Old English NounVerb Combination egesa stod”. Studies in Medieval English Language and Literature 8.25–37. Wood, Francis A. 1899. “Understand, Guess, Think, Mean, Semasiologically Explained”. Modern Language Notes 14.129–131. Wrenn, C. L., ed. 1973. Beowulf: With the Finnesburg Fragment. 3rd edn. rev. by W. F. Bolton. London: Harrap.



Homophones and the stabilization of orthography in nineteenth-century and early twentieth-century English Susan Kermas University of Lecce, Italy

1.

Introduction

Normally, when we think of homophones, we think of words that sound alike, words of different origins, which have become homophones as a result of sound change. However, since the emergence of two orthographic systems — the slightly more phonemic American system, on the one hand, and the more etymologically-oriented British system, on the other — more and more pairs of homophones result from the attribution of variant orthographic forms to extended meanings rather than from a shift in pronunciation of words of diverse derivation. It is normal for lexis to expand and evolve in order to meet the demands of an ever-changing society. What emerges, however, from my analysis of homophones of relatively recent coinage is that lexicographers are aiming at the same strict delimitation of meaning and mono-referentiality of lexis that scientists such as Boyle sought to attain in the seventeenth century (Gotti 2002: 66). Though it is undeniable that spelling has become increasingly phonemic on both sides of the Atlantic in the past two and a half centuries, since the publication of Webster’s dictionary (1828) the creation of new pairs of homophones such as check/cheque, draft/draught, ensure/insure and curb/kerb reveals not only an ever-increasing awareness of the advantages of graphemic distinction of words in cases of polysemy but also the inadequacy of the term homophone itself. Indeed, in spite of general consensus around Hobbs’s early definition of homophone as “one of two or more words pronounced the same but different in spelling and meaning” (1936: 1), the specification that “it is needful to state

152

Susan Kermas

that homophonous words must be different words, else we should include a whole class of words which are not true homophones” (Bridges 1919: 3) seems to explain the weakness of the term. There is evidence that spelling reformers such as Bullokar and Gill already recognized the desirability of a graphic distinction of homophones (Venezky 1970: 31). Hume (1617), who noted the importance of retaining the original c, s and sc graphemes in words of Latin origin with /s/ phoneme in order to distinguish “words of one sound and diverse signification” (Venezky 1970: 20), certainly encouraged the creation of an endless number of homophones such as ascent/assent, cell/sell, council/counsel and cite/site/sight, based on etymological differentiation. More importantly, “during the seventeenth century some spelling variants came to be associated, often according to the direction of spelling book compilers, with distinct fields of reference of what is etymologically a single word” (Scragg 1994: 78). As a result, eighteenth-century lexicographers gradually began to adopt variant orthographic forms in their dictionaries as a means of distinguishing one sense of a word from another. Cruden first recorded the distinction between flour and flower, for instance, in his Concordance (1738), and though Johnson did not include this particular differentiation, he anticipated other modern distinctions such as metal and mettle in his dictionary (1755; cf. Potter 1990: 72).1 If there seems to be no difficulty in including these pairs in lists of homophones today, it is increasingly difficult to establish when pairs such as check/ cheque, draft/draught, ensure/insure and curb/kerb — all excluded from Bridges’ list of homophones (1919) — actually become homophones. It is the awareness of the increase of pairs of homophones resulting from diverging meanings of the same word rather than converging pronunciation of different words that has brought to light the unacceptability of Bridges’ rigid stance.2 It is for this reason, too, that the gradual lemmatization of variant orthographic forms of draught, check and ensure, which all have extended

1.According to Scragg, “Dr. Johnson was being old-fashioned in retaining both meanings under the heading flower in his dictionary of 1755” (1974: 78). In effect, according to OED citations at flower, Milton is the last example of the French spelling in reference to ‘flowering plants’ but at flour, there is a citation illustrating flower spelling as late as 1809. 2.Kay and Wotherspoon remark on the fact that “the desire to distinguish homophones may still be a factor in certain newer developments. In Australian English, for example, we find wrap, meaning ‘praise’, ‘to praise’. The Australian National Dictionary lists this meaning under rap, although all but one quotation are spelled with ·wrÒ, perhaps to distinguish it from rap with the opposite meaning, as in ‘to take the rap’” (2002: 136).

Homophones and the stabilization of orthography in English

commercial meanings, will be traced in this paper, in order to illustrate that once a homophonic variant form gains lemma status, it is impossible to deny that it is also a homophone. There is reason to believe that variant forms with extended technical meanings tend to lemmatize more readily than those with less specialized connotations; therefore, a comparison with the emergence of curb/kerb is deemed necessary.

2. Obsolescence of pre-OED dictionaries In Early Modern English, the existence of a dual standard of spelling was normal and Johnson himself adopted one form in his private correspondence, another in his dictionary (Osselton 1984: 123), but during the nineteenth century most orthographic variant forms disappeared. If we examine the Oxford English Dictionary (OED) citations for the obsolete meaning of cheque (counterfoil of a bank bill, draft, etc.), we will note that both Johnson (1755) and Barclay (1774–82) have check as headwords, whilst both citations based on real usage (1706 and 1708) have Frenchified forms (checque and cheque, respectively). Likewise, in the modern banking sense of the word, four of the eight citations based on real usage between 1774 and 1886 illustrate the renewed popularity of the French forms, and the remaining four, which have the so-called ‘correct’ check form, illustrate the gradual influence of dictionaries on usage. In the 1850s, there is greater awareness that dictionaries should no longer base their evidence on tried-and-tested classics. Rather they should be looking at the flourishing newspaper press which better represented the spoken and written idiom of the mid-nineteenth century (cf. Green 1996: 445). Lexicographers such as Clarke (1855) and Craig (1858) — both interested in real usage — include cheque as a cross-reference in their dictionaries and Boag (1852) even gives it main headword status. Similar findings are drawn from examination of entries for the banking sense of draft. In the 1820s, dictionaries keep to the traditionally ‘correct’ form (draught) in spite of evidence of widespread use of the phonetic form since the last two decades of the eighteenth century, and though Todd (1839) already anticipates the modern distinction by over ten years, it is only in the 1850s that dictionaries such as those by Boag (1852), Nuttall (1856) and Craig (1858) give

153

154

Susan Kermas

precedence to draft in the banking sense.3 The fact that Clarke (1855) includes both forms without specifying any distinction probably reflects real confusion stemming from polysemous overload and widespread belief that draft is “corrupted from draught” (Boag 1852: 433). The next pair of words to be examined is slightly different in that the form insure is properly a mere variant of ensure, and still occasionally appears in all the surviving senses according to the OED.4 In general usage, however, it is now limited to the financial sense with reference to ‘insurance’ of life or property, in which the form ensure is completely obsolete. Citations in the OED give preference to insure in the financial sense since the seventeenth century, there being only two citations for ensure in the commercial sense between 1693 and 1747 at ensure and one at insure 1635. All the others have insure in the commercial sense from 1663 to 1883. In spite of this obvious tendency to use insure in the financial sense and ensure in a more general sense since the mid-eighteenth-century, there is a tendency for early nineteenth-century dictionaries not to differentiate the terms. Indeed, though most dictionaries in the 1850s have both forms as headwords, the modern distinction of usage is first made by Cooley (1861), and dictionaries such as Nuttall’s continue to exclude ensure on the basis of its obsolescence as late as 1856. Likewise, mid-nineteenth-century dictionaries give curb(-stone) headword status notwithstanding the fact that none of the OED citations for this period have this form. Though kerb(-stone) is a cross-reference in most dictionaries, it is never considered to be the main form and Craig (1858) even suggests it is a form used in London only. In short, though the dichotomy between real usage and lemma status is more evident in the ensure/insure and curb/kerb pairs, there can be no doubt that dictionaries seriously lag behind real usage in all four pairs examined. In this case study alone, both the content and layout of different editions of dictionaries are often similar. Entries for check, draught, draft, ensure, insure and curb are identical in both the 1835 and 1863 editions of Richardson’s dictionary, and Clarke makes no amendments in his 1865 edition, in spite of a change of printers. There is also evidence of plagiarism as there is no difference between

3.The last OED citation with draught in the banking sense is 1790. All the other citations between 1786 and 1861 have draft (OED, at draught and draft). 4.In corroboration of this, several instances of insure in the general sense of the word occur in a random selection of the lexeme in the BNC.

Homophones and the stabilization of orthography in English

Boag’s (1852) and Craig’s (1858) treatment of ensure and insure, and in the case of draught and draft entries, there are only minor differences in punctuation and capital letters. Technically, at this stage, it is impossible to define our lexical pairs as homophones simply because lexicographers have not officially sanctioned the use of the variant forms, and what is all the more disconcerting is that the lemmatization of such forms is unlikely in such adverse circumstances because there is no attempt to update dictionaries.

3. Search for authority Though there is evidence to support the view that there is greater awareness of usage in the 1880s and 1890s, and editors such as Findlater (1882) specify that check is usually written cheque, there is, nonetheless, much concern for ‘correct’ usage. Skeat, for instance, considers cheque “a pedantic spelling of check, from confusion with exchequer” (1885: 77) and explains that “it is really a name given to a draft for money, of which one keeps a memorandum or counter-check” (ibid.). Ogilvie (1898), whose original source was Webster, has check as a headword and cheque as a mere cross-reference. Clearly, it became more and more evident that “what all such dictionaries required was a single authority upon which they could draw” (Green 1996:445). Johnson — the English authority — was obsolete and Webster was American. The New English Dictionary (after 1895 the Oxford English Dictionary) was still in progress. In the interim, however, Murray, the OED editor, prepared two dictionaries for Routledge in 1890 and 1892, in which cheque has complete lemma status, and together with Bradley produced a revised edition of Hart’s Rules for Compositors and Readers (1897), which also gives absolute recognition of the French spelling. In a similar way, draft only gains momentum in the 1890s. Indeed, though Findlater admits that draft is the spelling in the banking sense, he considers it a “corruption of Draught” (1882: 139). Skeat does no more than admit that “draft is a phonetic spelling” (1885: 124), and makes no distinction of spelling according to meanings. Though Murray includes neither draft nor draught in the revised edition of Hart’s Rules for Compositors (1897), he does give draft headword status in the

155

156 Susan Kermas

sense of “an order for money” in both Routledge dictionaries. In The Imperial Dictionary of the English Language there is an important claim that the two variant orthographic forms serve as an illustration of “a principle in language, namely, that when, through considerable variation in spelling, forms originally identical appear as different words, different shades of meaning are assigned to each” (Ogilvie 1898: 94); and cheque and check are given as further examples. It is all the more evident that lexicographers were in need of a British authority in the 1880s as regards insure. Indeed, though Findlater includes both ensure and insure in his dictionary and allows both senses in the insure entry, Skeat is completely oblivious to real usage and includes ensure only, as dictionaries such as Todd’s (1839) had in the 1830s. Murray also continues to include both meanings in his insure entry and to treat ensure as a cross reference only in the Routledge dictionaries. It is only in Hart’s Rules for Compositors (1897) that he finally specifies that ensure means ‘make safe’ and insure is used ‘as in a society’. The detailed information about usage given in The Imperial Dictionary of the English Language (Ogilvie 1898) is a breakthrough. It not only states that ensure and insure “now present an example of differentiation of form when a new idea is developed rendering such distinction desirable” (1898: 171) but also gives information about assure in reference to ‘life insurance’. Little progress is made as regards the differentiation of curb and kerb. In all dictionaries, kerb continues to be treated as a variant form, and neither form is included in Hart’s Rules for Compositors and Readers (1897). Clearly, with the exception of curb and kerb, there is greater recognition of semantic differentiation of all pairs of words examined; however, it is still premature to define them as pairs of homophones without the general consensus of all dictionaries.

4. American vs. British usage Later, at the turn of the century, very few significant changes occur. The various editions of Hart’s Rules for Compositors continue to give cheque as the correct form, whilst some dictionaries — especially those based on historical principle — give both forms with the specification that cheque is the more usual form. However, what does emerge in the first decade of the twentieth century is an ever-increasing nationalism. Indeed, according to the preface to Nuttall’s Standard Dictionary of the English Language, “in the Orthography […] the

Homophones and the stabilization of orthography in English

innovations of the Transatlantic school have been entirely rejected, and the recognised authorities of our own country in the matter followed as more congenial to English taste and English feeling” (Wood 1906: viii). Statements such as “cheque, though merely a variant of check, is in British usage clearly and usefully differentiated from it with the sense bank-draft, check being chiefly American in this sense” (Fowler 1926: 75) become more and more frequent.5 In effect, dictionaries no longer focus uniquely on graphemic differentiation as a means of distinguishing two meanings but also as a means of separating British from American usage. Though dictionaries based on historical principle continue to give both forms, cheque is always given precedence, and from the 1930s, the specification that check is American is also added. Similarly, preference continues to be given to the phonetic form of draft in the banking sense in the first two decades of the twentieth century but specification of American usage as regards draft occurs much later than for check. No doubt this is partly due to semantic overload of draught and draft
; however, American usage itself was slow to stabilize. Indeed, in both the Thorndike Century Senior Dictionary (1941) and Funk & Wagnall’s New ‘Standard’ Dictionary of the English Language (1949), they are treated as variant forms and it is not until the late 1950s that American dictionaries treat draft as the main form for all meanings of the word.6 As a consequence, American usage labels for draft begin to appear in British dictionaries only thirty years after the same specification for check (e.g. Geddie 1959). In the case of ensure and insure, orthographic overlap is all the more marked and during the first two decades of the twentieth century, most dictionaries treat them as alternative forms. Though there is general consensus that insure is the only form possible in the commercial sense, only Hart’s Rules for Compositors

5.In effect, at the beginning of the twentieth century there was a fashion for American accountants to adopt cheque “as a convenient means of distinguishing between a bank check (to which it is applied) and check in the sense of verification” (Mencken 1936: 391–392), and printers’ stylesheets often included this form, especially for books intended for circulation in England as well as America. However, Mencken was of the opinion that these works should be revised on the basis that “American spelling is plainly better than English spelling, and in the long run it seems sure to prevail” (ibid.: 393). 6.Indeed, this reflects a real decline in usage in American English and in confirmation of this, whilst the Brown Corpus of American English (1961) has only one instance of draught, the Frown Corpus (1991) has none. LOB and FLOB — the corresponding corpora of British English — also reveal a downward trend from twenty-two to nine instances; however, draught continues to be used in traditional contexts of draught cider and so on.

157

158

Susan Kermas

and Readers and The Concise Oxford Dictionary of Current English opt for ensure in the general sense, in 1897 and 1911 respectively. What makes the distinction between British and American usage different as regards insure and ensure is that whilst this distinction is well consolidated in the 1920s and 1930s in British usage, both forms are still used indifferently in American English as late as the 1940s. Interestingly, in Webster’s New International Dictionary (1957), ensure is treated as the better literary term in the sense of ‘make sure’ and in The Random House Dictionary of the English Language (1966) it is given headword status. There is indeed evidence to support the view that ensure is more widespread in American usage than most British lexicographers admit. Though the Longman Dictionary of the English Language (1984) specifies that insure is chiefly North American for ‘to ensure’, according to Manser (1988), the word insure is only sometimes used in place of ensure in American English. Indeed, a comparison between the Brown (1961) and Frown (1991) corpora of American English confirms a marked upward trend of ensure from eight to thirty-seven instances in the latter corpus. Dictionaries such as the Cambridge International Dictionary of English (Procter 1995) do indeed indicate that in American usage insure is only a variant form in the general sense of the word. It is difficult to establish why so little attention is paid to curb and kerb in British dictionaries, and why curb should still be considered the main form until the publication of the Oxford English Dictionary (1933), other than that they are not of commercial interest. American dictionaries begin to specify that kerb is a Britishism in the 1940s and it is only in the 1980s — at a time when British English is submerged with computer programs and disks — that British dictionaries at last specify that curb is American and that it should not, therefore, be used in British English.7

5. Concluding remarks Today it is generally agreed that the pairs of words examined in this paper are homophones, though there is some reservation as regards ensure and insure. Carney includes check/cheque, draft/draught and curb/kerb in his lists of pairs of homophones resulting “from variant spellings of what was once the same word”

7.Of the dictionaries examined in this case study, Thorndike (1941) is the first to state that kerb is British. Funk & Wagnall’s (1949) also includes kerb and kerbstone but does not specify that they are British.

Homophones and the stabilization of orthography in English 159

(1997: 38); however, he does not explicitly state that the term homophone is strictly applicable to ensure/insure in spite of the obvious tendency to associate insure with the financial sense and ensure with the general sense of ‘make sure’. It is indeed this particular pair of lexemes that draws our attention to the fact that words of the same derivation — no matter how widespread the distinction may be — cannot be classed as homophones if they have not gained full lemma status. In the case of ensure/insure, insure — the alternative form of ensure — is occasionally found in real usage in British English, and the OED, which has gained the authority that lexicographers sought at the end of the nineteenth century, also treats it as an occasional alternative form of ensure in all surviving senses. Indeed, though there is general agreement that there is no need to exclude words of the same derivation from lists of homophones — as Bridges did in 1919 — there is not the same consensus about the applicability of the term homophone to variant orthographic forms that have not completely lemmatized. Today the average definition of homophone focuses more on the sound than on the origin, and definitions such as “words which sound alike but are written differently and often have different meanings” (Richards et al. 1992: 168) are commonplace. The fact that Bridges includes pairs such as story/storey illustrates the fallibility of his definition because there is no general agreement as to “whether these names for the floor and the tale are etymologically the same word or not” (Fowler 1959, at storey), and also in view of the fact that story is still considered the correct form for ‘floor of a house’ in printers’ dictionaries in the 1950s (cf. Hart 1952: 23), and the differentiation of the spellings according to different meanings is still a probationer as late as 1959 (cf. Fowler 1959, at storey). What does appear to be particularly important today is that pairs of homophones resulting from the diverging of British and American forms can only be defined as such once the variant forms have lemmatized to such an extent that there is no orthographic overlap between the two varieties of English. Indeed, since the 1980s, dictionaries on both sides of the Atlantic are adamant about the forms to be used and it does appear that — within the British context at least — lexicographers are so concerned about what Trudgill calls the Americanisation catastrophe — the “process of cultural and linguistic homogenisation, brought about mainly by the influence of the electronic media” (Trudgill 2002: 147) — that etymologizing is used as a means of counter-balancing this trend. It is no doubt this process that has speeded up the stabilization of orthography and given more pairs of homophones in British English on the one hand, and obscured the term homophone on the other. However, though books on

160 Susan Kermas

correct usage on both sides of the Atlantic specify that storey/story, check/cheque, draft/draught and curb/kerb are differentiated according to meaning in British English only (cf. Wilson 1993; Howard 1994), they are not quite so adamant about ensure/insure.8 It is, in conclusion, this particular pair of words that makes us aware of the intrinsic link between homophone and lemma status and makes us realize that if one of the lexical items in a pair of homophones can in any way be treated as a variant form in either American or British English, then the term homophone loses its validity.

References Boag, John. 1852. The Imperial Lexicon of the English Language. Edinburgh & London: A. Fullerton. Bridges, Robert. 1919. “English Homophones”. Society of Pure English, Tract II. Oxford: Clarendon Press. British National Corpus (BNC). Oxford: Oxford University Computing Service (http:// info.ox.ac.uk/bnc) Brown Corpus. 1961. International Computer Archive of Medieval and Modern English (ICAME): http://www.hit.uib.no/icame/cd Carney, Edward. 1997. English Spelling. London & New York: Routledge. Clarke, Hyde. 1855. A New and Comprehensive Dictionary of the English Language, as Spoken and Written. London: John Weale. ———. 1865. A New and Comprehensive Dictionary of the English Language. London: Virtue Brothers. Collins, Howard F. 1933 [7th edn]. Authors and Printers’ Dictionary revised by Humphrey Milford. London: Humphrey Milford. The Concise Oxford Dictionary of Current English. 1911, ed. by Henry Watson Fowler & F. G. Fowler. Oxford: Clarendon Press. Cooley, Arnold J. 1861. A Dictionary of the English Language. London: W. & R. Chambers. Craig, John. 1858. A New Universal Etymological, Technological, and Pronouncing Dictionary of the English Language. London & New York: Routledge. Findlater, Andrew, ed. 1882. Chamber’s Etymological Dictionary of the English Language. London: Chambers. FLOB Corpus. 1991. International Computer Archive of Medieval and Modern English (ICAME): http://www.hit.uib.no/icame/cd Fowler, Henry Watson. 1926. A Dictionary of Modern English Usage. Oxford: Oxford University Press.

8.According to Wilson, “Americans frequently prefer insure when it comes actually to writing or buying insurance policies, whereas both insure and ensure have the generic sense of making certain of outcomes” (1993: 251).

Homophones and the stabilization of orthography in English

———. 1959. A Dictionary of Modern English Usage. Oxford: Clarendon Press. Frown Corpus. 1991. International Computer Archive of Medieval and Modern English (ICAME): http://www.hit.uib.no/icame/cd Funk, Isaac K., ed. 1949. Funk & Wagnall’s New ‘Standard’ Dictionary of the English Language. New York: Funk & Wagnall. Geddie, William, ed. 1959. Chamber’s Twentieth Century Dictionary. London: Chambers. Gotti, Maurizio. 2002. “The Development of English as a Language for Specialized Purposes.” Domain-specific English ed. by Giuseppina Cortese & Philip Riley, 65–86. Bern: Peter Lang. Green, Jonathon. 1996. Chasing the Sun — Dictionary Makers and the Dictionaries They Made. New York: Henry Holt. Hart, Horace. 1897. Rules for Compositors and Readers revised by James A. H. Murray & Henry Bradley. Oxford: Clarendon Press. ———. 1952. Rules for Compositors and Readers revised by Geoffrey Cumberlege. Oxford: Oxford University Press. Hobbs, James B. 1936. Homophones and Homographs — An American Dictionary. Jefferson: McFarland. Howard, Godfrey. 1994. The Good English Guide — English Usage in the 1990s. London: MacMillan. Kay, Christian J. & Irené Wotherspoon. 2002. “Wreak, wrack, rack, and (w)ruin.” Sounds, Words, Texts and Change. Selected Papers from 11ICEHL, Santiago de Compostela, 7–11 September 2000 ed. by Teresa Fanego, B. Méndez-Naya & E. Seoane, 129–143. Amsterdam & Philadelphia: John Benjamins. LOB Corpus. 1961. International Computer Archive of Medieval and Modern English (ICAME): http://www.hit.uib.no/icame/cd Longman Dictionary of the English Language. 1984, ed. by Randolph Quirk. Harlow: Longman. Manser, Martin H. 1988. Bloomsbury Good Word Guide — Spelling, Punctuation, Pronunciation, Grammar, Jargon and Buzz Words. London: Bloomsbury. Mencken, H. L. 1936 [4th edn]. The American Language — An Inquiry into the Development of English in the United States. New York: Alfred A. Knopf. Murray, James Henry. 1890. Routledge’s Illustrated Dictionary. London: George Routledge. ———. 1892. A Companion Dictionary of the English Language. London: George Routledge. Nuttall, Peter Austin. 1856. Webster’s Pronouncing Dictionary of the English Language, critically revised, and adapted to the present state of English literature. London: George Routledge. Ogilvie, John. 1898. The Imperial Dictionary of the English Language revised by Charles Annandale. London: Blackie. Osselton, Noel E. 1984. “Informal Spelling Systems in Early Modern English: 1500–1800.” English Historical Linguistics: Studies in Development ed. by N. F. Blake & Charles Jones, 123–137. Sheffield: University of Sheffield. The Oxford English Dictionary. 1933, ed. by James Henry Murray et al. Oxford: Oxford University Press. Oxford English Dictionary Online (OED). 1989 [2nd edn.] ed. by John Simpson & Edmund Weiner. Oxford: Oxford University Press www.oed.com Potter, Simeon. 1990. Our Language. London: Penguin.

161



162 Susan Kermas

Procter, Paul, ed. 1995. Cambridge International Dictionary of English. Cambridge: Cambridge University Press. The Random House Dictionary of the English Language. 1966, ed. by J. Stein. New York: Random House. Richards, Jack C. et al. 1992 [2nd edn] Dictionary of Language Teaching and Applied Linguistics. Harlow: Longman. Richardson, Charles. 1835. A New Dictionary of the English Language. London: William Pickering. ———. 1863. A New Dictionary of the English Language. London: Bell & Daldy. Scragg, Donald G. 1994. A History of English Spelling. Manchester: Manchester University Press. Skeat, Walter. 1885. A Concise Etymological Dictionary of the English Language. Oxford: Clarendon Press. Thorndike, Edward L. 1941. Thorndike Century Senior Dictionary. Chicago: Scott & Foresman. Todd, Henry, ed. 1827/1839. Johnson’s Dictionary. Boston: Jenks & Palmer. Trudgill, Peter. 2002. Sociolinguistic Variation and Change. Edinburgh: Edinburgh University Press. Venezky, Richard L. 1970. The Structure of English Orthography. The Hague: Mouton. Webster’s New International Dictionary. 1957, ed. by Philip B. Gove. Springfield, Mass.: Merriam. Wilson, Kenneth G. 1993. The Columbia Guide to Standard American English. New York: Columbia University Press. Wood, James, ed. 1906. Nuttall’s Standard Dictionary of the English Language. London: Frederick Warne.



Kailyard, conservatism and Scots in the Statistical Accounts of Scotland Robert McColl Millar University of Aberdeen

1.

Introduction

From the mid-18th century on, Scotland was faced with concerted changes to its economic and social structure. An agrarian country was transformed into one where the majority lived in urban centres. One element of these changes was the status and treatment of the autochthonous languages of Scotland: Scots and Gaelic. Whilst the latter’s integrity as a separate linguistic system was never in question, industrial development and agrarian “improvement” led to the disruption of its linguistic population base. With Scots, the situation was more complex. A largely independent written language in the 16th and early 17th centuries had been almost wholly subsumed into English by the mid 18th century. There is every reason to believe, however, that the great majority of non-Gaelic Scots continued to speak the language until the changes brought about by the development of an urbanised and aspirant middle class in the early stages of industrialisation. As I have shown elsewhere (Millar 2000; Millar 2003), these changes were initially popular among a significant part of the opinion-forming classes of Scotland — a great majority of the correspondents to the first Statistical Account of Scotland, published in the 1790s. Often the Scots tongue was used by these informants to show disdain for the “old ways”. As time passed, however, it became increasingly evident that at least the short-term human cost of these changes would be immeasurable. In the Lowlands the combination of subsistence agriculture, near total employment, and a system of poor relief based upon a population forced to take ‘handouts’ from the contributions of their ‘betters’ only in direst necessity

164 Robert McColl Millar

nearly worked; the dislocation caused by improvement meant that many were compelled to offer their labour in new industries. The voluntary system of social welfare (and also social control) broke down. The influx of immigrants into the new industrial centres from the Highlands and Ireland exacerbated the problem. Often not speakers of English (never mind the local vernacular), some of the Highlanders and the majority of the Irish were even more threatening to the radical Protestant lowlanders in their Catholicism. New, urban, forms of both bigotry and poverty developed (McCaffery 1998). During the middle years of the 19th century, the population of Scotland’s urban centres grew exponentially. The growing urban middle classes, overwhelmingly Protestant — often dissenting — in confession, began to make their presence felt (MacLaren 1989). At first, this newly enfranchised population were enthusiastic in their support for social and economic change; many remained so. Nevertheless, a hankering after a simpler time when the majority of the population lived in rural communities where social distinctions were not divisive and where the poor were God-fearing, couthie (Robinson 1985: s.v. couth … ~ie … adj. … of persons agreeable, sociable, friendly, sympathetic) and pawkie (Robinson 1985: s.v. pawk … ~ie 1 wily, crafty; shrewd, astute; stubborn … 2 having a matter of fact, humorously critical outlook on life, characterized by a sly, quiet wit…) but knew their place, began to grow as the full implications of industrialisation became apparent. This sentiment expanded as workers began to organise politically during the 19th century. This romanticisation of a rural conservative utopia is often termed kailyard in Scotland. Strictly speaking, the kailyard refers only to a literary school of largely Scottish authors writing towards the end of the 19th century. A ‘classic’ kailyard novel, as discussed by Blake 1981, Campbell 1981 and Knowles 1983, would include many of the features already detailed for the conservative dream. A central element of these novels is the use of Scots to show ‘pawkiness’ among rural ‘worthies’. By extension, kailyard has come to mean any attempt to equate Scottish identity with a view of the past (and present) which is inherently conservative and can be quite mawkish at times — the most famous example being the musical film Brigadoon. As has already been mentioned, Scots was thrust underground as a literary language from the 17th century on. It naturally remained as the spoken language of the great majority of the (non-Gaelic) population. Throughout the succeeding period, however, Scots continued to be used, not merely for literary purposes, but also in (often radical) print journalism, as Donaldson (1986) points out. Part of the language’s appeal under these circumstances is its

Kailyard, conservatism and Scots in the Statistical Accounts of Scotland 165

association with ‘genuine’ sentiment — even with truth. It was the language of the ‘common man’. Elements of these partially competing traditions — the rural, conservative, ‘pawkie’ tradition and the largely urban, radical, ‘pawkie’ tradition — are to be found in the returns to the three Statistical Accounts, and in particular in the New Account and the Third Account. The conservative element would doubtless have been encouraged by the radical changes in Scottish society during the course of the 20th century, such as the growth of active socialism and political nationalism and the general move away from ‘traditional’ values. In this essay I will analyse discussion of language use and in particular examples of the use of Scots in these returns within the framework of the discourse which contains them, and then discuss whether such highly visible interest in, and use of, the disparaged tongue actually did its maintenance and development any favours.

2. Corpus and methodology The material analysed in this essay derives from the New Statistical Account (abbreviated to SA2), published in the 1830s and 1840s, and the Third Statistical Account (SA3), published from the 1950s to the 1990s. These accounts attempt an economic, social and cultural portrait of all the parishes in the Kingdom of Scotland. Naturally, given the time between the different Accounts, there were differences in production methods. This is seen in particular in the apparently ‘seamless’ production of the New Account in comparison with the rather slipshod presentation of the Third Account, despite the latter’s attempt at centralised editorial production. Nevertheless, the authorial and editorial precepts set down by Sir John Sinclair for the first Account in the 1790s (as discussed in Millar 2000 and 2003) were continued to a surprisingly large extent in the succeeding Accounts. This means that many of the idiosyncrasies of a highly localised authorship were maintained across the period, even if the clerical dominance of the first two Accounts shifts towards a schoolteacherdominated Third Account. The analytical framework employed in this essay is informed both by critical discourse theory, as formulated by (among others) Norman Fairclough (1992) and Ruth Wodak (1999), and by some of the insights developed by Tony Crowley (1989 and 1996) on the ideology of Standard English during the period. I assume an ideological dimension of some consistency in the comments made on language use and the overt expression of language attitudes.

166 Robert McColl Millar

This does not mean, however, that I believe that the authors of the various returns to the Accounts shared a common political or social ideology in any strict party sense. Such a view would be manifestly unjust to the individuals involved. I would claim, however, that a common background — educational and social — along with the sense of partial dislocation from the local community often felt by “authority” figures such as ministers of religion or schoolteachers (who make up the large part of the correspondents to the Accounts under consideration) would lend a common world-view on a variety of issues — social and linguistic — whether conscious or unconscious. Although on occasions I have placed linguistic use and attitudes expressed in the New Account alongside similar instances in the Third (or vice versa) without comment, I do not intend to suggest by this that there were no changes in views in the century or so between the two Accounts. Many of the bluntly undemocratic, paternalist and condescending views expressed by correspondents in the 1830s and 1840s would have been unacceptable in the post-War period. It is striking to note the extent to which some views (or analogous views) have been maintained over the period, however.

3. Analysis In the New Account return from Lochwinnoch, rural virtues are emphasized. Part of the rhetorical import of this message is carried by discussion of the use and nature of the local dialect: (1) SA2 Renfrew, Lochwinnoch The kingdom of Strathclyde comprehended the shires of Dumbarton, Renfrew, Lanark, Ayr, &c. and consisted of an independent British or Celtic people. It was conquered by the Saxons about the year 1000 or 1100, two hundred years after the other provinces in the lowlands. Hence the language of the west of Scotland, and the Renfrewshire dialect, contains far more words derived from the Gaelic than that of many other counties. The inhabitants of this parish spoke this kind of Scotch dialect exclusively till the public works introduced people from all parts of Scotland and even from Ireland, which has modified it somewhat. (98)

Interestingly, both of these features are connected to the idea of Britishness, something — as we might expect — of a loaded concept at the time. Certainly the kingdom of Strathclyde was a British kingdom (although it is striking that

Kailyard, conservatism and Scots in the Statistical Accounts of Scotland 167

the correspondent from Lochwinnoch collates this with the Gaelic influence from the Western Highlands undoubtedly felt in that part of Renfrewshire); the modern use of British as a coded reference to the Union and its aftermath (and its preference by Scottish intellectuals over Scottish) is well documented (for a recent discussion of these issues, see McLeod 2000). In the revolutionary climate in which the report was written, British, conservative, constitutional (in the sense expressed implicitly and to a degree explicitly in the original Account, of supporting a Whig constitutional monarchy as established in 1688 in alliance with a parliament elected by a small number of voters qualified by their position in society) virtues are being emphasized over the new, urban, implicitly foreign, radicalism. It is probably no accident that the area where these views are expressed is on the edge of the Clydeside conurbation, at the time, as later, a centre of political radicalism. Another element of this can be found in the discussion of the movement of population into Lochwinnoch from other parts of Scotland and also Ireland, thus causing linguistic — and, implicitly, societal — change. Expression of this worry can also be found in the following: (2) SA2 Lanark, Old or West Monkland The true Lowlanders in this district speak with great plainness the patois of the country, but they have a few expressions scarcely intelligible to their neighbours. For example, the word infidel is considered as synonymous with idiot; and when a man says, “Do you think I am an infidel?” a frequent interrogatory among the handicrafts, he merely means he is no fool, but knows what he is about. The no less common expression, “will you never deval?” merely means, will you never give over. (655–656)

Again a dialectic of purity, of ‘truth’ and ‘plainness’ is set up against the foreign, the other
: Irish or Highland; particularly important, it could be argued, in an area even closer to the industrial heartland of the West of Scotland. The use of local (implicitly, Scots) idiom is very much part of this process. Inherent in many of these ideas is a strong sense of connection to place, regularly associated with the use of Scots, as also suggested in: (3) SA2 Angus, Kirriemuir It has been remarked, that the people of Angus, as well as that of Aberdeen and the Mearns are in many respects different from the rest of the Scottish nation. These countries having been the chief part of the Pictish kingdom, it is probable that they retain the greatest share of the characteristics of that particular people; — having been less subject to the

168 Robert McColl Millar

invasion of the English, but more frequently intruded upon by the adventurers of the north of Europe than the population of the southern lowlands. The language of the people, accordingly, is in many respects different from that which is current in the more southern districts of Scotland. Dr. Jamieson, who lived long in this neighbourhood, remarks, in his introduction to his Scottish Dictionary, that the language of his country is spoken here in greater purity and more copiously than in any part of Scotland with which he was acquainted. His words are, “Having resided for many years in the County of Angus, where the old Scottish is spoken with as great purity as any where [sic] in Great Britain, I collected a vast number of words unknown in the southern and western dialects of Scotland. Many of these I found the classical terms in the language of Iceland, Sweden and Denmark.” It was here, accordingly, that he formed the resolution of writing a dictionary of the Scottish language. (179–180)

It is who did the invading which is important. Interestingly, the local form of speech is validated by its connection with scholarly work. As I point out elsewhere (Millar 2003: 320–321), correspondents to the Accounts, products of a culture of the book on a range of levels, are inclined to see their ‘reality’ in a light produced by their interpretation of authority. Implicit is the idea that the truly Scottish is to be found in the more remote areas. Interestingly, J. M. Barrie, a central figure in the Kailyard school, was a native of Kirriemuir in Angus. Similar ideas were still found in Moray in the 1950s: (4) SA3 Moray, St Andrews Lhanbryd Despite the educational advance and the inroads of the Southrons, the language heard on market day at Elgin or in the homes of farmers and field workers remains racy, of the soil, and of honourable ancestry (228)

Thus Scots is associated intrinsically not with Scotland as a whole, but with the local, normally rural, scene; it is also associated with “ancestral” — not contemporary — virtues. This sense of highly local virtue can sometimes be a double-edged sword: (5) SA3 Selkirk, Selkirk The ‘watters’, they are called in Selkirk: there was a Souter [a native of Selkirk] once who was asked had he seen the eclipse; “Naw, [sic: no quotation marks] he replied, “Ah wis up the watters that day”, and for the people of Selkirk the weekend is preferably spent in these delectable vales. (359)

Kailyard, conservatism and Scots in the Statistical Accounts of Scotland 169

Here local people are portrayed almost as simple-minded, and certainly highly parochial: a view supported by the — reported — use of Scots. This is even more prominent in (6) SA2 Kirkcudbright, Tongland The “refulgent lamp” of the glow-worm is often visible. On hot moist evenings I have seen multitudes of these beautiful insects scattered like “sparkling gems” over our meadows. From a bog, about a mile from the manse, I have frequently brought a plentiful crop to my garden, which, for many nights after their transportation, they would continue to illuminate. I had thrown seven one night into a grass plot in front of my house, and was much amused, next evening, at the alarm of one of my servants, who rushed suddenly into my room, exclaiming, that the grass before the door was “in a bleeze.” The poor woman, who never in her life had seen so many a glow-worm, had some reason for her apprehensions, as many a “bleeze” is less brilliant than the lustre of these earth-born pleiades. (87)

The use of quotation marks around the vernacular word, alongside the unhighlighted use of the learned pleiades, implies a number of things. In the first place, the academic assumption of the scholarly above the natural is played out. Moreover, an ironic sense of distance is constructed on the part of an amused middle-class observer, who, we can assume, is announcing that he would never himself be either taken in by events of this type nor use that kind of language. This rather condescending attitude is also to be found in (7) SA2 Ayr, Dundonald As a whole, the people are orderly, intelligent, comfortable as to food and raiment, and contented with their condition, except when some wiser heads would persuade them to the contrary; “but even than, they winna mind it lang.” (679)

Here, agitation for social change is seen as being the product of outside agencies, “wiser” than the forgetful — and naturally conservative — Scots-speaking parishioners. The corollary to this is that there are occasions in the Third Account in particular where local people and their speech are shown as being couthie and pawkie, for instance (8) SA3 Roxburgh, Yetholm When the writer came to preach for Yetholm he was very nervous as it was his first experience of preaching for a kirk. When John [a local man of Romany blood] came into the vestry to carry up ‘the books’ just before

170 Robert McColl Millar

the service, he turned and said: ‘Now laddie, speak oot and dinna gie a dump for yin o’ them.’ John was beadle for a period of 45 years, and during that time there were five ministers of the parish. On one occasion the writer’s brother was visiting Yetholm and asked John how he and the minister got on. John’s reply was: ‘Do ye ken, he (pointing to me) is the fifth meenister I have had under me and we get on grand for what he doesn’t ken, I tell him, and what I dinna ken, he tells me’. (358–359)

or: (9) SA3 Banff, Cullen It is worthy of note, that whereas formerly candidates for election were prepared both to give and to take with no quarter asked for or given, the modern candidate is a much more sensitive person and more liable to take a serious view of certain sallies, than to shake his heckler’s hand and call him a “bonnie fechter”. It is recalled that a former wheelwright once assailed his butcher rival at a nomination meeting with “Ay, John, ye’re a great man amon’ nowt (cattle), but a great nowt (nothing) amon’ men”. (255)

Yet reports of this type, whilst highly complimentary to the speaker in question, and conscientious in their presentation of setting, are inevitably involved in the presentation of a passing age; by implication, the language used is of a passing age as well. Inherent in all of these excerpts is a degree of conservatism. This is rather more explicit in a number of other returns. This conservatism may be political, and based on pragmatism, in this case about the results of the changes felt after 1945, when a relatively radical Labour administration instituted reforms which led to a state health care system and the bringing into public control of a variety of key industries: (10) SA3 Aberdeen, Lumphanan Only last week the writer met a road worker, who had gone to live in a village council house, and asked him how he liked his new quarters. “Nae very weel”, was the reply. “There’s unco little freedom jammed in atween twa neebors, an’ nae oot-rin at the back for hens or bees. I cud aye tak’ abeen £20 a year oot o’ th’ gairden i’ th’ auld place; an’ noo here A’m abeen £20 a year extra for rent an’ taxes. That mak’s mair than £40 a year doon on th’ower-turn. Practically aboot a poun’ a week oot o’ ma pey. Of course, th’ wife his a’thing handy i’ the hoose here, an’ nae ootgaun on an ill day t’ th’ wall for water — but it’s some dear”. (397)

Kailyard, conservatism and Scots in the Statistical Accounts of Scotland

Or it can reflect the position of capital and property within society: (11) SA3 Aberdeen, Social Life Against these figures labour people argue that they cannot get all their supporters to the poll because the party organisation is not good enough … There may be something else — more than one country man has been heard to say, “I daena like the Tories but we maun vote wi’ the folk wi’ the money else they’ll send it a’ out o’ the country.” (109)

Yet even here the modal maun “must” implies agreement with such a view. There is a Scottish tradition that the use of Scots implies truth-speaking, no matter how blunt, a tradition first employed in the liberal/radical press in the late nineteenth century, and handed down to the more conservative press today, as demonstrated in the Sunday Post (a highly popular newspaper which adheres to socially and to some extent politically conservative values, published in Dundee by the D. C. Thomson press). Indeed, a connection with the D. C. Thomson press is mentioned in one of these contexts, associated, it should be stressed, with the Bible: (12) SA3 Aberdeen, Social Life As one old lady said, long ago, “There are just twa writers for me — Job in the Bible and Annie Swan in The People’s Friend. They both bring a’thing out a’right in the end. (118)

The double distancing effect in the use of “old” and “long ago” should be particularly noted. Much of this conservatism, at least in the Third Account, is not so much political as it is social. This can be seen in yet another return from the Third Account volume on Aberdeenshire (the editor of this volume appears to have been particularly fond of this type of rhetorical displacement): (13) SA3 Aberdeen, Social Life It should be said of them [the tabloid Sunday papers, such as The News of the World] that they have brought the word “homosexuality” into the vocabulary of people who are still wondering what it can really be about. That sort of thing is a wonder and a puzzlement; something vaguely understood and instinctively dreaded. As one farmer said, after reading a report of on-going in London, W. C.1, “Well, well. It cannae dae me nae hairm, but I’m glad our bull canna read.” (117)

Again a polarity is constructed between the ‘unnatural’ acts associated with Middle-Class London and the rural reality; again Scots is employed to represent

171

172 Robert McColl Millar

‘truth’, but a ‘truth’ which inevitably allies itself with a passing age. In the earlier Accounts this ‘political’ aspect is not so visible: quite probably because the large majority of the population were not permitted to participate in the political process. Nevertheless, a pawkie commentary on social matters is permitted: (14) SA2 East Lothian, North Berwick A few years ago, an incident occurred on the Bass, expressive of a strong lingering desire to retain the chapel, occasionally at least, for its original destination. A young lady, in the presence of her father, was here solemnly confirmed in her Romish faith and profession, and the due ritual services were gone through in the presence also of the keeper of the Bass and his boat assistant. On the conclusion of the solemnities, the priest turned to the keeper, and asked him, with due decorum, if he would not now also kneel down before the altar, and follow them in similar dedication and worship. “Me?” said the Protestant Presbyterian James, “Me? Na, na, am thankfu’ there’s mair sense gi’en me. — I wad just as soon, Sir, fa’ doon and worship one o’ thae puir solan geese about us,” (pointing to the myriads around him) “than e’er gang on wi’ ony sic mockery.” My friend and parishioner James remains an invincible adherent of the Reformation, and also, as well may be conjectured, the Bass being ever before him, a stern abhorrent of prelatic tyranny and regal despotism. (331)

This story should be read in the context of Catholic emancipation during the period and of large-scale Irish immigration. Yet the middle-class commentator frees himself from any accusation of bigotry by displacing views onto a working-class speaker of ‘proverbial’ Scots. To what extent are such views genuine representations of the way Scots speakers felt at a given time? Doubtless many Scots shared — and share — the conservative social attitudes with which the writers of the Accounts associate them. Yet even if this is the case, their views are being presented in a Standard English discourse by an essential outsider who is framing an argument of his (occasionally her) own desire and design. One of the central building blocks of this process is the supposed truth-representing qualities of the Scots tongue. Whether it benefited — and benefits — from such a common refraction is questionable. Indeed, at times it seems that the intention is half-mocking. Commentators were inclined to celebrate the local, the particular; on the other hand, it is possible to see that a tone of ridicule may be embraced in the discussion of local wit, for instance:

Kailyard, conservatism and Scots in the Statistical Accounts of Scotland

(15) SA2 Ayr, Dreghorn In reference to what is here stated [that the people are clannish], the following anecdote is told. A good woman of Dreghorn, who had lost her husband, not seeming very disconsolate in her state of bereavement, was remonstrated with by her neighbour on her apparent hardness of heart, and want of sensibility. She pleaded guilty to the charge, but excused herself by saying, “It’s true, Janet, am no so muckle o’ercome as I ought to be; but you see John, puir man, was no o’ our parish.” (525)

Of course there is a long tradition of anecdotes of this sort; the people about whom these stories were told would probably both have told stories of this type themselves and also laughed at the story told. This misses the point somewhat, however. No-one has asked them what they think. A largely invisible (middle class) commentator has placed them in print. This is shown most strongly where the narrator breaks into the discourse: (16) SA3 Aberdeen, Tarves As an old man with a wide whisker once said to the present reporter, “Just put a right coo wi’ a good pedigree intae ony park at Tillycairn and in time she will calve a prize bull.” Well, maybe: but there is a certain something in management, which includes choosing the right cow. (495)

This is a double bind. Middle Class commentators appear to want local people to remain their ‘natural’ selves, thus preserving an aura of ‘true’ Scottishness; at the same time, this ‘authenticity’ is held up for mild ridicule, with the suggestion that the modern world has moved on. Scots as a vibrant focus of national identity must inevitably be caught in this vice.

4. Discussion What can be made of these examples? Scots is, as has been suggested, associated with a passing age — a rural, frugal, protestant, socially cohesive and paternalistic Scotland: the kailyard. What is most striking about this is that these views are placed, by largely middle class commentators, on the lips of the working class. More covert, perhaps, is the sense of mockery — even of the ridiculous — with which the use of Scots is associated. This is in line with Macaulay’s discussion (1991: 185–186) of the use of Scots by the Middle-Class citizens of Ayr. If you do not come from such a background, you may be tempted not to see yourself as being a “Scots” speaker; even if you do come from a rural background, you

173

174 Robert McColl Millar

may not be willing to associate yourself with the ideas expressed. As commentators on language maintenance and the Ausbau of vernaculars have demonstrated, a language, to survive, must be kept in the public eye. In that sense the language attitudes and language use in the Statistical Accounts, and the kailyarding tradition from which it draws, suit the prerequisites of language maintenance. Yet the foregrounding which the language is given is highly problematical. In the first place, Scots is placed within a discourse which treats Standard English overwhelmingly as the default. As Heinz Kloss (1978 and elsewhere) points out, it is not enough that a language has a written presence: it needs to be used in a considerable number of contexts, both literary and (in particular) non-literary, not only dealing with local matters. This cannot be fulfilled by uses such as this. More worrying, perhaps, is Fishman’s repeated assertion (1991 and 2001) that if a language is not used in a wide range of domains — both spoken and written — it will inevitably begin to decline where it is used. Furthermore, the discourse places Scots in the past. The language is viewed as decaying in the new, urban, society (and perhaps even as lacking in relevance in such a framework). This may in fact be the case, as a number of modern studies, most notably Macafee (1994), have suggested. Yet the discourse used has the makings of a self-fulfilling prophecy. Uses of, and attitudes about, Scots of this type inevitably paint the vernacular into a corner when it comes to its associations — even among native speakers — with an at least partially invented past.

5. Conclusion The ‘decline’ of Scots in the 19th and 20th centuries was not caused by the language use and language attitudes employed by contributors to the Statistical Accounts of Scotland
; at least not solely by them. That is too great a change to be associated with one single cause. Yet the Accounts were (and to some extent are) a major cultural product of the ideologies of the educated in Scotland, who also happened to be the opinion-formers, both overtly and covertly, of the views of the succeeding generations. The Accounts can therefore be seen as acting as a mirror of the general language uses and language attitudes of that group. The association of Scots with a desired or disparaged past, and its use in a way that was at least at times mocking, demonstrates that the language attitudes of a large part of this class encouraged this decline, particularly since it is they, and not the actual speakers, who control the written use of Scots.

Kailyard, conservatism and Scots in the Statistical Accounts of Scotland

References Blake, George. 1981. Barrie and the Kailyard School. London: Barker. Campbell, Ian. 1981. Kailyard. Edinburgh: Ramsay Head. Crowley, Tony. 1989. The Politics of Discourse: the standard language question in British cultural debates. London: Macmillan. ———. 1996. Language in History. Theories and Texts. London: Routledge. Donaldson, W. 1986. Popular Literature in Victorian Scotland: Language, Fiction and the Press. Aberdeen: Aberdeen University Press. Fairclough, Norman. 1992. Critical language awareness. London: Longman. Fishman, Joshua. 1991. Reversing Language Shift. Clevedon: Multilingual Matters. ——— ed. 2001. Can Threatened Languages Be Saved? Reversing Language Shift Revisited: A 21st Century Perspective. Clevedon: Multilingual Matters. Kloss, Heinz. 1978. Die Entwicklung neuer germanischer Kultursprachen seit 1800. 2nd edn. Düsseldorf: Schwann. Knowles, Thomas D. 1983. Ideology, art and commerce: aspects of literary sociology in the late Victorian Scottish Kailyard (= Gothenburg Studies in English 54). Göteborg: Acta Universitatis Gothoburgensis. Macafee, Caroline. 1994. Traditional dialect in the modern world: a Glasgow case study (= Bamberger Beiträge zur englischen Sprachwissenschaft 35). Frankfurt am Main: Lang. Macaulay, Ronald K. S. 1991. Locating dialect in discourse: the language of honest men and bonnie lasses in Ayr. Oxford: Oxford University Press. McCaffery, John. 1998. Scotland in the Nineteenth Century. London: Macmillan. MacLaren, A. A. 1989. “Patronage and Professionalism: The ‘Forgotten Middle Class’ 1760–1860”. The Making of Scotland: Nation, Culture and Social Change ed. by David McCrone, Stephen Kendrick & Pat Straw, 133–147. Edinburgh: Edinburgh University Press. MacLeod, James Lachlan. 2000. The second disruption: the Free Church in Victorian Scotland and the origins of the Free Presbyterian Church. Phantassie, East Linton: Tuckwell Press. Millar, Robert McColl. 2000. (with the assistance of Dauvit Horsbroch) “Covert and Overt Language Attitudes to the Scots Tongue expressed in the Statistical accounts of Scotland”. The History of English in a Social Context: a contribution to Historical Sociolinguistics (= Trends in Linguistics Studies and Monographs 129) ed. by Dieter Kastovsky & Arthur Mettinger, 169–198. Berlin & New York: Mouton de Gruyter. ———. 2003 “‘Blind attachment to inveterate customs’. Language use, language attitude and the rhetoric of improvement in the first Statistical Account.”. Insights into Late Modern English (= Linguistic Insights 7) ed. by Marina Dossena & Charles Jones, 311–330. Frankfurt am Main: Lang. Robinson, Mairi, ed. 1985. The Concise Scots Dictionary. Aberdeen: Aberdeen University Press. Wodak, Ruth. 1999. The discursive construction of national identity. Edinburgh: Edinburgh University Press.

175



176 Robert McColl Millar

Sources The Society for the Benefit of the Sons and Daughters of the Clergy, eds. 1841–45. The New Statistical Account of Scotland. Edinburgh: Blackwood. Various, eds. 1951–1996. The Third Statistical Account of Scotland. Various publishers. References to the Accounts are given in the following way: (SA2/3; Pre-1975 County [where appropriate, without -shire suffix]; Parish [or, occasionally, for SA3, specialised topic]; Page reference to particular volume)



A sociolinguistic approach to the Norse-derived words in the glosses to the Lindisfarne and Rushworth Gospels* Sara M. Pons-Sanz Queens’ College, Cambridge

1.

Introduction

The study of the Norse-derived words first recorded during the Old English period has mainly revolved around the identification of the terms, as well as the contrast between their predominantly technical nature and the everyday character of those first attested during the Middle English period.1 There are, of course, some important exceptions to this general tendency. For instance, in 1989 Fischer studied the process of the introduction, accommodation and eventual domination of the Norse loanword lagu over the native œ(w). ¯ One year before Thomason and Kaufman had presented their theory about the spread of ‘Norsified English’ from the Midlands to the old kingdom of Deira and later on

*My thanks are due to Richard Dance, D. Gary Miller and Martin J. Syrett, who have patiently read and commented upon earlier versions of this article. I am also very grateful to Andreas Fischer and Theo Vennemann for having provided me with off-prints of their respective articles mentioned in this paper. 1.I follow Townend (2002: xv) in the use of ‘Old Norse’ to designate the language spoken by Scandinavians during the Viking Age, and in the use of ‘Norse’ as a linguistic term as opposed to ‘Scandinavian’, which is employed with cultural or historical reference. I follow Haugen (1950) in interpreting ‘loanwords’ as a type of ‘loan’. Because of the difficulty involved in differentiating between what Haugen classifies as ‘loan-blends’ and ‘hybrid formations’ (see Dance, 2003: §3.4.3), I use the all-encompassing ‘Norse-derived word’ to refer to all varieties of loans, as well as new-formations within English utilizing a Norse loan and a native affix with no direct Norse model. I owe this phrase to Dance (2003), although my use is slightly different to his; see Dance (2003: §3.3).

178 Sara M. Pons-Sanz

to Northumberland, i.e. from the core of the so-called Danelaw to the peripheral areas. This may be a sensible explanation for the expansion of the Norsederived and Norse-influenced vocabulary. However, it is my aim in this paper to present some evidence in favour of the importance of the peripheral areas of the Scandinavian settlements during the initial years of the linguistic contact. To do so, I shall concentrate on Aldred’s glosses to the Lindisfarne Gospels and Owun’s glosses to the Rushworth Gospels. I will analyse them according to the model devised by Rogers and Schoemaker (1971) to account for the diffusion of new ideas and inventions, and adapted by Granovetter (1983) and Milroy and Milroy (1985) to sociological and sociolinguistic studies.

2. The model and its applicability to Anglo-Scandinavian Northumbria This model explains that the transmission of innovations, either technological or linguistic, only occurs in those cases in which social groups with differing standards are forced to interact, i.e. to create weak ties between themselves, and not in those cases in which interaction is restricted to the creation of strong ties between members of the same group. Granovetter (1983: 1361) suggests that “the strength of a tie is a (probably linear) combination of the amount of time, the emotional intensity, the intimacy (mutual confiding) and the reciprocal services which characterise a tie”. Weak ties are more likely to link members of different small groups than strong ones, which tend to be created within particular groups. Thus, while strong ties give rise to social cohesion, they lead, paradoxically, to overall fragmentation. It is likely that in any society there will be more weak than strong ties, and that many more individuals will be reached through the former than through the latter. This model can be applied to the contrasting situation in northern and southern Northumbria as regards the Scandinavian presence. Southern Northumbria, with its centre at York, was highly influenced by Scandinavian invaders and settlers after Halfdan’s division of lands in 876.2 The old kingdom of Bernicia, where most of St Cuthbert’s lands were located, was not equally affected by these newcomers, and, indeed, Samuels (1989: 111) excludes the areas beyond the Tees from those which were part of the ‘Scandinavian belt’.

2.See Hill’s (1981) Map no. 68, which should be treated with a caveat in mind because he does not specify whether the represented Grimston-hybrids and place-names with -by also include those best assigned to a post-Conquest period.

A sociolinguistic approach to the Lindisfarne and Rushworth Gospels 179

Watts’s 1988–89 and 1995 studies indicate that the area most Scandinavianised before the Norman Conquest was that between Barnard Castle in the west and Hart in the east. He connects the Norse-derived and Norse-influenced placenames in the mid and low Tees Valley with Halfdan’s land-sharing in 876, and associates those on the coast with the division of lands carried out by Ragnald after the second battle of Corbridge (late 910s). Many of these lands were previously owned by St Cuthbert, and, in fact, the community may have recovered some of them reasonably early.3 Morris (1977: 98–99; 1981: 227) prefers to associate the Norse-derived and Norse-influenced place-names around Gainford-onTees either with Halfdan’s land-sharing in 876, or with the “sons and daughters of the original settlers further south and east who preferred to exploit relatively rich land in north Yorkshire and south Durham, to relatively poor lands in the areas of primary settlement”. In any case, these territories would have been occupied by Scandinavian settlers by the time when Aldred wrote his glosses. There are also Norse-derived and Norse-influenced place-names to the west of Chester-le-Street, most of which are assigned to a post-Conquest period by Watts (1988–89: 57), and in the Upper Teesdale and Upper Weardale, where there would have been infiltration of Hiberno-Norse settlers over the Pennines. However, only the area around Sadberge was Scandinavianised enough for the formation of a wapentake to take place. Both Watts and Morris indicate that the settlement of the newcomers would have involved a process of ‘infilling’ which must have led to the cultural and political mixture reflected in Anglo-Scandinavian art throughout the south of County Durham.4 The differences in the scale of the Scandinavian presence may have created divergent structures in the social relations between the newcomers and the locals. One could hypothesize that, in those areas where the number of Scandinavian settlers was lower, they would have been forced to create a larger number of weak ties with the native population at an earlier stage than those people who settled in areas with an important Scandinavian presence. Originally, the social life of the latter may have been oriented towards forming strong ties among themselves. This hypothesis agrees with Björkman’s (1900–02: 6) suggestion that the Scandinavians would have given up their idiosyncratic characteristics earlier in the borderline areas than in the core areas of settlement. It

3.See Hall (1984) on the possessions of the community. 4.See, for instance, Bailey (1980: chs 8 and 10) and Lang (1984) on the integration of the Tees Valley in the trends of Anglo-Scandinavian art. cf. also Pons-Sanz 2000: 181.

180 Sara M. Pons-Sanz

is also in keeping with the indication in Watts’s studies on place-names in northern Northumbria that Old Norse died earlier there than in southern Northumbria, as suggested by the lower rate of retention of the Norse inflections. The implication of the aforementioned hypothesis is that the texts produced in the territories under the control of the community of St Cuthbert may initially show a higher degree of Norse influence than those produced further south. However, the latter may present deeper Norse influence in a later period, owing to the longer contact of the two languages in this area, and to the higher number of people who would have created weak ties with the Norse speakers once they proceeded to integrate themselves fully in the Northumbrian community. As already explained, this paper is only concerned with the initial phase. Milroy and Milroy (1985: 380) indicate that in the sociological model just explained prestige does not play a significant role in the transmission of linguistic innovations. However, when considering the linguistic situation to which it is applied here, it is important to note the differing diglossic situations concerning Old English and Old Norse in the areas under study. The territories in southern Northumbria were under the control of Scandinavian leaders from 876 until 954, whereas the border of what remained English Northumbria, ruled by ealdormen in Bamburgh and the community of St Cuthbert, is usually placed on the river Tees. This may account for the presence of a high number of technical loans in the southern territories even at an early stage.5

3. The corpus of study One part of the corpus comprises the Aldredian glosses to the Lindisfarne Gospels (London, British Library, MS. Cotton Nero D.iv), edited by Skeat (1871–1887). It would take me too far from the object of this paper to discuss the details of the exact dating of the composition of these glosses. It should be sufficient to explain that Ross, Stanley and Brown (1959–60: 32) conclude that the glosses to the Durham Ritual, also attributed to Aldred (pace Drago, 1977:61), must be later than those to the Lindisfarne Gospels. These authors date the glosses to the Durham Ritual to 970 from the information given in the colophon,

5.This slight modification of the model agrees with Vennemann’s (2002: 240) view that “military, legal, and constitutional terms are central concepts of a society and as such reflect the language of the ruling class. If such terms are of foreign origin, the language of the ruling class is likely to have been a foreign language”.

A sociolinguistic approach to the Lindisfarne and Rushworth Gospels

and this leads them to propose the dates c950 x 970 for those to the Lindisfarne Gospels. The second part of the corpus of this study includes Owun’s glosses to part of the Rushworth Gospels (Oxford, Bodleian Library, MS. Auct. D.ii.19), also edited by Skeat (1871–1887). Owun’s work, known as Rushworth2, consists of the glosses to St Mark’s gospel from 2.15, St Luke’s gospel, and St John’s gospel, except for 18.1–3. Whereas the Aldredian glosses have for a long time been considered to represent a late northern Northumbrian dialect, Owun’s are identified as evidence of tenth-century southern Northumbrian,6 even though the exact location of the place where the glosses were written is still a matter of dispute.7

4. The Norse-derived words in the glosses to the Lindisfarne and Rushworth Gospels A study conducted by Pons-Sanz (2000: Ch. 5) has shown that, in contrast with the general trend, the Norse-derived words from specialist and technical registers in the Aldredian glosses to the Lindisfarne Gospels are a minority in comparison with those of a miscellaneous character. This is true even if we only take into account the words generally accepted to be Norse-derived and those which probably are; the ratio is seven to eleven. If we also count the possible Norse loans, then the ratio becomes eight to sixteen. A. Words generally accepted to be Norse-derived8 1. Social terms: br¯yd(h)lo¯p (< OE br¯yd ‘bride’ + ON hlaup ‘running’ and ON brúðhlaup ‘marriage’) hold (< ON höldr ‘ranking below a jarl’)

6.This distinction was already established at the end of the nineteenth century by Lindelöf (1893: 298–302; 1901: iii–iv). See Bibire & Ross (1981: 99) for a summary of the morphological differences between the glosses to the Lindisfarne Gospels and Rushworth2. 7.See Bibire & Ross (1981: 98–99) and Coates (1997) on this matter. The rest of the glosses to the Rushworth Gospels, i.e. the so-called Rushworth1, are, for the most part, independent from Aldred’s work. The glossator, Farman, was from Mercia, which makes his glosses irrelevant for this study. 8.See, for instance, Hofmann (1955: §§237–43, 247 and 252), Holthausen (1934: s.vv. br¯ydhlo¯p, hold 3, ðrœl, ¯ ðı¯r, fle¯ge, o¯ran, b¯y, cost 1, dearf, eggian, seorðan, sang 2, ðweng), Peters (1981: 88–99 and 103–105) and Serjeantson (1935: 64–69).

181

182 Sara M. Pons-Sanz

2. 3. 4. 5.

ðı¯r (< ON þír ‘female servant’) ðrœl ¯ (< ON þræll ‘slave’) Legal terms: sacle¯as (< ON saklauss ‘innocent, without charge’ but with OE -le¯as) Nautical terms: fle¯ge (< ON fley ‘a kind of swift ship’) Monetary term: o¯ra(n) (< ON aurar ‘monetary unit’) Miscellaneous terms: b¯y (< ON b¯y ‘village, farmstead’) cost (< ON kostr ‘manner, way, custom’) dearf (< ON djarfr ‘bold, daring’; the hybrid complexes dearflic and dearfscipe have the loanword as their base) geeggian (< ON eggja ‘to egg on, incite’) serðan (< ON serða ‘to have improper sexual intercourse with someone, rape’) song (< ON sæng ‘bed’) tre¯ (< ON tré instead of OE tre¯ow ‘tree’) ðweng (< ON þvengr instead of OE ðwang ‘bond, tie’)

B. Probable Norse loans 1. Miscellaneous term: hundrað (< ON hundrað instead of OE hundred ‘hundred’)9 C. Possible Norse loans 1. Monetary term: fe¯orðing (< ON fjórðungr ‘fourth part’, but with OE fe¯orða ‘fourth’ + -ing)10 2. Miscellaneous terms: afolic (< ON öfugr ‘perverse’ + OE -lic)11

9.Its Norse origin is supported by Björkman (1900–02: 163), the MED (s.v. hundred), the OED (s.v. hundred), Thomason & Kaufman (1988: 294) and de Vries (1961: s.v. hundrað). Hofmann (1955: §248) and Peters (1981: 112–13) assign a native origin to this term. 10.The OED (s.v. farthing) associates the word with the Norse term, and suggests that it could have been adopted from it. However, this word is not mentioned by Björkman (1900–02), Hofmann (1955), Peters (1981) or Serjeantson (1935). De Vries (1961) does not include its possible Norse etymon in his dictionary. 11.Björkman (1900–1902: 20, n.1) suggests the possible Norse origin of the word, and Hines (1991: 424) includes it in his list of “Scandinavian loan-words in the late 10th-century

A sociolinguistic approach to the Lindisfarne and Rushworth Gospels

gesparrian (< ON sparra ‘to close, lock’)12 getryccan (< ON tryggja ‘to trust’)13 macalic (< ON makligr ‘appropriate, similar’ instead of OE gemæclic ‘conjugal’)14 sunset (modelled on ON sólarset ‘sunset’, but with OE sunne ‘sun’ and set ‘seat, habitation’)15 Another of the peculiarities of these glosses is the fact that some of the Norsederived words used by Aldred have not been recorded in any Old English text other than these glosses and (in some cases) those to the Rushworth Gospels, which were written under Aldred’s influence:16 b¯y (as an autonomous word), dearflic, dearfscipe, geeggian, fle¯ge, hundrað, song, ðı¯r, as well as getryccan and sunset. As Table 1 shows, Owun used almost the same Norse-derived words as Aldred except for those which appear in the sections glossed by Farman, or those which

Northumbrian glosses”. The OED (s.v. awk) and de Vries (1961: s.v. öfugr) give the Northumbrian form *afuh as native. Hofmann (1955), Peters (1981) and Serjeantson (1935) do not mention this term. 12.It is assigned a Norse origin by Holthausen (1934: s.v. gesparrian) and Serjeantson (1935: 68). Peters (1981: 110–111) prefers to analyse it as a native word, but accepts that its use may have increased thanks to the Norse cognate. It is not mentioned by Björkman (1900–02) or Hofmann (1955). 13.Hofmann (1955: §243) rejects the Norse origin of the verb on the grounds that we would expect it to have retained the voiced stop (cf. geeggian), and that the Norse verb had the meaning ‘to make firm or trusty’, whereas Aldred used it intransitively, as a gloss for “confidete”. Hofmann prefers to connect this verb with OE getricce ‘contented’, which is used in Æthelwold’s translation of the Benedictine Rule, a work completed in southern England in the 960s. Nonetheless, this work contains another possible Norse loanword, viz., sce¯ot
183

184 Sara M. Pons-Sanz

Table 1.Attestation of the Aldredian Norse-derived words in Owun’s glosses Aldredian words included in the sections glossed by Owun

Words generally accepted to be Norsederived

Aldredian words included in the sections glossed by Farman

Words found as Aldred’s first or only alternative fi accepted by Owun

cost dearf dearflic dearfscipe serðan

br¯yd(h)lo¯p b¯y? fle¯ge o¯ra(n) sacle¯as tre¯ ðı¯r ðrœl ¯

Probable Norse loans Possible Norse loans

Words found as Aldred’s first or only alternative fi rejected by Owun

Words not found as Aldred’s first alternative in a multiple gloss fi rejected by Owun geeggian song

hundrað afolic gesparrian

fe¯orðung macalic sunset

getryccan

are only used as a second (or third) alternative in one of Aldred’s multiple glosses, which Owun tended to avoid by choosing the first alternative.17 However, when glossing Mark 15.11, Owun’s decision not to follow the Aldredian gloss for “concitauerunt”, viz., “gewæhton vel geeggedon”, was not due to his dislike for multiple glosses because he accepted Aldred’s first alternative, and proposed “gicedun” as the second one. The importance of this example may be enhanced by the possible implications of the selection of the glosses as explained by Hines (1991: 410–11): “Alternative glosses placed in second place are unlikely to be words that should be less familiar to the reader […] if anything, they should be more familiar, ensuring that the meaning is not missed or mistaken”. Although Hines may be right, it would be a tedious task to test his assumption because frequency tables for each word used in multiple glosses would have to be constructed. Instead, another statistical approach seems to be

17.Song is the first alternative once (Mark 16.5), but this passage is lost in a lacuna in Rushworth2.

A sociolinguistic approach to the Lindisfarne and Rushworth Gospels

Owun chooses an unrelated form to the alternatives proposed by Aldred

Owun maintains the first two alternatives proposed by Aldred in the same order

6.73%

19.17% Owun maintains the first two alternatives proposed by Aldred but inverts the order

39 111

18.82% Owun simplifies Aldred’s multiple glosses by choosing the second alternative

109

11 1

1.89% 0.17% Owun accepts one of the alternatives proposed by Aldred in a multiple gloss but replaces the other

308

Owun simplifies Aldred’s multiple glosses by choosing the first alternative

53.19%

Figure 1. Owun’s attitude towards Aldred’s multiple glosses in St John’s gospel

more appropriate. The study of Owun’s attitude towards the multiple glosses written by Aldred in the gospel according to St John, randomly chosen among the three gospels which he glossed, shows the rarity of Owun’s choice when glossing “concitauerunt”. It implies that he avoided geeggian because the acceptance of Aldred’s first alternative and the replacement of the second occurs only once in the glosses to the gospel of St John, as shown in Figure 1.18 When Owun chose the first Aldredian alternative in John 16.33, where “getreuað vel gelefeð vel getryccað” render “confidete”, and in Luke 22.12, where “bedd vel song” render “stratum”, he was following his preferred way of dealing with multiple glosses. Getryccan and song could, in principle, have also been chosen, though, because Owun selected the second alternative on 109 occasions in St John’s gospel. His choice, together with the fact that the two words are only recorded in the glosses to the Lindisfarne Gospels, may again show a dialectal difference between the two glossators. Even so, it would be foolhardy to press this argument any further because the choice of the first variant is, in fact, Owun’s most common procedure.

18.These percentages can also be used as evidence against Hines’s view because we may have expected Owun to choose the most common alternative among those provided. Therefore, if any, the first alternative in a multiple gloss may have been the most familiar.

185

186 Sara M. Pons-Sanz

Figure 1 should present sufficient evidence that, although Owun followed the Aldredian glosses closely, his work was not slavishly determined by Aldred’s. Three words require closer attention. Aldred rendered “domicilium” in Mark 5.3 as “hus vel lytelo by”, a multiple gloss which Owun decided to maintain, but with some modifications, viz., “hus vel byinge lytle”. The fact that he added a morpheme used to create deverbal nouns may point to his identification of the Aldredian term not as the Norse noun, but as a word related to the verb b¯ya, a form of OE bu¯an regularly used to gloss “habitare” and “possidere” in the Lindisfarne Gospels. B¯ya may itself have been formed by the assimilation of the English verb to the imported Norse noun, and, even if this were not the case, we could still argue that its use may have increased because of the influence of the Norse term.19 It is difficult to believe that Owun would not have known the Norse place-name formative -by because he would have been surrounded by places bearing it. One could argue that he had recognised the term, but preferred to use a deverbal noun instead, in the same way that he decided to change Aldred’s word order. Alternatively, a semantic explanation for Owun’s choice can also be proposed. According to Cleasby and Vigfusson (1957: s.v. býr), ON býr was used mainly with the senses ‘town, village’ or ‘farm, landed estate’. Therefore, Owun, being familiar with the meaning of the Norse term, may have found Aldred’s use of this word to refer to a single building inappropriate. This is not the only case in which he changed the ending of a Norse loan. He added the correct ending to sacle¯as in John 15.25, i.e. the morpheme -ne, which indicates that we are dealing with the accusative singular of the strong form of a masculine adjective. Thus, whereas Aldred’s form may present the first step in the integration of the hybrid, in which the word is not yet a proper member of the system, Owun’s could reflect a step further in this process. However, we should not place too much importance on the use of a grammatically inappropriate form by Aldred because grammatical confusion is frequently seen also with native adjectives.20 Aldred used the form “hundrað” instead of OE hundred to gloss “centum” on twelve occasions. In five of the six examples in Rushworth2 (Mark 4.8, Mark 4.20, Mark 6.40, Mark 14.5 and Luke 15.4) Owun preferred the form “hundreð”, and used “hundred” once (Mark 14.5). Owun’s forms have the fricative consonant

19.See Hines (1991: 412) for an argument in favour of Norse influence on this verb, and Hofmann (1955: §247) for an opposing view. 20.See Ross (1937: 105–106).

A sociolinguistic approach to the Lindisfarne and Rushworth Gospels 187

of the Norse noun in most cases but also the English vowel. So, whereas the Aldredian word could be described as a loanword, Owun’s should rather be interpreted as a Norse-influenced or a native form. In conclusion, Owun may have known some of the Norse-derived words used by Aldred (br¯yd(h)lo¯p, fle¯ge, o¯ra(n), sacle¯as, tre¯, ðı¯r, ðrœl, ¯ fe¯orðung, macalic, sunset), whereas he does not appear to have been as familiar with others (geeggian and hundrað); the case of b¯y is difficult to interpret. Owun’s familiarity with other Aldredian Norse-derived words cannot be known because they do not appear in the portion of the gospels which he glossed. However, it is worth noticing that under no circumstances did he choose the second (or third) alternative in a multiple gloss when it was a Norse loan, nor did he choose a Norse-derived word as a different alternative to those provided by Aldred, even though he was familiar with both procedures, as Figure 1 shows. Thus, whereas in the Aldredian glosses to the Lindisfarne Gospels there is a predominance of miscellaneous Norse-derived words over technical ones,21 in Owun’s glosses we find six plausible or probable technical loans as opposed to one or two everyday loans, depending on whether we include b¯y. If we also count the possible Norse loans, the disproportion increases slightly. This difference, however, should be taken with the caveat in mind that some of the everyday Norse-derived words used by Aldred occur in the section glossed by Farman; therefore, Owun did not have much chance to use them.

5. Analysis of the data through the sociolinguistic model The results as they stand seem to support the hypothesis suggested in Section 2 on the basis of the sociolinguistic model used in this paper. Owun’s community was one in which strong ties among the Scandinavian population would have been more important initially. The arrival of the Hiberno-Norse leaders and, especially, the West Saxon kings in 954 would have boosted the development of weak and strong ties among the locals and the Scandinavian settlers because, in Hadley’s (2000: 306–07) words, the “divergent interests of the indigenous

21.If we only take into account the Aldredian terms which appear in the part followed by Owun, and count only those words generally accepted to be Norse-derived and those which probably are, the miscellaneous terms relinquish the majority (five against six, which still is a much smaller disparity than in Owun’s glosses). The miscellaneous terms regain the majority if the possible Norse loans are also taken into account.

188 Sara M. Pons-Sanz

population and the Scandinavian settlers may have been thrown into confusion”. Owun’s work appears to reflect the situation in which Old Norse was the prestigious language of government and its connections with economy, law and social stratification, as well as naval and martial activities. Still, the infiltration of this language in the everyday life of the native population may not have reached as yet the same level as in Aldred’s dialect, in which a higher number of weak ties between the two communities would have been working for a longer period. Thus, in Aldred’s dialect the adjective dearf has already become fully integrated in the linguistic system, and the hybrids dearflic (cf. ON djarfligr ‘bold, daring’) and dearfscipe (cf. ON djarfleikr ‘boldness’) may be native newformations rather than loan-translations. Even though Aldred used a significant number of Norse-derived words, his glosses do not present the borrowing of any structure words or a clear morphological influence from Old Norse, despite the great simplification of forms and the confusion of grammatical gender in his glosses.22 Structure words and grammatical features are the lowest in the ‘scales of adoptability’ or ‘accessibility’, as argued, among others, by Haugen (1950: 223–24) and Lefebvre (1984: 14). The importation of these elements can only take place when the speakers of the two languages live together for a prolonged period as a single close-knit community. This implies that the contact between the Scandinavians and the people in the lands of St Cuthbert had not been long enough at Aldred’s time for this process to take place. What we find in Aldred’s glosses is, therefore, what we would expect to find by applying the model, and taking into account temporal and linguistic constraints.

6. Conclusion Some assumptions have been made, and some problems have been overlooked so as to test the explanatory capacity of the sociolinguistic model used in this paper: 1. Almost nothing is known about Aldred. He may have come from the southern part of Northumbria, or from the area within the wapentake of

22.See Ross (1937) on the process of morphological simplification in these glosses and idem (1936) on Aldred’s confusion of grammatical gender. See Stein (1986) and Thomason & Kaufman (1988: 275ff.) for an explanation of these processes as a consequence of the AngloScandinavian contact. See Jones (1967, 1970 and 1988), Kastovsky (2000) and Millar (2000) for a vision of these processes as internal developments in Old English. See Markus (1988) for an integrative view.

A sociolinguistic approach to the Lindisfarne and Rushworth Gospels 189

Sadberge, which would lead to a different interpretation of the findings of this paper. The linguistic features of the glosses seem to indicate that we are dealing with a different dialect from that of Owun, which would argue against the first option, whereas the second one still stands as a possible drawback for the validity of our conclusions. 2. The facts that Owun followed Aldred’s work, that he did not gloss the whole of the Rushworth Gospels, and that he had a tendency to simplify the Aldredian multiple glosses make the interpretation of the linguistic evidence from Rushworth2 very problematic. They do not allow us to have a clear view of the situation in southern Northumbria at a time almost contemporary with Aldred’s. However, by following these assumptions, and disregarding some of the problems, it has been possible to show that it may be misleading to present the Norse-derived words first recorded during the Old English period as a mere list of terms organised by semantic fields without any reference to their temporal and topographical distribution. The linguistic situation in northern Northumbria may have been somewhat different from that in southern Northumbria during the second half of the tenth century. This situation would contradict, at least partially, Thomason and Kaufman’s (1988) theory on the spread of ‘Norsified English’. The importance of these findings and their dependence on the correct selection of the aforementioned assumptions make it necessary to apply the same sociolinguistic model to later data from northern and southern Northumbria, and to other peripheral areas of the Scandinavian settlements. This might show whether the procedure followed in this work is appropriate, or needs to be thoroughly revised.

References Bailey, Richard N. 1980. Viking Age Sculpture in Northern England. London: Collins. Bibire, Paul & Alan S. C. Ross. 1981. “The Differences between Lindisfarne and Rushworth Two”. Notes and Queries 226.98–116. Björkman, Erik. 1900–1902. Scandinavian Loan-Words in Middle English. 2 vols. Halle: Max Niemeyer. Blakeley, L. 1947–1948. “Accusative-Dative Syncretism in the Lindisfarne Gospels”. English and Germanic Studies 1.6–31. Cleasby, Richard & Gudbrand Vigfusson, eds. 1957. An Icelandic-English Dictionary. 2nd ed. with a supplement by William A. Craigie. Oxford: Clarendon.

190 Sara M. Pons-Sanz

Coates, Richard. 1997. “The Scriptorium of the Mercian Rushworth Gloss: A Bilingual Perspective”. Notes and Queries 242.453–458. Dance, Richard. 2003. Words Derived from Old Norse in Early Middle English: Studies in the Vocabulary of the South-West Midland Texts. (= Medieval and Renaissance Texts and Studies 246). Tempe: Arizona Centre for Medieval and Renaissance Studies. Drago, Adriana. 1977. “Le glosse anglosassoni del Durham Ritual”. Studi di filologia e di literatura tedesca in onore di Nicola Accolti Gil Vitale ed. by Giancarlo Bolognesi & Giorgio Sichel, 35–62. Florence: Olschki. Fischer, Andreas. 1989. “Lexical Change in Late Old English: From œ¯ to lagu”. The History and the Dialects of English: Festschrift for Eduard Kolb ed. by Andreas Fischer, 103–114. (= Anglistische Forschungen, 203.) Heidelberg: Carl Winter. Granovetter, M. 1983. “The Strength of Weak Ties”. American Journal of Sociology 78.1360–1380. Hadley, D. M. 2000. The Northern Danelaw: Its Social Structure, c. 800–1100. London: Leicester University Press. Hall, D. J. 1984. The Community of Saint Cuthbert: Its Properties, Rights and Claims from the Ninth Century to the Twelfth. Ph.D. dissertation, University of Oxford. Haugen, Einar. 1950. “The Analysis of Linguistic Borrowing”. Language 26.210–231. Hill, D. 1981. An Atlas of Anglo-Saxon England. Oxford: Blackwell. Hines, John. 1991. “Scandinavian English: A Creole in Context”. Language Contact in the British Isles: Proceedings of the Eighth International Symposium on Language Contact in Europe, Douglas, Isle of Man, 1988 ed. by P. Sture Ureland & George Broderick, 403–427. (= Linguistische Arbeiten, 238.) Tübingen: Niemeyer. Hofmann, Dietrich. 1955. Nordisch-englische Lehnbeziehungen der Wikingerzeit. (= Bibliotheca Arnamagnæana, 14.) Copenhagen: Munksgaard. Holthausen, F., ed. 1934. Altenglisches etymologisches Wörterbuch. (= Germanische Bibliothek: Reiche 4: Wörterbucher, 7.) Heidelberg: Carl Winter. Jones, Charles. 1967. “The Functional Motivation of Linguistic Change: A Study of the Development of the Grammatical Category of Gender in the Late Old English Period”. English Studies 48.97–111. ———. 1970. “Some Features of Determiner Usage in the Old English Glosses to the Lindisfarne Gospels and the Durham Ritual”. Indogermanische Forschungen 75.198–219. ———. 1988. Grammatical Gender in English: 950 to 1250. London: Croom Helm. Kastovsky, Dieter. 2000. “Inflectional Classes, Morphological Restructuring, and the Dissolution of Old English Grammatical Gender”. Gender in Grammar and Cognition: I Approaches to Gender; II Manifestations of Gender ed. by Barbara Unterbeck et al., 709–727. (= Trends in Linguistics: Studies and Monographs, 124.) Berlin: Mouton de Gruyter. Lang, James. 1984. “The Hogback: A Viking Colonial Monument”. Anglo-Saxon Studies in Archaeology and History 3.85–176. Lefebvre, Claire. 1984. “Grammaires en contact. Definition et perspectives de recherche”. Revue québecoise de linguistique 14.11–47. Lindelöf, Uno. 1893. “Beiträge zur Kenntnis des Altnorthumbrischen”. Mémoires de la Société Néo-Philologique a Helsingfors 1.219–302.

A sociolinguistic approach to the Lindisfarne and Rushworth Gospels

———. 1901. Die südnorthumbrische Mundart des 10. Jahrhunderts: Die Sprache der sog. Glosse Rushworth2. (= Bonner Beiträge zur Anglistik herausgegeben von Prof. Dr. M. Trautmann, 10.) Bonn: P. Hanstein. Markus, Manfred. 1988. “Reasons for the Loss of Gender in English”. Luick Revisited: Papers Read at the Luick-Symposium at Schloß Liechtenstein, 15.–18. 9. 1985 ed. by Dieter Kastovsky, Gero Bauer & Jacek Fisiak, 241–258. (= Tübinger Beiträge zur Linguistik herausgegeben von Gunter Narr, 288.) Tübingen: Gunter Narr. MED = Middle English Dictionary. 1952–2001, ed. by Hans Kurath et al. Ann Arbor: University of Michigan Press. Millar, Robert McColl. 2000. System Collapse, System Rebirth: The Demonstrative Pronouns of English, 900–1350 and the Birth of the Definite Article. Oxford: Peter Lang. Milroy, James & Lesley Milroy. 1985. “Linguistic Change, Social Network and Speaker Innovation”. Journal of Linguistics 21.339–384. Morris, Christopher D. 1977. “Northumbria and the Viking Settlement: The Evidence for Land-Holding”. Archaeologia Aeliana 5th ser. 5.81–103. ———. 1981. “Viking and Native in Northern England: A Case Study”. Proceedings of the Eighth Viking Congress, Århus, 24–31 August 1977 ed. by Hans Bekker-Nielsen, Peter Foote & Olaf Olsen, 223–244. (= Mediaeval Scandinavia: Supplements, 2.) Odense: Odense University Press. ———. 1984. “Aspects of Scandinavian Settlement in Northern England: A Review”. Northern History 20.1–22. OED = Oxford English Dictionary. 1989. 2nd edn, ed. by John A. Simpson & Edmund S. C. Weiner. Oxford: Oxford University Press. Peters, Hans. 1981. “Zum skandinavischen Lenhngut im Altenglischen”. Sprachwissenschaft 6.85–124. Pons-Sanz, Sara María. 2000. Analysis of the Scandinavian Loanwords in the Aldredian Glosses to the Lindisfarne Gospels. (= Studies in English Language and Linguistics: Monographs, 9.) Valencia: Department of English and German Philology (University of Valencia). Rogers, Everett M. & F. Floyd Schoemaker. 1971. Communication of Innovations: A Cross Cultural Approach. 2nd edn. New York: Free Press; London: Collier-Macmillan. Ross, Alan S. C. 1936. “Sex and Gender in the Lindisfarne Gospels”. Journal of English and Germanic Philology 35.321–330. ———. 1937. Studies in the Accidence of the Lindisfarne Gospels. (= Leeds School of English Language: Texts and Monographs, 2.) Leeds: School of English Language (University of Leeds). ———. 1979. “Lindisfarne and Rushworth One” Notes and Queries 224. 194–199. Ross, A. S. C., E. G. Stanley & T. J. Brown. 1956–60. “Part I: Some Observations on the Gloss and the Glossator”. Evangeliorum Quattuor Codex Lindisfarnensis: Musei Britannici Codes Cottonianus Nero D.iv ed. by T. D. Kendrick et al., vol. II, book 2, 5–33. Oltun: Urs Graf. Samuels, M. L. 1989. “The Great Scandinavian Belt”. Middle English Dialectology: Essays on Some Principles and Problems ed. by A. McIntosh, M. L. Samuels & M. Laing, 106–115. Aberdeen: Aberdeen University Press. Serjeantson, Mary S. 1935. A History of Foreign Words in English. London: Routledge and Kegan Paul.

191



192 Sara M. Pons-Sanz

Skeat, Walter W., ed. 1871–1887. The Holy Gospels in Anglo-Saxon, Northumbrian, and Old Mercian Versions, Synoptically Arranged. Cambridge: Cambridge University Press. Stein, Dieter. 1986. “Old English Northumbrian Verb Inflexion Revisited”. Linguistics across Historical and Geographical Boundaries: In Honour of Jacek Fisiak on the Occasion of His Fiftieth Birthday ed. by Dieter Kastovsky & Alexander Szwedek, vol. I, 637–650. (= Trends in Linguistics: Studies and Monographs, 32.) Berlin: Mouton de Gruyter. Thomason, Sara Grey & Terence Kaufman. 1988. Language Contact, Creolization and Genetic Linguistics. Berkeley & London: University of California Press. Townend, Matthew Owen. 2002. Language and History in Viking Age England: Linguistic Relations between Speakers of Old Norse and Old English. (= Studies in Early Middle Ages, 6.) Turnhout: Brepols. Vennemann, Theo. 2002. “Key Issues in English Etymology”. Sounds, Words, Texts and Change: Selected Papers from 11ICEHL, Santiago de Compostela, 7–11 September 2000 ed. by Teresa Fanego, B. Méndez-Naya & Elena Seoane, 229–252. (= Current Issues in Linguistic Theory, 224.) Amsterdam: John Benjamins. de Vries, Jan, ed. 1961. Altnordisches etymologisches Wörterbuch. Leiden: Brill. Watts, Victor. 1988–1989. “Scandinavian Settlement Names in County Durham”. Nomina 12.17–63. ———. 1995. “Northumberland and Durham: The Place-Name Evidence”. Scandinavian Settlement in Northern Britain: Thirteen Studies of Place-Names in their Historical Context ed. by Barbara Crawford, 206–213. London: Leicester University Press.



Haplology in English adverb-formation Amanda Pounder University of Calgary

1.

Introduction

One of the phenomena regularly cited as an example of haplology is zeroformation in the English derived adverb, as in (1), just where there is a base ending in the sequence /li/ (Stemberger 1981, Menn and McWhinney 1984). (1) Adj daily Æ Adv daily a. Fred undertook his daily chores with grace. b. Fred mopped the stairs daily.

In this case, the /li/-sequence in the adjective base arises as a result of affixation of ly, deriving an adjective from a noun. Further, adverb formation is avoided altogether in Standard English with some adjective bases derived from a noun by means of ly, as shown in (2): (2) a. b. c. d.

Fred was a friendly giant. ?Fred waved the mop at us friendlily.

*Fred waved the mop at us friendly. Fred waved the mop at us in a friendly manner.

Fowler (1946:327) admits e.g. she nodded queenly, which corresponds to friendly in the example in (2c), but just fifty-odd years later, it seems barely, if at all, acceptable in Standard English. When the /li/-sequence does not correspond to the suffix ly, forms like those in (3) are acceptable to some speakers at least. (3) sillily holily jollily

¨ silly ¨ holy ¨ jolly

194 Amanda Pounder

However, both of these types are and have long been the object of comment in grammars and elsewhere, as illustrated in (4). In (4a), for example, godlily and holily are treated as equally offensive, despite the difference in structure. (4) a.

Bayly (1772): but if the adjective end in ly, as godly, holy, the adjective is used adverbially, the ear not admitting the repetition godlily, holily … b. Fogg (1796): Adjectives that end in ly must not be made into adverbs by the addition of ly, but the phrase in such a manner must supply the place of an adverb.

To the data involving current English (1)–(3), we can add examples from earlier stages of Modern English whereby in the syntactic context of two adverbs in sequence or in conjunction, only one undergoes ly-affixation, as in (5) and (6). (5) adverb adverb:

I live exceedinge chargeablie (Copley 1572 in Christie 1897: 3) talked exceedingly civil (Woodforde 5.6.1778 in Beresford 1924–31)

(6) adverb and adverb:

and it yet goes on faire and softly (Osborne 1653 in Smith 1928: 7)

This paper will begin by considering the question of whether ‘haplology’ constitutes an apt term for referring to the phenomena observed for adverbformation in English, and will show that they may rather be interpreted as paradigmatic selection, i.e. the selection of an alternative morphological operation in the presence of potential identity. Section 2 integrates the phenomena into a Processes and Paradigm framework (Pounder 2000). This is a framework which features the construct of the morphological paradigm as an organizing principle of the lexicon and the morphological component. Section 3 of the paper sketches the historical development of the deadjectival wordformation paradigm in English. Morphological haplology and paradigmatic selection depend on speakers’ evaluation of euphony or lack thereof; such an evaluation can be considered typical of a pre-standard, orally-focused phase of a language. We can contrast this with considerations of consistency of morphological marking and elimination of variants associated with a language standard based on written language. We will show that while in Modern Standard English grammatical considerations typical of written language have generally come to prevail in adverb-formation, euphony still plays a role, so that its influence cannot be ruled out, even in a mature standard.

Haplology in English adverb-formation 195

2. Haplology and English adverb-ormation 2.1 Haplology and paradigmatic selection Haplology can be defined in a traditional manner as the avoidance of repetition of phonological material, ranging from a segment to one or more syllables, within a word. Standard haplology examples, presented as purely phonological, always seem to involve morphology, as the repetition of material comes about through compounding or affixation, as in (7). (7) Homeric Greek: amphiphoreús Æ amphoreús “two-handled pitcher” compound stem amphi(Hock 1986: 109)

We will be focusing on a type of morphological haplology in which an affix identical to part of the base (i.e. to another affix in sequence or to part of the lexical stem) does not appear. Within morphological haplology, we can again distinguish two types, depending on the extent to which the morphology of the base is involved. That is, the process can be sensitive to the presence of an identical affix only, as in the case of Ancient Greek, as shown in (8), or on the other hand may be triggered by a phonological sequence irrespective of status, as in Swedish (9). éphthimai ephthíme¯n perf.mid.1sg pluperf.mid.1sg b. lélumai elelúme¯n perf.mid.1sg pluperf.mid.1sg c. elpídzo¯ ¯elpika pres.indic.1sg perf.indic.1sg (based on Stemberger 1981: 799)

(8) a.

< phthi“perish” < lu“redeem” < elpid“hope”

In (8a), we see that the e-prefix of the pluperfect (8b) does not appear when an e-prefix is already present; however, if the lexical base begins with an /7/ (8c), then e-prefixation, here marking the perfect, takes place notwithstanding. (9) början början (*börjanen) sg def.sg “beginning” (based on Stemberger 1981: 798)

There are other options in response to the potential repetition of sequences. Given the appropriate phonological and possibly morphological conditions, an

196 Amanda Pounder

alternative morphological process may be applied to the base. In (10a) we see that an alternative suffixation has been used. Here, the suffix corresponding to “definite” and “plural” is na, rather than the default en in the inflection of neuter nouns. (10) a.

barn: barn barnen “child” pl def.pl b. äpple
: äpplen äpplena “apple” pl def.pl (based on Stemberger 1981: 801)

Where an alternative process is chosen, as in (10b), we can speak of paradigmatic selection, meaning that where the base meets specified criteria, a morphological operation from a set of available alternatives is chosen. In this case, the suffix na is available, not only for bases in en, but for other base types (common gender nouns, plural bases ending in /Vr/), where identity avoidance does not come into play. Let us stress here that paradigmatic selection and morphological haplology are to be considered distinct phenomena, although with identical ultimate motivations in these cases. Returning now to English adverb-formation, we can propose the cases in (1), (6), and (7) as candidates for morphological haplology. Namely, there seems to be a clear avoidance of ly-suffixation where this would involve repeating a /li/ syllable or ly suffix. Given examples like (1), it would seem that it is the suffix -ly that is not appearing, rather than a sequence from the base, although this cannot be determined with certainty, as there are no examples of haplology under near-homophony merely. However, we must also consider the possibility that we are, rather, dealing with paradigmatic selection. The paradigmatic alternative would be a conversion operation, as conversion is an active formal process in the English word-formation system in general (11), and within adverb formation in particular, as shown in (12). (11)

sugarN coverV tidyADJ

Æ Æ Æ

sugarV coverN tidyV

(12)

slowADJ fastADJ terribleADJ

Æ Æ Æ

slowADV fastADV terribleADV

The latter two examples from (12) belong, of course, to informal registers and non-standard dialects, but are part of the repertoire of probably all speakers.

Haplology in English adverb-formation 197

Beyond this, the speaker has the choice of moving to serial compounding as an alternative process, again in informal registers and non-standard dialects (the elements in question being -wise and -like), or a circumlocution (in a XADJ manner, for example). Let us consider briefly available accounts of haplology to apply in the domain of English adverb-formation. Stemberger (1981), Menn and McWhinney (1984), and Yip (1998) all reject, for various aesthetic and empirical reasons, an interpretation of haplology as deletion of affixed material: why add something only to delete it? They therefore have to account for the fact that the semantic or category change (Adj Æ Adv) is added. In Stemberger’s analysis, for example, a template-like rule (‘an English adverb ends in ly’) vacuously applies to the base. (13) and (14) show how this works: the morpheme responsible for adverb formation has been added, and is in some way present, but surface material maps onto both it and the morpheme responsible for adjective formation (13) or the base itself (14), resulting in what Stemberger calls ambimorphemicity. We shall return to an evaluation of this account after looking at the paradigmatic selection account. (13)

Adv Adj

M

N

M

day

ly

(based on Stemberger 1981: 814f) (14)

Adv Adj

M

silly

(based on Stemberger 1981: 814f)

2.2 Identity avoidance in a process and paradigm framework The essential components of a Process and Paradigm framework (Pounder 2000) are the commitment to the idea that morphological form and content are functions of the base, these functions represented as rules and mapped onto

198 Amanda Pounder

each other in operations which include the condition of their application; a separation of morphological form and content, distinct from form-meaning relations in lexemic units; a morphological component distinct from lexicon, syntax, and phonology, which contains the repertoire of rules, rule elements (affixes), and operations organized paradigmatically. The lexicon likewise has a paradigmatic or network structure, rather than a list structure; however, it is not a dynamic or creative component. In this framework, the morphological paradigm is defined as the set of possible paths departing from a base to produce correct word-forms (for inflection) or lexemes (for word-formation). These paths are determined by the operations and the conditions on them; some lexemic items may state requirements for particular choices of path in the lexicon. As indicated by the definition, both inflection and word-formation are considered to be fundamentally paradigmatic in nature, and the paradigm in each case is based on the same basic principles. Whereas the paradigm is a derivative construct, in the sense that it is defined by and can be deduced from the operations, it has a distinct raison d’être
: the paradigmatic form-meaning relations of synonymy and polysemy, defined as commonality of semantic function and form rule respectively, help define the morphological system of a given language. In the present case, we need to consider what elements are required for the subpart of the English word-formation system responsible for deadjectival adverb formation. We will restrict ourselves here to derivation. Let us propose the following for Modern English: (15) Rule Element:

1.

· ly
; FR1; … Ò

(15) shows a sign representation for the one affix available in the modern English system. The second side of the sign is the semantic representation; as it is not considered in this framework that affixes ‘have’ meaning in the way that lexical units (lexemes, word-forms) do, the content here is an indexing of the form rule, call it Form Rule 1 here, to which the affix belongs. The third side of the sign contains relevant conditions of the affix’s use, which we will ignore here. (16) Form Rule: 1. · X ≈ ly
; ‘FR1’; N, Adj … Ò 2. · X Æ X; ‘FR2’; N, V, Adj … Ò

The modern system contains two form rules used in the formation of adverbs (17). The first is available to noun bases for adjective formation as well as to adjective bases, as shown in the third component of the rule. As the affixation is not tied to content here, only one rule for both base types is required. The

Haplology in English adverb-formation 199

first component of the rule expresses that ly is affixed to the base in an appropriate way, while the second expresses the abstract content of the rule, which we will not explore here. Form Rule 2 is similarly available to different base types, and enjoys frequent use in English. Its form side expresses that the output is formally identical to the input base. (17) Semantic Rule:

1. · Z(‘X’); ‘SR1’; N, V, Adj … Ò

There is one semantic rule in this subsystem, namely a zero function of the meaning of the base (17): the lexical meaning of the output is identical to that of the input base. As in the case of the rules we have seen thus far, it is not particular to deadjectival adverb-formation, but rather is available to different base types. (18) Syntactic Rule:

1. · ΣAdj Æ ΣAdv; ‘ΣR1’; Adj … Ò 2. · ΣAdj Æ ΣAdv-*preV; ‘ΣR1’; Adj… Ò

Finally, the subsystem contains two syntactic rules, nearly identical, that are both responsible for the category change and all that that implies (e.g. syntactic distribution of adverbs). The form side of Rule 1 can be read ‘The syntactics [i.e. the lexico-syntactic and combinatory properties] of an adjective become the syntactics of an adverb’. As expressed by this, the rule is available only to adjective bases. Rules 1 and 2 clearly overlap, but the latter is more restricted: adverbs are created which may not appear in preverbal position, although there are no other syntactic restrictions. (19) Operations 1. · X ≈ ly
; ‘FR1’; … Ò ; · Z(‘X’); ‘SR1’; … Ò · ΣAdj Æ ΣAdv; ‘ΣR1’; …Ò

‘Op1’;

sc: Adj, *X = /x + li/… oc: Slot I; Resyllabify

Operation 1 (19) maps the rule for affixing ly onto that providing the zero semantic function and onto that providing the category change from adjective to adverb, and this mapping is expressed in the first component of the rule. The content side of the operation includes an indication of its productivity and status within the system; here that will mean high productivity, among other things. The third component of the operation has two parts: the stem conditions (sc) state conditions on the stem that qualify it for undergoing the operation, and the operation conditions (oc) state how the operation is to take place. In this case, the operation applies to adjective bases (which narrows down the application domain of the formal and semantic rules); stem conditions may

200 Amanda Pounder

also be phonological or semantic in nature, or refer to a subclass of a lexicosyntactic category. Not all speakers have the restriction whereby bases ending in /li/ may not undergo the operation; others may have the variant shown in (20). Speakers using Operation 1a will be able to produce sillily, uglily, and so on, whereas speakers using Operation 1 will not (and this does not mean that both operations may not figure in the repertoires of some speakers, or of a speaker at different times in his life). The operation is applied in ‘Slot I’, that is, a suffixation to the base occurs. (20) 1a.

· X ≈ ly
; ‘FR1’; … Ò · Z(‘X’); ‘SR1’; … Ò · ΣAdj Æ ΣAdv; ‘ΣR1’; …Ò

; ‘Op1’;

sc: Adj, *X = x + li… oc: Slot I; Resyllabify

‘Resyllabify’ indicates that in the phonology there must be resyllabification and degemination of /ll/, this occurring when a base ends in a nucleic /l/, as in able, simple, double
: the result will be ably, simply, doubly, and not e.g. *doublely. (21) 2.

· X Æ X; ‘FR2’; … Ò ; ‘Op2’; · Z(‘X’); ‘SR1’; … Ò · ΣAdj Æ ΣAdv-*preV; ‘ΣR1’; …Ò

sc: Adj; l.c.: Time Dist oc: …

Operation 2 is a conversion operation and maps the rule producing a formal identity onto the same semantic rule as in Operation 1, and onto the more restricted syntactic rule. The content of the operation includes an indication that, in the general case (that means where it is not required by the lexical entry of the base, or where it is chosen as an alternative to Operation 1 for which the base in question has failed to qualify), it is assigned to an informal register and is thus stylistically marked. The stem conditions state that it applies to adjective bases, and that one of the eligible lexical-semantic classes is Time Distribution adjectives (the use of lexical semantic-classes does not represent a restriction to these only). (22) Subparadigm · X ≈ ly
; ‘FR1’; … Ò · Z(‘X’); ‘SR2’; … Ò · ΣAdj Æ ΣAdv; ‘ΣR1’; …Ò

; ‘Op1’;

sc: Adj, *X = x + li… oc: Slot I; Resyllabify

· X Æ X; ‘FR2’; … Ò · Z(‘X’); ‘SR1’; … Ò · ΣAdj Æ ΣAdv-*PreV; ‘ΣR1’; … Ò

; ‘Op2’;

sc: Adj; l.c.: Time Dist oc:

XAdj

Haplology in English adverb-formation 201

In (22) we have a portion of the morphological component of modern English. A base AdjX enters into the paradigm and has a choice of two paths, one leading to Operation 1 and one leading to Operation 2. As the content sides of the operations are identical, we have a case of paradigmatic synonymy (in spite of the fact that, in the general case, there will be a register difference). If the base ends in /li/ or ly, or is lexically specified for it, it may follow the path to Operation 2; otherwise, it undergoes Operation 1, if a stylistically neutral effect is desired. A dialect speaker or a speaker selecting the informal register may also select Operation 2 for any base as long as the syntactic restrictions are met. In this manner, the criteria for selection are expressed in the operations themselves: lexical selection is provided for, selection based on phonological structure of the base, syntactic position, and register. We also have lexical entries such as (23) meaning that if one wants to make an adverb, one has to choose Operation 2. (23) fast Adj; /fæst/; ‘fast’ …

Op.2!

With the representations of rules and operations given in this section, we have been able to produce the correct forms, taking the possibilities proper to different registers or styles into account, while also allowing for different outcomes from one and the same base, as manifested by different speakers or the same speakers on different occasions. The operation representations take account, as far as is possible, of the range of criteria that can influence the choice of one operation over another: lexical selection, semantic type, phonological structure, register/style, and syntactic function/position. Purely phonological accounts are not able to do this. We can assume at the same time that the natural tendency to avoid identity is operative as part of self-monitoring or -evaluation mechanisms and will be evident in speaker utterances. Let us now contrast the paradigmatic selection account with haplology accounts. First of all, contra Stemberger (1981) and Menn and McWhinney (1984), given the independently necessary existence of Operation 2 responsible for conversion, there is no reason to suppose ambimorphemicity in cases like she runs 10 km daily. That is, seeing that daily is a legitimate adverb according to the word-formation system of Modern English, we do not need to suppose that the hearer/reader interprets the ly as being both adjectival, deriving the Time Distribution adjective from the Time Unit noun, and adverbial, mapping onto the category change. Secondly, the syntactic distribution of Time Distribution adverbs supports the paradigmatic selection approach. As indicated earlier,

202 Amanda Pounder

Operation 2 creates adverbs with a peculiar syntactic distribution: they may not stand before the verb. This distribution is illustrated in (24) and (25). (24) that was a pretty nice performance she sang beautifully well she ran fast Honest, I saw it happen! (25) *she fast ran *she beautiful sang

To this there are only three current exceptions, as far as I know, shown in (26). (26) He had long wished for a silver pocket knife. Max is likely coming with Judy. I sure wish he would show up.

Now, if adverbs like daily had only undergone the denominal adjective-forming operation, we would expect that they would be permitted the same syntactic distribution as ly-adverbs, that is, we would expect their phonological structure to qualify them for preverbal position. However, this does not appear to be the case. Time Distribution adverbs in -ly cannot appear in preverbal position, although other Time adverbs can (27). (27) He regularly mopped the floor. He often mopped the floor. *He daily mopped the floor. His mother had wanted him to donate to the Women’s Shelter, which he soon/often/regularly did. *His mother had wanted him to donate to the Women’s Shelter, which he yearly did.

Likewise, if daily and other Time Distribution adverbs were the result of deletions, we should expect these sentences to be acceptable. Time Distribution adverbs have the same distribution as adverbs derived without ly. It seems clear, then, that these are adverbs converted from derived adjectives: they are more complex than these. Let us now consider adverb form in conjunction constructions. First of all, textual evidence shows that there are various possibilities (28). (28) a.

by my greedie adversarie on whom God’s iustice did notoriouslie and speedilie appeare (Copley 1573 in Christie 1897: 19) b. Mr. Cox spoke … most elegantly and genteelly (Woodforde 12.10.1767 in Beresford 1924–31)

Haplology in English adverb-formation 203

c. the great art of living easy and happy (Boswell 1762 in Pottle 1950:63) d. it strikes smooth and gently upon the sense (Collier 18th c., from Knorrek 1938: 103) e. she talks very sensible and well (Woodforde 21.5.1778 in Beresford 1924–31)

In (28a,
b), we see that both adverbs have undergone ly-affixation; in (c), however, neither does. (28d) shows the distribution that Knorrek (1938) discusses, wherein the first adverb has undergone conversion rather than ly-affixation. I have found almost no textual evidence so far for a pattern in which the first adverb is ly-affixed but the second converted; two examples are given in (29). (29) Everything was done decently, handsome and well (Woodforde 22.5.1771 in Beresford 1924–31) … that they behave themselves dutifully and obedient1 (Evelyn, 1704)

Clearly, the choice is stylistic; there is no grammatical imperative in this case. This fact provides some support for the paradigmatic selection analysis: each adverb is formed separately. That we are indeed talking about paradigmatic selection here, rather than deletion of a pre-selected ly sequence or affix, receives further support from German data. Up into the nineteenth century, we observe the optional inflection of just one conjoined attributive adjective before a neuter noun (nom/acc sing), as in (30). (30) ein roth-Ø und weisses Gesicht “a red and white face” (Modern Standard German: ein rotes und weißes Gesicht) (Adelung 1782: 637)

As in the case of the adverbs, an unsuffixed form of the attributive adjective is legitimate within the system for this type of noun (rot Gesicht is all right) at this time. This is different from the contemporary case where, when two complex forms are conjoined, only one — the second — bears the affix that in principle both are intended to receive (31). (31) brot- und ehrgeizlos “bread- and ambitionless” Natürlich- und Unnatürlichkeit “natural- and unnaturalness”

1.This example was supplied to me by Teresa Fanego Lema (source: Century of Prose Corpus).

204 Amanda Pounder

Here, punctuation and intonation mark the gap; the structures are not legitimate or interpretable without the second conjunct (*brot-). In these cases, one could indeed talk of deletion or ellipsis; the los or keit must be supplied by the speaker or hearer. Not so in the case of the German adjectives or the English adverbs. It therefore seems that paradigmatic selection is a better account of the phenomena shown in (28). Finally, let us consider adverbial modification, where one adverb is modified by another. As we saw in the examples in (5), we have the full set of options realized in Early Modern English and earlier Modern English: (6) adverb adverb

I live exceedinge chargeablie (Copley 1572 in Christie 1897: 3) talked exceedingly civil (Woodforde 5.6.1778 in Beresford 1924–31)

talked exceeding civil and I live exceeding chargeable are certainly possibilities likewise, as are exceedingly civilly and exceedingly chargeably. In these cases, the choice is a stylistic one, whereby the attempt to avoid repetition of the ly may occur. Beckmann (1880) claims that ly-affixation of the modified adverb is far more usual; this is not borne out in Woodforde, rather the contrary, but may be statistically valid across texts. Again, exceeding is acceptable at this time as an adverb (she is exceeding civil), and so is civil (she behaved very civil towards us): there is no more reason than in previous cases to assume anything other than paradigmatic selection. In all cases we have described, then, an analysis of identity avoidance through paradigmatic selection of operation rather than haplology as described in familiar sources seems possible and preferable.

3. Historical sketch of English adverb formation 3.1 Principal developments The principal features of the deadjectival adverb formation system of English develop as follows (for details, see Pounder 2001, Uhler 1926 (Old English), Nevalainen (1994, 1997), (Early Modern English)). In Old English, the default suffix is e (heard, hearde “hard”, “hard”; ryght, ryghte “right”, “right, rightly”). In later Old English, an additional suffix lic, normally used in denominal adjective derivation, may precede e-suffixation (soþ “true”, soþe or soþlice “truly”). The

Haplology in English adverb-formation 205

productivity of lic-suffixation in this role increases; full synonymy with simpler adverbs is the usual result (Uhler 1926). In late Old English and Middle English, the increasing weakening of vowels in unstressed syllables affects the e-suffix in adverbial derivation, although this e is maintained for some time (Minkova 1991, Duggan 1994). It would then seem that ly-affixation becomes reanalyzed as adverbial formation. With the loss of e-suffixation, adverb forms are reinterpreted, leading to the conversion operation (Operation 2). In written English, ly-suffixation appears to be the most productive process, as it is textually the most frequent. From Middle English to Early Modern English, increasing restrictions on the syntactic distribution of converted adverbs can be observed. In Standard English into the nineteenth century, this corresponds to increasing social stigmatization of converted adverbs, the loss of most free use in formal register, and the loss of choices like exceeding civilly/exceedingly civil; smooth and gently. In these cases, the principle of identity avoidance loses over considerations deriving from ‘standard values’. 3.2 Systemic change in the English adverb-formation system In this section, we will show how the system changes, using the representation format for operations outlined in 2.2. (32) shows the deadjectival adverb paradigm in the first stage of Old English: there is just one operation at this time. (32) Old English I: XAdj

· X ≈ e
; ‘FR1’; … Ò · Z(‘X’); ‘SR1’; … Ò · ΣAdj Æ ΣAdv; ‘ΣR1’; … Ò

; ‘Op1’;

sc: Adj … oc: Slot I …

· X ≈ lic; ‘FR2’; … Ò · Z(‘X’); ‘SR1’; … Ò · ΣAdj Æ ΣAdj; ‘ΣR2’; … Ò

; ‘Op3’;

sc: Adj … oc: Slot I …

· X ≈ e
; ‘FR1’; … Ò · Z(‘X’); ‘SR1’; … Ò · ΣAdj Æ ΣAdv; ‘ΣR1’; … Ò

; ‘Op1’;

sc: Adj … oc: Slot I …

(33) Old English II:

XAdj

The subparadigm in (33) looks at first glance similar to that posited for contemporary English. However, one subtle difference is crucial. Note the syntactic rule

206 Amanda Pounder

in ‘Operation 3’. This operation forms synonymous adjectives, not adverbs, from adjectives. Adjectives produced from it do serve, like other adjective bases, as inputs for this subparadigm, and thus may undergo Operation 1. It is not necessary that the adjective in question be in frequent use as an adjective for it to undergo subsequent derivation in this paradigm (although some are). Under this scenario, the adverb-formation subparadigm itself will not have changed for Old English II. Paradigmatic relations, on the other hand, will have changed, with a new set of synonymy relations emerging. There may be synonymous adjectives sharing the same root, and there may be two synonymous adverbs likewise sharing the same root. (34) Middle English/Early Modern English after disappearance of final schwa on adverbs · X ≈ ly
; ‘FR1’; … Ò · Z(‘X’); ‘SR1’; … Ò · ΣAdj Æ ΣAdv; ‘ΣR1’; … Ò

; ‘Op1’;

sc: Adj … oc: Slot I …

· X Æ X; ‘FR2’; … Ò · Z(‘X’); ‘SR1’; … Ò · ΣAdj Æ ΣAdv; ‘ΣR1’; … Ò

; ‘Op2’;

sc: Adj; l.c.: Time Dist oc:

XAdj

At this stage, there is little difference in the distributions of these operations; their content sides will express that Operation 1 is more productive, although it will not express any register difference. Any ‘identity avoidance’, except that relative to the Time Distribution class, will be sporadic and stylistic; speakers may choose Operation 2 if they do not like the sound of two ly’s, but they may prefer Operation 1 in any given instance (35). Resyllabification may sporadically occur with Operation 1, but need not (36). ly-less adverbs rarely appear in preverbal position, but occasionally do so (37). (35) pray therefore see that this notice be as timelyly given them as you can (Pepys 1675 in Heath 1955: 33) and let the matter dye friendlyly (Pepys 1676/8 in Heath 1955: 59) (36) and howe simplely and sincerely I meane to proceed (Copley 1575 in Christie 189: 43) (37) your sendyng to Caster is evyll taken among my lordes folkes (J. Paston Jr. 1476 in Gairdner 1910: 150) You dayly and howerly soe comble me with … deeds of your worthyness (B. St. Michel 1672 in Heath 1955: 20)

Haplology in English adverb-formation 207

an attachment to have bin made upon the hundreds of pounds you have wrongfull in past made and taken to your use (Dr. T. Pepys 1664 in Heath 1955: 8) to Jentellmen …. hath bid as much to the full as it Can posible be worth (J. Pepys Sr. 1674 in Heath 195: 41)

As we move towards the present system, the status of Operation 2 changes, as there is more and more pressure to avoid it unless required, particularly in writing; the syntactic restriction, previously a strong tendency, is added. Here, then, we have seen that the most dramatic change to the system took place in Old English and Middle English. Since that time, changes have affected primarily the status and productivity of the conversion operation, and syntactic and phonological restrictions on Operations 2 and 1 respectively. We turn now to the role standardization has played in the history of deadjectival adverb formation in English.

4. Standardization and systemic change in the English adverb-formation system If we focus on the increasing restrictions on the conversion operation and its eventual social stigmatization, standardization automatically comes to mind. Generally speaking, there appear to be essential differences between the morphological systems of non- or pre-standard languages as opposed to standardizing or standardized languages (Pounder 2000). One of these differences is relevant here: the role of the lexicon is relatively more important in standardized systems, and there are more limits on productivity in an absolute sense, meaning that non-authorized people may feel inhibited from creating new words at need using productive operations. Speakers will ask ‘is that a word?’ and if it is not authorized by being in a dictionary, may protest its use or avoid it, and so on. Another, related, difference is a negative value attached to paradigmatic synonymy, which itself is a variant of the ‘avoid variation’ principle of standardization: if two (or more) derivations with the same root and the same semantic and syntactic rules exist, one must disappear or a lexical, semantic, stylistic, or other difference must distinguish them. And of course, one should not create or maintain paradigmatic synonyms. This is an expression of the famous ‘Blocking Principle’ for morphology, which is not a systemic language principle but a prescriptive principle of standard languages (Pounder 2000). The prescribed avoidance of paradigmatic synonymy did a great deal for

208 Amanda Pounder

the retirement of the conversion operation as a neutrally-valued option for the formal register. Stein (1997) has suggested that the nature of change may depend on whether we are talking about a standard or a non- or pre-standard language. For example, we might expect that change will be geared more towards written than spoken language in the case of standardization. The written standard language might require a consistent and explicit marker of lexical category in the case of adverbs. Non-standard change might reflect rhythmical and stress factors more than change in a standard, where presumably one would not be sensitive to such characteristics. Rohdenburg (1997) shows this to be the case for German dialect inflection. In the remainder of this paper, we will consider on the one hand the treatment of adverb-formation in English grammars, as well as the changes we have observed. A number of hypotheses are possible here. Firstly, we can propose that ‘standard’ change will require consistency of formation type (emphasis on lexicon, free choice devalued), independent of aesthetic considerations (identity, rhythm…). Secondly, ‘non-standard’ or ‘pre-standard’ change/paradigmatic selection will be based on spoken-language considerations (identity, rhythm…) What we find is that both types of change take place. Consistency of formation type is indeed striven after: to a great extent, in a formal register, the freedom of paradigmatic selection is lost in the contexts described in this paper. For example, in the case of adverb modifying adverb, in most cases we have no choice: we must select the ly-adverb, no matter whether that produces two ly-adverbs in sequence. Some stylists will suggest a new formulation (in an exceedingly civil manner rather than exceedingly civil or exceeding civilly). In the case of adverbs in conjunction, it is no longer acceptable to select the conversion operation for the first conjunct (smoothly and gently is the only option in formal register). Here we see typical developments to be expected under standardization. The comments of contemporary grammarians encourage and support this change (e.g. 38): (38) Don’t use adjectives when adverbs are required. Don’t say, for instance, This pear is uncommon good, but This pear is uncommonly good. (Censor:76)

On the other hand, contemporary grammarians seem to be extraordinarily sensitive to the aesthetics of adverb formation: they insist that the repetition of ly is unpleasant to the ear, and should be avoided (39). (39) a.

(= 4a) … but if the adjective end in ly, as godly, holy, the adjective is used adverbially, the ear not admitting the repetition godlily, holily. (Bayly 1772: 60)

Haplology in English adverb-formation 209

b. Certainly the ear, which will overrule judgement and theory, taught the use of some adjectives as adverbs without adverbial termination … (Bayly 1772: 61) c. ly is cut off from exceedingly when the next word ends in ly … They used me exceeding discreetly (Lennie 1815: 76) d. In the proper disposition of adverbs, the ear carefully requires to be consulted as well as the sense. (Lennie 1815: 76)

Eighteenth- and nineteenth-century grammarians therefore support, for the most part, the choice of the conversion operation or circumlocution for adjective bases like timely
; they do not, however, propose the consistent solution of *dailily (40). (40) The adverbial termination ly is improperly annexed to adjectives ending in ly. Instead of sillily, friendlily, holily etc., a circumlocution must be used. (Angus 1812: 220)

It is clear that the spoken — or declaimed — language is in the minds of the grammarians, even if the demands of written language have increased in importance. At the same time, our first hypothesis does seem to be supported by the data.

5. Conclusions This paper considered the avoidance of subsequent /li/-syllables and ly-suffixes in English, within the word and within constructions. Both haplology and paradigmatic selection accounts for the phenomenon were considered, and it was shown that the latter is preferable for a number of empirical and theoretical reasons (principally, the alternative operation is required independently, as is the need to make a selection in other situations). It was also shown that the system itself has changed little since Middle English, while the motivation for making a particular paradigmatic selection in a given situation has changed a great deal. The developments observed should be considered typical of the standardization process and its emphasis on the written language, but it is clear that spoken-language considerations are still valid and considered legitimate by grammarians.

210 Amanda Pounder

References Adelung, Johann. [1782]. 1971. Umständliches Lehrgebäude der Deutschen Sprache: Zur Erläuterung der Deutschen Sprachlehre für Schulen. 2 vols., facs. Hildesheim: Olms. Angus, William. 1812. A New System of English Grammar. Glasgow: Brash & Reid. Bayly, Anselm. 1772. A Plain and Complete Grammar of the English Language. London: J. Ridley. Beckmann, E. 1880. “Über die doppelformigen englischen Adjektiv-Adverbien”. Archiv für das Studium der neueren Sprachen and Literaturen 64: 25–27. Beresford, John, ed. 1924–31. Diary of a Country Parson. 5 vols. Oxford: Oxford University Press. Campbell, George. 1776. The Philosophy of Rhetoric. 2 vols. London. Censor (= Oliver Bunce). n.y. Don’t! A Manual of Mistakes and Improprieties More or Less Prevalent in Conduct and Speech. London: Field & Tuer. Christie, Richard, ed. 1897. Letters of Sir Thomas Copley. New York: Franklin. Duggan, Hoyt N. 1994. “The Role and Distribution of ly-adverbs in Middle English Alliterative Verse”. Loyal Letters: Studies on Mediaeval Alliterative Poetry and Prose ed. by L. A. J. R. Houwen and A. MacDonald. 131–154. Groningen: Forsten. Fogg, Peter. [1792/1796]. 1970. Elementa Anglicana. facs. Menston: Scolar Press. Fowler, Henry, ed. 1946. A Dictionary of Modern English Usage. 3rd edn. London: Oxford University Press. Gairdner, James, ed. 1910. The Letters of Samuel Pepys and his Family Circle. Oxford: Clarendon Press. Heath, Helen, ed. 1955. The Letters of Samuel Pepys and his Family Circle. Oxford: Clarendon Press. Hock, Heinrich H. 1986. Principles of Historical Linguistics. Berlin: Mouton de Gruyter. Knorrek, Marianne. 1938. “Der Einfluß des Rationalismus auf die englische Sprache”. Beiträge zur Entwicklungsgeschichte der englischen Syntax im 17. und 18. Jh. Breslau: Priebatschs Buchhandlung. Lennie, William. 1815. The Principles of English Grammar, Comprising the Substance of all the most approved English Grammars extant, briefly defined, and neatly arranged; with Copious Exercises in Parsing and Syntax. 3rd edn. Edinburgh: Oliver & Boyd. Menn, Lise & Brian McWhinney. 1984. “The Repeated Morph Constraint”. Language 60: 519–541. Minkova, Donka. 1987. “Adjective Inflexion Relics and Speech Rhythm in Late Middle English and Early Modern English”. Papers from the 5th ICEHL ed. by Sylvia Adamson et al. 313–336. Amsterdam: John Benjamins. Nevalainen, Terttu. 1994. “Aspects of Adverbial Change in Early Modern English”. Studies in Early Modern English ed. by Dieter Kastovsky. 243–259. Berlin: Mouton de Gruyter. ———. 1997. “The Processes of Adverb Derivation in Late Middle and Early Modern English”. Grammaticalization at Work: Studies of Long-Term Developments in English ed. by Matti Rissanen et al. 145–189. Berlin: Mouton de Gruyter. Pottle, Fred, ed. 1950. Boswell’s London Journal 1762–1763. New York: McGraw-Hill. Pounder, Amanda. 2000. Processes and Paradigms in Word-Formation Morphology. Berlin: Mouton de Gruyter.



Haplology in English adverb-formation

———. 2001. “Adverb-Marking in German and English: System and Standardization”. Diachronica 18: 301–358. Rohdenburg, Günter. 1997. “Grammatical Variation and the Avoidance of Stress Clashes in Northern Low German”. Taming the Vernacular: From Dialect to Written Standard Language ed. by Jenny Cheshire & Dieter Stein. 93–109. London: Longman. Smith, G. C. Moore, ed. 1928. The Letters of Dorothy Osborne to William Temple. Oxford: Clarendon Press. Stein, Dieter. 1997. “Syntax and Varieties”. Taming the Vernacular: From Dialect to Written Standard Language ed. by Jenny Cheshire & Dieter Stein. 35–50. London: Longman. Stemberger, Joseph. 1981. “Morphological Haplology”. Language 57: 791–817. Uhler, K. 1926. Die Bedeutungsgleichheit der altenglischen Adjektiva und Adverbia mit und ohne -lic (-lice). Heidelberg: Winter. Webster, Noah. [1789]. 1967. Dissertations on the English Language. facs. Menston: Scolar Press. Yip, Moira. 1998. “Identity Avoidance in Phonology and Morphology”. Morphology and its Relation to Phonology and Syntax ed. by Stephen Lapointe et al. Stanford: CSLI. 216–246.

211



Uses of Scottish place-names as evidence in historical dictionaries* Margaret Scott Scottish Language Dictionaries (formerly OED3)

Historical dictionaries provide a wealth of information which can often be very significant for place-name research. An investigation of the range and meaning of a place-name element is usually enhanced by the knowledge of the range and meaning of the equivalent lexical item. Conversely, place-name studies have an important contribution to make to historical lexicography. This is particularly true for the history of the Scots language, the earliest extant literary example of which dates from the late fourteenth century. However, toponymic material raises some difficult taxonomic questions for the lexicographer, not least because he or she must devise a method by which it may be decided whether a piece of evidence is treated as a word or as a name. This paper seeks to examine some of the lexicographical advantages which may be gained from a detailed consideration of placename evidence in the context of the Germanic languages of the British Isles. Examples focus particularly on Scottish material, and have been taken from the corpus of Germanic place-name elements in southern Scotland which I have compiled, and from the Third Edition of the Oxford English Dictionary (OED3). During the twentieth century, the toponymic corpus of England became established as a valuable resource for the early history of the English language. Many influential publications on English place-name vocabulary have been produced during the last eighty years by the English Place-Name Society, including reference works such as Smith (1956), a new edition of which is currently in progress at the University of Nottingham (Parsons & Styles with

*I am very grateful to Philip Durkin, Carole Hough and Joshua Pendragon for their comments on drafts of this paper.

214 Margaret Scott

Hough 1997, Parsons & Styles 2000). More recently, Gelling and Cole (2000) have examined a wide range of topographical place-name elements, and many of the detailed studies by Hough (e.g. 1995, 1998) and Kitson (e.g. 1997, 1998) have revealed words and meanings which were previously unidentified. Collectively, these findings have made a substantial contribution to our understanding of the vocabulary of Old and Middle English. Despite the ancient linguistic connections between the territories now known as England and Scotland, Scottish place-names have seldom been taken into consideration in studies of historical vocabulary, even though many Old and Middle English place-name elements have direct equivalents in Early and Middle Scots. One of the main reasons for this omission is that the Scottish Place-Name Survey is less advanced than its English counterpart, with much important material remaining unpublished. However, invaluable raw material can be found in the archives of the Scottish Place-Name Survey, and in Dixon (1947), Macdonald (1937) and Williamson (1942), three PhD theses which were written at the University of Edinburgh over fifty years ago. Of these, only Macdonald’s work on West Lothian names was published in book form (1941). Nevertheless, the value of Scottish place-name evidence has not been entirely overlooked. As Sir William Craigie pointed out in his introduction to the first volume of the Dictionary of the Older Scottish Tongue (DOST): The continuous Scottish record begins only with Barbour’s Brus in 1375, and the contemporary Legends of the Saints. Before that date the scanty evidence has mainly to be gleaned from the place-names and personal designations occurring in the Latin charters or other documents relating to the southeastern counties, or from legal terms employed in a Latin context in the early laws. These sources have been carefully examined for this evidence, which enables the Scottish form of a considerable number of common words to be traced back to the 12th or 13th cc. (DOST
: I, vii)

Craigie is perhaps a little unjust in his reference to the “scanty” pre-literary evidence. A surprising quantity of material is available for analysis, and it can be used for much more than simply identifying older examples of common words. Significantly, in the later volumes of DOST, the contribution of place-name material receives more recognition and is given a much more thorough and detailed treatment. In the discussion of Middle Scots feld, felde “field” for example, it is noted that a number of place-names dating from the pre-literary Scots period contain this element, including Wytefeld (c1200), Westfeld (1294),

Uses of Scottish place-names as evidence in historical dictionaries

Lyllochefylde (1293) and Prestfeld (1327) (DOST s.v. feld(e n.).1 Similarly, at the entry for hall “large and spacious building, esp. one which is the residence of a magnate”, the dictionary records early uses of the word in the names Blachall (1329) and Halton (1345–50) (DOST s.v. hall n.). In the entry for fidlar “fiddler”, the earliest evidence for the word is found in the place-name Fiddillariscroftis (1457), Fidlariscroft (1458), and the onomastic examples have been included in the entry as square-bracketed quotations at the beginning of the quotation paragraphs (DOST s.v. fidlar n.). A number of scholars have made a significant contribution to Scottish onomastics in recent years. W. F. H. Nicolaisen deserves special recognition, as he has used place-name evidence to examine a wide variety of linguistic issues from sound-change to ‘onomastic dialects’ (e.g. Nicolaisen 1976, 1993, 1995, 2001). Simon Taylor’s work has also considered many aspects of Scottish onomastics, and recently addressed aspects of the relationship between placenames and lexicography at the forty-fifth Conference of the Scottish Medievalists, which celebrated the completion of DOST (Taylor 2002). Other scholars including Barbara Crawford, Ian Fraser, Gillian Fellows-Jensen, Carole Hough and Doreen Waugh have made a significant contribution to Scottish placename studies. Consequently, the gap between Scottish and English toponymy is gradually being bridged. A particularly relevant example of this is found in Hough’s discussion of the Scottish place-name Bemersyde (Hough 1999). In literary sources the word that forms the first element of this name is attested with the meaning “trumpeter”, but she has shown that it was also used in a transferred sense to denote a type of bird, specifically the bittern. This usage can be found both in English place-names derived from West Saxon b¯ymere, including Bemerton in Wiltshire and Bemersley in Staffordshire, and in the Scottish place-name Bemersyde in Berwickshire, derived from Anglian be¯mere (Hough 1999). There are many other Scottish place-names which similarly deserve to be considered for their contribution to our knowledge of the early lexicon of the British Isles. In my own research, I have compiled a corpus of the Germanic elements so far identified in the place-names of southern Scotland. From this body of material, it is possible to identify many examples of terminology

1.In this paper I will follow the established convention of italicizing historical forms of place-names, and place-names which are ‘lost’, i.e. no longer in use. Please note that the modern place-names given in passages quoted from OED3 do not follow this convention and are also italicized.

215

216 Margaret Scott

recorded in place-names earlier than in literature, and terminology which is unattested or of low incidence in the literary corpus, for which the place-name evidence is often particularly significant. For example, Macdonald notes that the name Priest Mill in West Lothian is first recorded in 1534 in the form Breistmyln(e), which represents the Scots compound breist-mill (Macdonald 1941: 47). This is the Scottish equivalent of English breast-mill, which Wright defines as “a water-mill in which the water goes in at the side or breast to turn the wheel” (English Dialect Dictionary (EDD) s.v. breast n. 5). In later spellings, the initial consonant is changed by folketymology, with the form Priestmyln being recorded from the later sixteenth century (Macdonald 1941: 47). The word breist-mill is not recorded in either DOST or the Scottish National Dictionary (SND), and so the place-name evidence allows the identification of a previously unrecognised Scottish compound. According to the Oxford English Dictionary, the English form breastmill is attested from 1674, and therefore a significant ante-dating for this word is preserved in the earliest historical form of the name Priest Mill, recorded a hundred and forty years earlier (OED s.v. breast n. 11). One slightly earlier example of the English compound is shown by the place-name Breast Mill Beck in Lancashire, recorded as Bristmylbeck in 1526 (Parsons & Styles with Hough 1997: 22). Both of these ante-datings deserve to be mentioned in any detailed discussion of the compound breast-mill, and the locations of these places, in West Lothian and Lancashire, may also serve to emphasise the shared vocabulary of the south of Scotland and the north of England. Scottish place-names can also be used to identify elements that are unattested or rarely attested in the literary corpus. The Ayrshire name Privick, which survives only in the name of Privick Mill, is recorded from the late twelfth century as Preueic (a.1177; Fraser 1874: II,1),2 and later as Prevyk (1379; Stuart et al. 1878–1908: III, 31), Previck (1429; Nicolaisen 2001: 103), and is likely to contain the rarely attested Old English peru “pear tree” compounded with Old English wı¯c “(dependent) farm” (Scott 2000). Tarbolton in Ayrshire, first recorded as Torboultoun (a.1177), appears to have been formed by the addition of Gaelic torr “hill” to an earlier name from Old English *boðl-tu¯n “houseenclosure, house-farm”, a compound which is attested only in place-names.3

2.I am very grateful to T. A. Hendry for alerting me to the existence of the early form Preueic (a1177). 3.Simon Taylor, personal correspondence.

Uses of Scottish place-names as evidence in historical dictionaries 217

The same Old English compound can be found in the lost East Lothian name Botheltun, as noted by Nicolaisen, who has also drawn attention to the Hedderwick place-names, in East Lothian and Angus, which contain the unattested Old English *hæddre “heather” (Nicolaisen 2001:100, 102). OE *rydding “a clearing” is possible in the lost Dumfriesshire name Batemanridding (1275–1329), and OE *wincel “nook, corner” is likely in Winchburgh, West Lothian, first recorded as Wincelburgh in the late twelfth century (Williamson 1942: 217; Macdonald 1941: 44). In cases such as these, the Scottish place-name evidence expands the known geographical and historical spread of the elements, and deserves to be recognised by any comprehensive account of the words themselves. These examples also reveal one of the most difficult taxonomic issues which has to be addressed in any analysis of early Scottish Germanic material. If a Scottish place-name element is clearly related to an Anglian Old English word, and there is historical support for an early coinage date, then the Scottish context in which it is found should not be ignored. This may appear to be a statement of the obvious, but early Scottish place-name elements have often been labelled “Old English”, and considered as early evidence for the English language, with no further consideration as to the information they may provide for the Scots language. While Nicolaisen (1976, 2001) includes chapters on onomastic source languages including Gaelic, Scandinavian and Cumbric, he does not include a chapter on Scots. Earlier commentators have frequently ignored Scots elements. Dixon interprets all the Middle and Modern Scots names with the element side as evidence for the Old English element sı¯de “side, slope of a hill or bank”, with no further regard to the meaning of side at any point during the Scots period (Dixon 1941: 136). In doing so, he has overlooked the development of the element, including for example its later use as a suffix in the sense “direction, district, region”, attested from the late eighteenth century (SND s.v. side n. 4). Similarly, place-names first recorded in the Middle Scots period have sometimes been regarded as evidence for Middle English. The element index to Macdonald’s work includes an entry for Middle English noke “nook, corner of land” (1947: 135), even though all of the evidence presented there clearly relates to the Middle and Modern Scots element neuk, which can mean “nook”, “projection of land” and “outlying or remote place”, and deserves both on formal and semantic grounds to be considered within its Scottish context. Williamson treats the early forms of the Berwickshire name Bastle as evidence for Middle English bastle “strong tower”, even though the oldest recorded spelling Foulenbastell (1614) is clearly Middle Scots in date (Williamson

218 Margaret Scott

1942: 172). Bastle, in the sense “bastille, fortified tower; siege-tower”, is attested in Scottish literary sources from the sixteenth until the early twentieth century (Concise Scots Dictionary (CSD) s.v. bastile n.), and it is very likely that careful analysis of bastle as a place-name element would serve to enhance our knowledge of the semantic range of the word. Certainly the context of the evidence should not be ignored. The onomastic applications of bastle in Scotland may also be quite different from those found south of the border, for as Williamson herself notes, “there are no cases of the use of this word as a place-name element in Northern England” (Williamson 1942:172). Bastel “castle tower” is only recorded in English place-names in the form bastile in one lost Cheshire field name, the Bastile Room, recorded in 1537 (Parsons & Styles with Hough 1997: 53). In cases such as this, it would be very unfortunate if Scottish onomastic material was not considered together with contemporary Scottish lexical material. Part of this problem is caused by the common convention of presenting the evidence for each English place-name element under the headword which corresponds to that element’s Old or Middle English etymon. This system is problematic in the case of English place-names coined in the early modern period or later, and is an inadequate vehicle for Scottish Germanic material. It has long been recognised that Scots words are often quite distinct from their English counterparts, in terms of many factors including semantic range, phonology and etymology, and place-name elements are no different in this respect. While it is often useful to consider English and Scottish evidence together, the superficial similarity of English and Scottish evidence can also be deceptive, and for the sake of this caveat, if not for other reasons, it is important to recognise the contribution of Scottish place-names to the history of Scots as well as English. The history of lexis in contextual use is the topic of primary concern for the Third Edition of the Oxford English Dictionary (OED3), but onomastic material from many parts of the British Isles, and sometimes beyond, is being incorporated into the revised text where relevant. When a lexical item in OED3 is derived from a place-name, the place-name itself will not be etymologized, on the grounds that the main focus of the dictionary is lexical, not onomastic, and place-names and words cannot be assessed using identical criteria. Although semantically transparent names never become truly fossilised, and therefore still retain some lexical meaning, it is nevertheless necessary for compilers of dictionaries to attempt to draw as clear a line as possible between the lexicon and the onomasticon. In OED3, if the accepted or supposed etymology of a place-name throws

Uses of Scottish place-names as evidence in historical dictionaries 219

light on some aspect of the use of the relevant headword in contextual use, then the information which is significant to the use and development of the word may be given in discursive text. For example, in the following entry, the name of the town of Masham in North Yorkshire differs in pronunciation from the word Masham, which designates a specific breed of sheep, and it is useful to refer to the etymology of the place-name in order to explain this difference: Masham, n. Brit. /Ámas(6)m/, /Ámaw(6)m/, U. S. /Ámæw6m/ [< Masham, the name of a town in North Yorkshire, England (freq. explained as < an Old English personal name + -ham: see ham n.3). The pronunciation /Ámas(6)m/ has traditionally been used by residents for the name of the town; /Ámaw(6)m/ is prob. a spelling pronunciation.] A breed of sheep produced by crossing Wensleydale or Teeswater rams with Blackface or Swaledale ewes; a sheep of this breed… (OED3 s.v. Masham n.)

Place-name evidence which antedates lexical evidence is also being incorporated into etymological notes where appropriate, as in the entries for math n.1 and mereing n., shown below. In the case of math n.1, much of the phonological evidence for the later development of the word indicates that the Old English form may have had a doublet with a short vowel. However, the two Devonshire examples of the name Meeth show a reflex of the Old English form with a long vowel, and are therefore relevant to the word’s history: math, n.1 Now arch. and Brit. regional…Forms: OE mæþ, mæð, ME (in compounds) mad-, math-, mathe-, 15–16 mathe, 15– math, 17 meath. [Cognate with Old Frisian me¯th (neuter), Old Saxon ma¯d- (in the compound ma¯ddag mowing day; Middle Low German ma¯t (feminine)), Old High German -ma¯d (in the compound a¯ma¯d (neuter) aftermath; Middle High German ma¯t (neuter/ feminine), German Mahd (feminine); cf. also the weak feminine noun from the same base represented by Old High German ma¯da (Middle High German ma¯de) swathe), ult. < the Germanic base of mow v.1 The quantity of the root vowel in Old English has conventionally been taken to be long on the evidence of the cognates cited above; however, the subsequent history of the word in English suggests that Old English mæþ ¯ may have had a doublet form mæþ with short vowel: cf. Dutch mad, mat (neuter) swathe, and also the possible parallel of two Germanic ablaut grades (one with long root vowel, the other with short) clearly shown by the cognates of the Verner’s law variant mead n.2 Almost all of the phonological evidence for the word after the Old English period points to the reflex of Old English mæþ (or to an otherwise unparalleled shortening);

220 Margaret Scott

evidence for the reflex of Old English mæþ ¯ occurs in a few place names, as Meda (1086), Meðe (1175; now Meeth, Devon), la Methe (1249; now Meethe, Devon), and in the 18th-cent. form meath reported in the works of William Ellis (see Soc. Pure Eng. Tract (1945) LXIV. 103). Although attested as an element in compounds, the word is not found as a simplex in Middle English except in the place name Le Mathes in Cambridgeshire (1221; also as the second element of compound place names (e.g. Wetemathis (1302), now Wheatmath Field, Cambridgeshire)).] I. Simple uses. 1. A mowing; the action or work of mowing; that which may be or has been mowed; the portion of a crop that has been mowed. See also aftermath n., beadmath n., day math n., lattermath n., undermath n. In R. E. Zupko Dict. Eng. Weights & Measures (1968) defined s.v. as ‘in Herefordshire equal to approximately 1 acre.. or to the amount of land that a man could mow in a day’, but app. on the basis of an instance of day’s math (see quot. 1820 s.v. day math n.)… (OED3 s.v. math n.1)

In the etymology of mereing n., there is a discussion of some of the early forms of the name of Westmorland, as there is some possibility that the late thirteenth century form West-mering contains the same word. However, it is also possible that West-mering instead shows influence from the word mere, and because that is the explanation we have preferred, the quotation has not been included in the quotation paragraph: mereing, n… Forms: 15, 19– mering, 15– mearing, 16 meeringe, 17–18 meering, 18– mereing; Sc. (chiefly north.) pre-17 mearin, mering, meryne, pre-17 17– mearing, 17 meering, merin, 18– (now rare) mearen, miren; Irish English 18– mearing, meerin, merin, 19– marin, mearin’, mearn. [< mere v.2 + -ing1. App. unattested in Middle English, although cf. the following: c1275 (?a1200) LAõAMON Brut (Calig.) 4972 Nome him scupte þe king & hehte þene stan West-mering [c1300 Otho Westenering]; a muchel dæl londes þe þer lið abuten nom þe king to his hond & hæhte hit West-merelinge lond. The above quot. occurs in a passage on the naming of Westmorland; cf. Old English Westmo¯ringa land, lit. ‘the land of those living west of the (Yorkshire) moors’ (cf. -ing3). The form West-mering perh. shows alteration of the second element of the Old English word after mere n.2 (cf. merestone n., and the following early forms of the place name Westmorland
: Westmerland (1129–33), Westmerieland (1183), Westmeriland (1190), Westmereland (1237)), rather than an isolated early attestation of the present word.

Uses of Scottish place-names as evidence in historical dictionaries 221

The following occurrence of mæring ¯ in Old English in a Kentish charter has also been taken by some as an early attestation of this word, but the word division of the MS. is misleading, and Haðemæring, ¯ a place name based on a personal name H(e)aðumær ¯ (cf. -ing3), is a more likely reading: OE Bounds (Sawyer 546) in W. de G. Birch Cartularium Saxonicum (1893) III. 36 Of suð tune andlang broces to haðe mæringe ðonne on suð healfe of haðe mæringe to stoccum. The location app. corresponds to a field whose name is variously recorded in later documents as Helmering (1526), Hellmerings (1620), and Hemmings (19th cent.)…] I. Simple uses. 1. a. A boundary; = mere n.2 1. Also fig. Now chiefly Irish English… (OED3 s.v. mereing n.)

Some Scottish place-name material has also been incorporated into revised entries. One example is the entry for mickle a., pron., n., and adv. At sense 1.a. it is noted that the adjectival sense “great or large in size, bulk or stature” was used at one time to designate a particular place or building, and that this usage is now only retained in place-names. In the etymological note, examples of this type of name are given from both Scotland and England, including Mickleover in Derbyshire, Meikle Black Law in Berwickshire and Muckle Flugga in Shetland: mickle a., pron., n., and adv. Now regional (chiefly Sc. and Eng. regional (north.)) and arch… [Cognate with Middle Dutch me¯kel, Old Saxon mikil (Middle Low German (rare) me¯kel; the commoner michel is a borrowing from Middle High German), Old High German mihhil, mihhal, mihhel (Middle High German michel, early modern German michel), Old Icelandic mikill, mykill (masculine accusative mikinn, mykinn, neuter nominative and accusative mikit, mykit), Norn (Shetland) mege (neuter), Norwegian (Nynorsk) mykjen, Norwegian (Bokmål) meget (now only as adverb)… …In present regional use mickle is the prevailing form in the north of England, and muckle in Scotland (although meikle is frequently met with as a place-name element); in Northern Ireland meckle (now rare or obsolete elsewhere) is common alongside mickle and muckle
; an isolated 20th-cent. instance of muckle in New England is recorded by Dict. Amer. Regional Eng. s.v. The archaistic use in non-regional poetry is rare, and almost confined to the form mickle. From early times sense 1a was used to designate a particular place, building, etc.; this use now survives only in place names, e.g. Michelmersh, Hants., Mickleover, Derbys. (cf. Magna Oufra (c1100)), Meikle Black Law, Berwickshire, Muckle Flugga, Shetland (in this example prob. a partial Anglicization of a Norn name), etc. Cf.:

222 Margaret Scott

eOE Bounds (Sawyer 512) in W. de G. Birch Cartularium Saxonicum (1887) II. 520 Þis is þara VI sulunga boc æt Miclangrafe. c1275 in W. A. Hulton Coucher Bk. Whalley Abbey (1848) 688, Partem cujusdam terræ..quod vocatur Mikel pughull et Litel pughull. 1379 Rolls of Parl. III. 69/1 Son College appellez Mokel Universite Halle en Oxenford. c1400 Brut (Rawl. B. 171) 50 Aurilambros & Vter .. assemblede a .. hoste forto come into michel Britaigne. 1459 Will J. Fastolf in Paston Lett. (1971) I. 90 Castre faste by Mikel Yermuth, in the shire of Norffolk. a1475 in A. Clark Eng. Reg. Godstow Nunnery (1911) I. 287, j other half acre..vppon mochel-forlonge…] A. adj. I. = great a. in various applications. Cf. much a. 1. 1. a. Great or large in size, bulk, or stature; = much a. 1a. Also designating a particular place, building, etc. (now only in place names: see etymological note) … (OED3 s.v. mickle a., pron., n., and adv.)

In the OED3 entry for malm n., the small-type etymological note discusses possible examples of the use of the element in English place-names. In the list of cognate forms given in the main part of the etymology, reference is made to the use of Old Swedish malm in the place-name Malmö, and to the Shetland place-name Maamy Soond in Fetlar. The place-name evidence allows the identification of cognate forms which would otherwise be unknown, and thus enhances our knowledge of the history and development of the word within the Germanic languages: malm, n… [Cognate with Old Icelandic málmr sand (in the place name Málm-haugar ‘sand-mounds’ (Old Swedish Malmöughe, now Malmö) in Sweden; cf. also the Shetland place name Maamy Soond ‘sandy sound’ in Fetlar), gold (because found in grains in alluvial deposits: cf. the phrase Rínar rauð-málmr ‘the red metal of the Rhine’), metal, ore (cf. Old Swedish malmber, Swedish malm, Danish malm ore), Gothic malma sand; also Middle High German malmen to crush (German zermalmen with prefix zer-). Cf. Dutch malen to grind, whirl (> maelstrom n.) and verbs in the other main Germanic languages except Old English and Old Frisian from the same Indo-European base; from IndoEuropean ablaut variants cf. meal n.1 (and with -m extension also Middle Dutch melm dust, dry sand, Old Saxon melm, Old High German melm dust), and mould n.1 (with -m extension also Middle Dutch molm, mollem (Dutch molm) mould, German Mulm dust). Apart from a single probable instance at sense 2, Old English mealm (prob. orig. with basic meaning ‘sandy, friable stone’) is attested only in the compound malmstone n., the adjective mealmiht sandy, chalky (cf. later malmy a.), and (unless independently < early Scandinavian) in the place names

Uses of Scottish place-names as evidence in historical dictionaries 223

Malmeton, Lincolnshire (a1066; now Manton), and Malmerby, North Yorkshire (1086; now Melmerby). The word is noted in field names (both as a simplex and as an element in compounds) from southern counties from the 15th cent. onwards (the earliest being Le Malme, Berkshire, 1439) in J. Field Eng. Field Names (1972) 132…] I. Simple uses. 1. a. A soft friable rock containing a high proportion of chalky material; spec. a greyish-white calcareous sandstone of the Upper Greensand of Surrey and Sussex in England (also called firestone). b. A light loamy soil formed by the decomposition of malm rock. Recorded earliest in malmstone n… (OED3 s.v. malm n.)

The etymologies of words in OED3 are often more comprehensive and detailed than those of earlier editions (cf. Durkin 1999, Durkin et al. 2002). A growing awareness of the value of toponymic material as an important supplement to lexical material has already increased the volume of scholarly research in this field over the last hundred years, and OED3 endeavours to take account of all relevant etymological scholarship in the course of the revision programme. Place-name evidence is being incorporated in cases where the toponymic material makes a significant contribution to the development of a lexical item. As the revision process continues, there will be many more such additions, and consequently the Germanic toponymicon of the British Isles may be more widely recognised for the wealth of linguistic information it contains.

References Craigie, W., et al., eds. 1937–2002. The Dictionary of the Older Scottish Tongue. Aberdeen: Aberdeen University Press; Chicago: Chicago University Press; London & Oxford: Oxford University Press. Dixon, Norman. 1947. The Place-Names of Midlothian. Ph.D. dissertation, University of Edinburgh. Durkin, Philip. 1999. “Root and branch: revising the etymological component of the OED”. Transactions of the Philological Society 97.1–50. ———, Anthony Esposito, Neil Fulton, Jane McCauley, Samantha Schad, Tania Styles & Joanna Tulloch. 2002. “From the etymology workshop: work in progress on the Oxford English Dictionary”. Oxford University Working Papers in Linguistics, Philology & Phonetics 7.225–242. Fraser, William, ed. 1874. The Lennox. Edinburgh: T. & A. Constable. Gelling, Margaret & Ann Cole. 2000. The Landscape of Place-Names. Stamford: Shaun Tyas. Grant, W., et al., eds. 1931–76. The Scottish National Dictionary. Edinburgh: Scottish National Dictionary Association.



224 Margaret Scott

Hough, Carole. 1995. “OE *græg ¯ in Place-Names”. Neuphilologische Mitteilungen 96.361–365. ———. 1998. “OE bru¯n in Place-Names”. English Studies 79.512–521. ———. 1999. “The Trumpeters of Bemersyde: a Scottish Placename Reconsidered”. Names 47.257–268. Kitson, Peter. 1997. “Old English Bird-Names (I)”. English Studies 78.481–505. ———. 1998. “Old English Bird-Names (II)”. English Studies 79.2–22. Macdonald, Angus. 1937. The Place-Names of West Lothian. Ph.D. dissertation, University of Edinburgh. ———. 1941. The Place-Names of West Lothian. Edinburgh: Oliver & Boyd. Nicolaisen, W. F. H. 1976. Scottish Place-Names. London: Batsford. New edn. 2001. Edinburgh: John Donald. ———. 1993. “Scottish place-names as evidence for language change”. Names 41.306–313. ———. 1995. “Is there a Northwest Germanic toponymy? Some thoughts and a proposal”. Nordwestgermanisch ed. by Edith Marold & Christiane Zimmermann, 103–114. Berlin: Walter de Gruyter. Oxford English Dictionary. 1884–1933. ed. by Sir James A. H. Murray, Henry Bradley, Sir William A. Craigie & Charles T. Onions; Supplement, 1972–1986, ed. by Robert W. Burchfield; 2nd. edn., 1989, ed. by John A. Simpson & Edmund S. C. Weiner; Additions Series, 1993–1997, ed. by John A. Simpson, Edmund S. C. Weiner & Michael Proffitt; 3rd edn. (in progress) OED Online, March 2000–, ed. by John A. Simpson. Oxford: Oxford University Press. Parsons, D. & T. Styles with C. Hough. 1997. The Vocabulary of English Place-Names: Á-BOX. Nottingham: Centre for English Name Studies. ———. & T. Styles. 2000. The Vocabulary of English Place-Names: BRACE-CÆSTER. Nottingham: Centre for English Name Studies. Scott, Margaret. 2000. “Privick: A Scottish place-name”. Paper presented at the Conference of the Society for Name Studies in Britain and Ireland, Bangor, April 2000. Smith, A. H. 1956. English Place-Name Elements. English Place-Name Survey vols. 25 & 26. Cambridge: Cambridge University Press. Stuart, J., et al. eds. 1878–1908. The Exchequer Rolls of Scotland. Edinburgh: H. M. General Register House. Taylor, Simon. 2002. “Walking the Marches of Language: Scots in Medieval Boundary Charters”. Paper presented at the forty-fifth Annual Conference of the Scottish Medievalists Society, Pitlochry, Scotland, January 2002. Williamson, M. G. 1942. The Non-Celtic Place-Names of the Scottish Border Counties. Ph.D. dissertation, University of Edinburgh. Wright, Joseph. 1898–1905. The English Dialect Dictionary. London: Henry Frowde.



On the stressing of French loanwords in English Ann-Marie Svensson University of Goteborg, Sweden

1.

Introduction

In the chapter on stress in Elements of Orthoepy, Robert Nares makes the following comment: Accent appears to be the most unstable part of the English language. We can all remember words differently accented from the present practice; and many might be collected which still are fluctuating, with their accent unsettled. (1784: 147)

The key words here are unstable, fluctuating and unsettled, and Nares comes back to the same theme in his chapter on “ancient and peculiar accentuation”: “In perusal of our best poets, we often meet with words accented in a manner not familiar to us, in a way which was usual in their time, but since has grown into disuse” (1784: 324). These two quotations indicate that the eighteenth century was a period of changing and fluctuating stress. Samuel Johnson, who gives stress in his dictionary from 1755, also comments on the instability of English stress: “rules for the accent or quantity are not easily to be given, being subject to innumerable exceptions” (Preface, unpaginated). The words used by Nares, Johnson and their contemporaries to illustrate stress fluctuation are all French loanwords introduced into English after the Norman Conquest when many of French loans began to invade the English language. Two conflicting stress patterns clashed, the Germanic system with initial root stress and the French one with the accent on the last heavy syllable. At first the French loans were stressed in the French way, which can be seen in

226 Ann-Marie Svensson

Middle English poetry, but as they became integrated into English, they gradually adopted a more English pronunciation, which often meant that the stress moved to the beginning of the words according to the Germanic pattern. Lass (1992:89) shows that, to begin with, the two systems could work together but that the Middle English process of vowel reduction and loss often led to variation and double forms, e.g. nacioun stressed NAcioun and naCIOUN, respectively. Also, Barber (1997:132–134) states that the existence of secondary stress in Middle English facilitated the use of such variant forms of French loans, not only in Middle English but also in Early Modern English. Minkova (1997: 144–145) points to the reluctance to stress the last syllable as the factor responsible for the early stress shift and also for fluctuating stress in Present Day English. In today’s English (Wells 2000), the great majority of the French loans are pronounced with the stress on the first syllable, but there are exceptions and words with fluctuating stress and also differences between British and American English. Some examples are shown in Table 1 below (stress is marked before the stressed syllable). Table 1.Examples of fluctuating stress in today’s English Disyllables

café chauffeur debris debut dressage garage gourmet massage paté touché

Trisyllables BrE

AmE

’--, -’’--, -’’-’-’--, -’’--, -’’-’-’-’--

-’-’-’-, ’--’-’-’’--, -’-’-’-’-

amateur attaché cabriolet cigarette colonnade consommé deficit dénoument matinée renaissance

BrE

AmE

’---,--’-’-’---,--’--’-, ’----’-, ’---’--, ’--’---, -’--’-’---’--,--’-

’----’--’’---,--’--’--’’---, -’---’--’--’-, ’---

How important for the stressing today are factors such as date of introduction into English, number of syllables or prefixes? The tables and diagrams in the following discussion will be based on the categories of loanwords in the list below. 1. 2. 3. 4.

200 disyllabic nouns introduced into English before 1500 200 trisyllabic nouns introduced into English before 1500 100 disyllabic nouns introduced into English 1500–1700 100 trisyllabic nouns introduced into English 1500–1700

On the stressing of French loanwords in English 227

5. 100 disyllabic nouns introduced into English after 1700 6. 100 trisyllabic nouns introduced into English after 1700 7. 200 prefixed disyllabic noun/verb pairs. The Modern English stressing of these words as given in Wells (2000) will be compared to that given by Johnson (1755). Citations can be found at the dictionary entries discussed. Stress differences between British and American English and some of the comments on stress and stress fluctuation made in early dictionaries and pronunciation handbooks will also be examined.

2. Nouns introduced before 1500 2.1 Disyllabic nouns As for the disyllabic nouns introduced into English before 1500, 97 per cent of the examined 200 nouns are (or can be) stressed on the first syllable in today’s English, both British and American (Wells 2000), and 99 per cent of these nouns have the same stress in Johnson (1755) as in Wells (2000). Furthermore, Levins (1570) gives 74 of these early disyllabic loans with their stress marked, and all of them have the same stress here as in Johnson (1755), which indicates that the stress shift probably started as early as the Middle English period. However, Johnson and his contemporaries sometimes comment on earlier changes or fluctuating stress as in barrier, introduced into English in the fourteenth century: “It is sometimes pronounced with the accent on the last syllable, but it is placed more properly on the first” (Johnson 1755). Johnson thus finds the last-syllable stressing of barrier old-fashioned, and so does Walker (1826), who comments on Pope’s fluctuating stressing of the word: Pope, by the licence of his art, pronounced this word in two syllables, with the accent on the last as if written bar-reér. “’Twixt that and reason what a nice barrier! “For ever sep’rate, yet for ever near.” Essay on Man, Ep. I. v. 215 And yet in another part of his works he places the accent on the first syllable as we always hear it in prose. “Safe in the love of Heav’n, an ocean flows “Around our realm, a barrier from the foes.”

Aspect is another fourteenth-century loan whose shift of stress is commented on by Johnson: “It appears anciently to have been pronounced with the accent on

228 Ann-Marie Svensson

the last syllable, which is now placed on the first”. Nares (1784: 329–330) believes the last-syllable stress “was undoubtedly the ancient accent” and quotes Shakespeare to illustrate this. Nares goes on to point to the shift: “In Dryden’s time the accent had changed its place”. 2.2 Trisyllabic nouns As for the early trisyllable loans, the majority, 84 per cent, of the examined 200 nouns, have (or can have) the stress on the first syllable today, and here, too, there is agreement between British and American English. There are, however, fewer trisyllables with first-syllable stress in Johnson (1755), 76 per cent, and eighteenth- and nineteenth-century linguists frequently comment on fluctuation between the first and last syllable and also between the first and second. Dobson (1968: 445–446) finds that Middle English secondary stress in educated speech survived longer in trisyllables than in disyllables and that this is one of the reasons for stress fluctuations in the sixteenth and seventeeth centuries. Reverie, introduced into English in the fourteenth century, is given penultimate stress by Johnson, but Walker gives first-syllable stress and comments on the earlier fluctuation between “the first and last syllable”. This word seems to have been some years floating between the first and last syllable, but to have settled at last on the former. It may still, however, be reckoned among these words, which, if occasion require, admit of either.

As for the stressing of retinue, another trisyllabic noun introduced into English in the fourteenth century, Nares (1784: 171–172) is in two minds: “I think I have sometimes heard rétinue accented regularly in colloquial use; but I know not of any poetical authority for it”. Walker also points to the fluctuation of retinue and gives both first- and second-syllable stress, but he prefers the former: This word was formerly always accented on the second syllable: but the antepenultimate accent to which our language is so prone in simples of three syllables, has so generally obtained as to make it doubtful to which side the best usage inclines. Dr. Johnson, Sheridan, Ash, Kenrick, Nares, Bailey, and Fenning, accent the second syllable; and Buchanan, W. Johnson, Perry, Barclay, and Entick, the first. Scott accents both, but prefers the first. In this case, then, analogy ought to decide for placing the accent on the first syllable.

Thus retinue appears to have been in the process of shifting the stress from the second to the first syllable, and this is also the case with revenue, introduced in

On the stressing of French loanwords in English 229

the fifteenth century. Nares comments that “[r]evenue is accented both ways by the best writers”, gives quotations from Shakespeare and Swift to illustrate fluctuation, and adds: But Swift, who piqued himself upon correctness: “Not so a nation’s révenues are paid.” (1784: 172)

Johnson finds that “its accent is uncertain”, and so does Walker: This word seems as nearly balanced between the accent on the first and second syllable as possible but as it is of the same form and origin as avenue and retinue, it ought to follow the same fortune. Retinue seems to have been long inclining to accent the first syllable, and avenue has decidedly done so, since Dr. Watts observed that it was sometimes accented on the second; and by this retrocession of accent, as it may be called, we may easily forsee that these three words will uniformly yield to the antepenultimate accent, the favourite accent of our language, conformably to the general rule, which accents simples of three syllables upon the first. Dr. Johnson, Mr. Nares and Bailey, are for the accent on the second syllable; but Dr. Ash, Dr. Kenrick, Buchanan, W. Johnson, Perry, Barclay, Fenning, and Entick, accent the first. Mr. Sheridan gives both, but places the penultimate accent first.

The phrase “since Dr. Watts observed that it was sometimes accented on the second” indicates a shift in the eighteenth century. Avenue was introduced into English in the seventeenth century and was thus a fairly recent acquisition when Watts wrote The Art of Reading and Writing English, which was published in 1721. Watts comments: The Metre also is favour’d sometimes by placing the Accent on different Syllables in some Words that will admit of it; as the Word Avenue must have the Accent on the first Syllable in this Line, Wide ávenues for cruel Death. But in the next Line it must be accented on the second Syllable; as, A wide Avénue to the Grave. (1721: 77–78)

Avenue belongs to the words introduced after 1500, and here we find more fluctuating stress in both disyllabic and trisyllabic nouns than in early loans.

230 Ann-Marie Svensson

3. Nouns introduced after 1500 3.1 Disyllabic nouns Nouns introduced after 1500, and particularly after 1700, display not only more fluctuation than early loans but also a decrease in the percentage of nouns with first-syllable stress. There are also clear differences between British and American English, as shown in Table 2 below. Table 2.Disyllabic French nouns in Present-Day English British English

American English

Introduced

’--

’-- / -’-

-’-

’--

’-- / -’-

-’-

before 1500 1500–1700 after 1700

94% 73% 57%

3% 8% 22%

3% 19% 21%

94% 67% 28%

3% 8% 10%

3% 25% 62%

In fact, the majority of the late disyllabic loans have kept the French stress on the last syllable in today’s American English, but in British English many of these loans, e.g. debut, paté, touché, have moved the stress. 3.2 Trisyllabic nouns Like the disyllables introduced after 1500, late trisyllabic loans display fluctuating stress, which is seen in 26 per cent of the examined trisyllables in British English and in 16 per cent of these nouns in American English, and there are also stress differences between British and American English. Thus words like matinée and negligée, introduced after 1700, are stressed on the first syllable in British English but on the last in American English. Another noun, champignon, introduced into English in the sixteenth century, has penultimate stress in Johnson (1755) and Walker (1826), and this stressing of the word has been kept in American English. As for champignon in British English, Wells (2000) gives first-syllable stress as first choice with last-syllable stress as an option. There is also stress variation between the penultimate in British English and the last syllable in American English, as in attaché and décolletée. A possible explanation for the stress differences between British and American English may be the greater equality between stressed and unstressed syllables in polysyllabic words in American English than in British English.

On the stressing of French loanwords in English

For obvious reasons, nouns introduced after 1700 are generally not found in Johnson (1755) or Walker (1826), but Walker has a comment on envelope, introduced into English in the eighteenth century: “This word… is always pronounced in the French manner by those who can pronounce French…” The pronunciation of those who did not know French is not given!

4. Noun/verb pairs The introduction into English of a large number of French prefixed disyllabic noun/verb pairs disrupted the original Germanic stress pattern of such doublets, i.e. the stress on the first syllable of the noun but on the root of the verb. However, the loans with Romance prefixes soon began to copy the Germanic pattern, and as early as the sixteenth century, linguists like Levins (1570) and Coote (1596) were aware of a stress distinction between the nouns and the verbs in prefixed noun/verb pairs. Lists of doublets with stress distinction can be found in some of the early pronunciation handbooks, e.g. Butler (1634) and Jonson (1640). Johnson (1755) gives a rule in the preface to his dictionary: Of dissyllables which are at once nouns and verbs, the verb has commonly the accent on the latter, and the noun on the former syllable; as, to descánt, a déscant; to cemént, a cément; to contráct, a cóntract. This rule has many exceptions. Though verbs seldom have their accent on the former, yet nouns often have it on the latter syllable; as, delíght, perfúme.

Johnson thus points to the exceptions to the rule, and in fact only 17 per cent of the 200 doublets in this study are stressed in the ‘ideal’ way in his dictionary. Walker calls the stress distinction between nouns and verbs “a peculiarity of accentuation”: There is a peculiarity of accentuation in certain words of two syllables, which are both nouns and verbs, that is not unworthy of notice; the nouns having the accent on the first syllable, and the verbs on the last. (1826: rule 492)

Walker frequently refers to this rule in his dictionary, but, like Johnson, he stresses the great majority of both the nouns and the verbs in these pairs on the last syllable. For instance, he gives the noun protest with both second- and firstsyllable stress and comments: The first pronunciation of this word is adopted by Mr. Sheridan, Mr. Scott, Dr. Kenrick, Mr. Smith, Mr. Perry, Buchanan, Barclay, Bailey, and Fenning; and

231

232 Ann-Marie Svensson

the second by Mr. Nares, Dr. Ash, Dr. Johnson, and Entick. As this substantive was derived from the verb, it had formerly the accent of the verb; and that this accent was the most prevailing, appears from the majority of authorities in its favour. But the respectable authorities for the second pronunciation, and the pretence of distinguishing it from the verb, may very probably establish it, in the detriment of the sound of the language, without any advantage to its signification. (Walker 1826)

Nares, who also points to the stress distinction between nouns and verbs, has a somewhat amusing theory as regards the shifting of stress in both the noun and the verb exile
: I take exíle to have been the more ancient accentuation. It seems to have been upon the change in the days of Elizabeth. I believe that the force of analogy changed the substantive first, and that ignorance compelled the verb to follow it. To exíle was not quite obsolete when Dryden wrote: he has used that verb both ways. (1784: 345–346)

However, exile is not the only prefixed verb that has moved the stress to the prefix. Table 3 below shows that not only nouns but also a number of verbs, e.g. convoy, discount, prefix, surcharge, moved the stress to the first syllable between 1755 (Johnson) and 2000 (Wells). Table 3.Noun/verb pairs 80% 70% 60% 50%

n ’--, v -’n -’-, v -’n ’--, v ’--

40% 30% 20% 10% 0% 1755

2000

On the stressing of French loanwords in English 233

5. Conclusion As Robert Nares expressed it in 1784, “[w]e can all remember words differently accented from the present practice” (147). There is still the instability of stress Nares talks about, but the words “with their accent unsettled” are not the same as those commented on by Nares and his contemporaries. Especially in the categories of late loans (nouns introduced after 1700) and noun/verb pairs there is still fluctuation. Nouns introduced during the last couple of centuries may still be regarded as ‘foreign’, perhaps especially in American English. The date of introduction thus appears to be a factor of importance, and loanwords may need time to become fully integrated as regards pronunciation. The prefixes in the noun/verb pairs may be the reason why these disyllabic nouns kept the stress on the root longer than other disyllabic loans. However, the fact that not only prefixed nouns, which would be expected to have firstsyllable stress but also prefixed verbs, are in the process of moving the stress, indicates that some of the Romance prefixes are no longer felt to be prefixes in the same way as the original Germanic prefixes. The fact that trisyllabic nouns like avenue and revenue sometimes displayed fluctuation between the penultimate and antepenultimate syllable in the eighteenth century and other nouns, e.g. consommé and deficit, display this variation today, supports the results shown by Minkova (1997) that the reluctance to stress the last syllable is a strongly influential factor in past and present stress fluctuation. It appears that we are still in the process of moving the stress of Romance loans so that they fit into the Germanic pattern.

References Barber, Charles. 1997. Early Modern English. 2nd edn. Edinburgh: Edinburgh University Press. Butler, Charles. 1634. The English Grammar, or The Institution of Letters, Syllables, and Woords in the English tung. Neudrucke frühneuenglischer Grammatiken, Band 4,1 (1910) ed. by A. Eichler. Halle: Max Niemeyer. Coote, Edmund. 1596. The English Schoole-maister. A Collection of Facsimile Reprints, No 98 (1968) ed. by R. C. Alston. Menston, Yorkshire: Scholar Press. Dobson, E. J. 1968. English Pronunciation 1500–1700. 2nd edn. Oxford: Oxford University Press.



234 Ann-Marie Svensson

Johnson, Samuel. 1755. A Dictionary of the English Language. Facsimile reprint 1968. Hildesheim: Georg Olms. Jonson, Ben. 1640. The English Grammar. A Collection of Facsimile Reprints, No 349 (1972) ed. by R. C. Alston. Menston, Yorkshire: Scholar Press. Lass, Roger. 1992. “Phonology and morphology”. The Cambridge History of the English Language. Volume II: 1066–1476 ed. by Norman Blake, 23–153. Cambridge: Cambridge University Press. Levins, Peter. 1570. Manipulus Vocabulorum. A Collection of Facsimile Reprints, No 169 (1969) ed. by R. C. Alston. Menston, Yorkshire: Scholar Press. Minkova, Donka. 1997. “Constraint ranking in Middle English stress-shifting”. English Language and Linguistics 1.135–175. Nares, Robert. 1784. Elements of Orthoepy. A Collection of Facsimile Reprints, No 56 (1968) ed. by R. C. Alston. Menston, Yorkshire: Scholar Press. Svensson, Ann-Marie. 1997. “Stress Shift in English Disyllabic Prefixed Noun/Verb Pairs”. Moderna Språk 2.130–140. ———. 2001. “Germanic vs. French stress in disyllabic loanwords in English”. A wealth of English. Studies in honour of Göran Kjellmer ed. by Karin Aijmer, 123–129. Göteborg: ACTA (= Gothenburg Studies in English 81). Walker, John. 1826. A Critical Pronouncing Dictionary, and Expositor of the English Language. 4th edn. London and Leipzig: Ernest Fleischer. Watts, Isaac. 1721. The Art of Reading and Writing English. A Collection of Facsimile Reprints, No 333 (1972) ed. by R. C. Alston. Menston, Yorkshire: Scholar Press. Wells, J. C. 2000. Longman Pronunciation Dictionary. London: Longman.



Like like love Comparing two modern English words diachronically Heli Tissari University of Helsinki

1.

Introduction

The meaning of the verbs to like and to love can overlap: (1) I like cats better than dogs (BNC: Conversation [KC0 6308]) (2) I love cats and hate dogs (BNC: Graham, Reason and spontaneity [CB1 31])

However, to like in its sense ‘to find agreeable/to have a taste for/to take pleasure in’ (s.v. OED like v1: 6) denotes less intense feeling than to love ‘to entertain a great affection or regard for’ (s.v. OED love v1: 1). Note the contrast between “better than dogs” and “hate dogs”. Because liking is considered less intense than loving, using love instead of like has been criticised. The OED remarks that love is used as a “frequent vulgarism for like” in the US. Lewis (1968 [1960]: 15) writes: “Most of my generation were reproved as children for saying that we ‘loved’ strawberries…” This study investigates whether to like in the above sense was and is used differently from to love in Late Middle (fifteenth-century), Early Modern (sixteenth and seventeenth centuries) and Present-day English, asking also why and how. Although the main question concerns the verbs like and love, relevant nouns of attachment, above all love and liking, are also included in the data, because concepts “may be expressed by two or more words” (Miller & JohnsonLaird 1976: 269). The data comes from seven corpora, the size of which varies between half a million and a million words. For this article, I collected data on like, because I have dealt extensively with love in my earlier work (Tissari 2003). While the

236 Heli Tissari

main focus of the analysis is on like, results for like are compared with compatible results for love, based on the same sources of data. 2. Corpora These sources consist of the Corpus of English Correspondence Sampler (CEECS), the Early Modern English period of the Helsinki Corpus (HC), Shakespeare’s plays in the OUP electronic edition of his Complete Works (SH), the Brown Corpus of American English, the Lancaster-Oslo/Bergen Corpus of British English (LOB), the Freiburg-Brown Corpus of American English (FROWN), and the Freiburg-LOB Corpus of British English (FLOB). This combination of corpora is more balanced than a comparison of the three first, Late Middle and Early Modern English corpora, with the British National Corpus (BNC), for example. CEECS, HC and SH together have about 2 million words, and the Present-day English data total 4 million words. This is little compared to using the 100-million-word BNC, but the data nevertheless covers both British and American English, and includes both 1960s and 1990s English. It is more reasonable to make comparisons between two pools of data, one of which is twice the size of the other, than two pools of data, one of which is 500 times the size of the other.1 Table 1.Love and like in the data, including all forms of the verb and noun love, the verb like, and the verbal nouns like and liking, the latter in the general sense ‘to please’ or ‘to take pleasure in’ Corpus

CEECS 1 CEECS 2 HC SH BROWN LOB FROWN FLOB

Dating

1418–1638 1580–1680 1500–1710 1610 ± 20 1961 1961 1991 1991

like

love

Absolute frequency

N/10,000

Absolute frequency

N/10,000

114 37 90 236 260 380 332 323

4.63 1.81 1.63 1.99 2.60 3.80 3.32 3.23

,792

18*

,498 2,684 ,418 ,526 ,528 ,478

9.0 33 4.1 5.2 5.3 4.8

*Figures for CEECS 1 and CEECS 2

1.For further information on the corpora and help in locating the examples see the manuals: Burnard 1995 for BNC, and the ICAME home page ·http://khnt.hit.uib.no/icame/manuals/Ò 110603 for the rest. As to lines from Shakespeare, their numbers are based on Spevack 1970.

Like like love 237

Table 1 suggests that a significant change has occurred in the relative frequencies of love and like. In the Early Modern data, love is much more frequent than like, while in the Present-day English data, love has become much rarer and like more frequent. In terms of frequency, the two words begin to resemble each other. Note, however, that there is great variation between the structure and content of the Early Modern English corpora, while the Presentday English corpora are more directly comparable with each other.

3. Love My research on love, verb and noun, relates to an issue which seems central in comparing love and like, i.e. the participants of love. These include the subject and object of the verb love, or more generally, the experiencer and cause(r) of the emotion (Diller 1994: 221–222). Lewis’s (1968: 15) remark on strawberries and the title of the chapter in which it appears, “Likings and loves for the sub-human”, suggests that the verb like should be preferred when the cause(r) is not human. I have included animals in this category, although arguably animals are often loved as if they were human (Lewis 1968: 32). To study the distinction between love and like, it is easier to see whether liking rather than loving is directed at non-human targets, than to create some other scale for the intensity of the emotion. However, my research on love not only distinguishes between human and non-human cause(r)s of emotion, but also divides the former into sub-categories. Following the Greek terms used by Lewis (1968), I suggest that there are at least four forms of love apart from loving non-humans, and that these occur in the domains of family, sexuality, friendship, and religion: 1. Family love, or storge, occurs in the domain of family. Its participants are members of the same family. 2. Sexual love, or eros, occurs in the domain of sexuality. In the ideal case, its participants are each other’s lovers. 2b. It is possible to distinguish more specifically between storge, eros and what I call storge-eros, i.e. marital love, in which the participants are spouses to each other. The domain of this love is marriage. However, this distinction is not relevant for the present study. 3. Friendly love, or philia, occurs in the domain of friendship. More specifically, it occurs between people who are not lovers or members of the same family. It thus covers philanthrophy as well as more ‘genuine’ friendship.

238 Heli Tissari

4. Religious love, or agape, occurs in the domain of religion. Its participants include God or people acting according to their faith. 5. Love of non-human participants, or khreia, occurs in the domain of the ‘rest of the world’. At least one of its participants is non-human. The following examples illustrate these forms of love as senses of the verb or noun love
: (3) Storge
: Tell me, my daughters … Which of you shall we say doth love us most … ? (SH: King Lear 1.01.51) (4) Eros
: Fy fy for shame Loue ye, and know not whome; but hir yonde, a Woman, We shall then get you a Wyfe, I can not tell whan. (HC: Udall, Roister Doister [UDALL] L.131) (5) Storge-eros
: I like to think he lived on, married, had children and a wife who loved him dearly. (FLOB: N N26: 69) (6) Philia
: I write both for the ardent love which I bere unto your Grace, and for the accomplishment off my dutie (HC: Tunstall, Letters [TUNSTALL] I, 136) (7) Agape
: The church is “the family of God”. The members of the “family” are drawn together by a common love for Christ and a sincere devotion to His Kingdom. (BROWN: Religion D09: 95) (8) Khreia
: And it was the House he loved. (BROWN: Press: Editorial B03: 22)

In my research on love I emphasise that the senses storge, storge-eros, eros, philia, agape and khreia ought to be seen as prototypical categories which can partly overlap. Note that example (7), labelled agape, shares characteristics with storge. Eros seems to be by far the most frequent sense of love in both Early Modern (49%) and Present-day English data (64%), and its proportion of the total meaning of love seems to have increased during the past five hundred years. Another, but less significant, increase can be seen in the proportion of khreia (14% to 19%). The role of khreia in the history of ‘love words’ becomes more significant when it is noted that its proportion of the meanings of the words affection and passion also seems to have grown. The proportion of agape in the meaning of love has hardly changed (7% to 6%), but storge (14% to 5%) and philia (16% to 6%) attest a decrease (Tissari 2003: 265–267, 295–300, 305–313). In terms of these senses, the meaning of love does not seem to be stable but to change between Early Modern and Present-day English. The meaning of like is also likely to have changed between these two periods, but the change need not concern the participants of liking.

Like like love 239

4. Like Impersonal uses of to like provide a well-known example of change in the history of English:2 Like, OE lician, originally meant ‘please’ and took its object in the dative. We may give a typical sentence in its various stages in the following way: (1) Þam cynge licodon peran. (2) the king likeden peares. (3) the king liked pears. (4) he liked pears. (Jespersen 1961 [1927]: 208–209)

Jespersen’s example shows how an Old English dative construction changes into a Present-Day English sentence with a reverse word order. He claimed that in sentence (1) the king is the object, while in sentence (3), the king is the subject. He suggested that the example illustrates a syntactic reanalysis of the verb like, based on changes in English word order. Although writers on the verb like often quote Jespersen (Allen 1986: 376, Denison 1993: 74–75, Mair 1988: 210), Allen (1986) strongly criticises Jespersen’s example, and says that one cannot accept it as such. Changes in the behaviour of to like were more subtle, and resulted from a gradual development which had several causes. Allen (1986: 401) dates the crucial issue, the introduction of structural case marking, as “beginning in the early thirteenth century, and becoming complete in the sixteenth”. Consequently, We can say that speakers born in the later part of the fifteenth-century had grammars with a few verbs whose lexical entries specified (optional) dative case of the subject. But by the turn of the century, language learners constructed grammars in which the most frequently occurring ‘impersonal’ expressions of the preceding generation had become fixed expressions specifying the form, rather than the case, of the subject. (Allen 1986: 403)

We can thus predict that the earliest Late Middle English data attest free variation in the grammatical role of the experiencer and cause(r) of liking, while the Early Modern English data prefers fixed impersonal expressions. By Presentday English, impersonal uses are likely to have vanished. The last impersonal

2.The term ‘impersonal’ “indicates that the human or animate experiencer of a mental state is not coded as the grammatical subject of a transitive SVO clause but as an object” (Mair 1988: 210). I thank Dr. Minna Palander-Collin for suggesting references for this section.

240 Heli Tissari

example of to like in the OED is (bold emphasis added): (9) a1850 ROSSETTI Dante & Circ. (1874) I. 41, I rode sullenly Upon a certain path that liked me not. (s.v. like v1: 1)

With respect to CEECS, it is of interest to note Mair’s (1988: 211–212) observation concerning Late Middle and Early Modern English correspondence: “one of the most commonly used text-type specific expressions of deference involves the verb like (as in “like it your highness to understand/know/etc.”), so that this particular usage is certainly over-represented in a corpus restricted to letters”. Such over-representation may facilitate the timing of the fixing of impersonal expressions.

5. Results on ‘like’ This section combines three points of view on like. Liking is first analysed like loving, as occurring in the domain of family, friendship, sexuality, religion, and either between or towards non-human participants. Impersonal uses are then looked at. Lastly, I point out that the newer would like (to) construction has taken over functions of impersonal expressions. 5.1 Cause(r)s of pleasure I focused on the cause(r) of pleasure — at times displeasure — because no ‘five forms of liking’ are as conceptually salient as the five ‘loves’. Rather, it seems natural to ask whom or what it is possible to like. The following are examples in which the cause(r) of pleasure is (10) from one’s own family: I did not like my father. (FLOB: G G15: 27) (11) sexually attractive: Why, I’ faith, methinks she’s too low for a high praise, too brown for a fair praise, and too little for a great praise. Only this commendation I can afford her, that were she other than she is she were unhandsome, and being no other but as she is, I do not like her. (SH: Much Ado About Nothing 1.01.176) (12) religiously motivated (?): [Speaker A:] And looke in the 12. chapter of the Reuelation, and you shall find that the Dragon, which the Popery doth worship in stead of God, is the Deuill. [Speaker B:] Trulie I like your wordes well, I am persuaded the deuill doth seduce and bewitch

Like like love 241

mens mindes … (HC: Gifford, A Handbook on Witches and Witchcraft [GIFFORD] B4R) (13) none of the above, but human: They just wanted to beat me up. They didn’t like black people. Who knows? Maybe each one of them was mugged by a black man, too. (FROWN: B B17: 11) (14) non-human: I rather like pacing around. That’s what I always do at home. (FLOB: N N04: 75)

The data clearly indicate that the cause(r) is highly unlikely to be religiously motivated — all the cases can be disputed — or a member of one’s own family, although the latter seem slightly more salient. Perhaps like is unfit for religious discourse, because the existence of God arouses stronger emotions. Both claiming and denying liking for a member of one’s family seem to allow room for love: if family love is instinctive, one might love despite a dislike of the personal traits of the beloved. But on the other hand, if family love is expected of everybody, it is easier to claim a dislike for one’s family than to bluntly admit a lack of love for them (example 10). The distinction between loving and liking is psychologically and socially intriguing. A member of the opposite sex fairly often causes liking in Shakespeare’s plays, in which like helps to belittle the romantic emotion and turns it into laughter. It is used for downplaying or highlighting embarrassment, which may arise either from social clumsiness or immoral behaviour. Dislike can serve as a disguise for loving, being a means of criticism but not necessarily a negation of attraction (example 11). Like seems to be frequently used for criticising people and their actions. At times it is associated with direct aggression. Dislike of somebody can ‘cause’ violence (as in example 13 above), and violent behaviour can incite pleasure (example 16 below). This ‘aggressive’ use is characteristic of slang: (15) “My boss ain’t gonna like it if you muck up the wipers.” (FLOB: P P26: 36) (16) The seventh man was Red Hogan, a wiry little puncher with a wild streak and a liking for hell-raising. (BROWN: Adventure N10: 102)

Dislike meant prejudice and misgivings even in the Early Modern period (cf. examples 12 and 13): (17) (Dan.) You neuer had no hurt done yet, had you by any witch? (Sam.) Trust me I cannot tell, but I feare me I haue, for there be two or three in our towne which I like not … (HC: Gifford, A Handbook on Witches and Witchcraft [GIFFORD] B1R)

242 Heli Tissari

The OED reminds us that the verb like has had the “qualified uses” to like well or ill, and that these survive in the “neutral”, “interrogative use with how” (s.v. like v1: 6f). In its actual context, the “neutral” use often anticipates agreement: (18) (Lafeu) His highness hath promised me to do it; and to stop up the displeasure he hath conceived against your son, there is no fitter matter. How does your ladyship like it? (Countess) With very much content, my lord, and I wish it happily effected. (SH: All Is Well That Ends Well 4.05.77)

Example (18) is a typical representative of the data in general, because usually the cause(r) of pleasure is non-human, as shown by Table 2. Table 2.Cause(r)s of pleasure. The table shows their distribution in the data on like Corpus

Cause(r) of pleasure Family member

Sexually attractive

Religiously Human being, Non-human Unclear motivated further motivation

CEECS 1 CEECS 2 HC SH

1 0 0 2

1 0 7 43

0 1 2 0

7 6 15 30

104 30 38 173

2 0 3 4

Subtotal 100%

0.4%

11%

0.6%

12%

74%

1.9%

Brown LOB Frown FLOB

1 0 5 2

5 13 15 10

0 0 0 0

23 23 36 28

228 344 276 282

3 0 0 1

Subtotal 100%

0.6%

3.3%

0%

8.5%

87%

0.3%

Total 100% 0.6%

5.3%

0.2%

9.5%

84%

0.7%

5.2 Impersonal uses Since the non-human cause of pleasure is often expressed by the pronoun it, it becomes pertinent to ask whether impersonal uses affect Table 2. Most impersonal expressions include it, although sometimes it is dropped:

Like like love 243

(19) Like it yow to yeve credens to the berer … (CEECS: 1483? Page, Richard [RPAGE] II, 157) (20) Also good madame, and it lyke you to wete, as on Friday next after Corpus Christi day I intend unto the marte wardes … (CEECS: 1478 Betson, Thomas [TBETSON] II, 47) (21) Sir, like you to remember, the conclusions of the matter taken betwixt you and Chapman of Stamford by Husee and me, that yee for to have his releas general should pay 100 (s.) …(CEECS: 1464 Rocliffe, Brian [BROCLIFFE] 12) (22) If yt like your mastership, yt were best for to shew the Kyngs grace that ye wold refuse none of his counsell … (CEECS: 1503 Emerson, George [GEMERSON] 177) (23) (Clown) Are you a courtier, an ‘t like you, sir? (Autolycus) Whether it like me or no, I am a courtier. (SH: The Winter’s Tale 4.04.729)

Most of the relatively fixed impersonal expressions appearing in the Late Middle and Early Modern English data can be grouped into two kinds of formulae.3 One kind precedes news and announcements; these are usually introduced by like it you (to know) (example 20). The other precedes requests, reminders and suggestions; these often begin with like it you to remember or if it like you (examples 21 and 22). The expression like it you to give credence to the bearer in (19) combines the announcement of news with the request to listen seriously to the letter-bearer’s message. Most of the polite, fixed impersonal expressions appear in correspondence, but their reappearance in Shakespeare suggests that they also belonged to oral discourse, especially between unequals. However, Shakespeare’s favourite phrase an ’t like you seems to indicate not only reverent politeness but also jocular or rustic speech (example 23), which might be a sign that it was becoming outdated. It also attests a compression of meaning which might partly parallel and relate to the history of please. The data suggest that, despite their frequency, impersonal uses of like were fairly fixed even in Late Middle English. The number of personal uses grows towards the seventeenth century, as shown by Table 3.

3.Both formulae fall under Richardson’s term ‘notification’, which Nevalainen (2001: 205) uses in her description of Early English correspondence.

244 Heli Tissari

Table 3.The decline of impersonal like CORPUS

Impersonal uses

Personal uses

Unclear

Total

CEECS 1 (1418–1638) CEECS 2 (1580–1680) HC (1500–1710) SH

56 = 58% 1 = 3% 25 = 31% 31 = 16%

39 = 40% 32 = 97% 55 = 69% 157 = 82%

2

97 33 80 191

3

The only part of the data in which impersonal uses are more frequent than personal is the sub-corpus CEECS 1, spanning from 1418 to 1638 as against the time span 1580–1680 of CEECS 2. The most dramatic contrast is also between these two sub-corpora, CEECS 2 attesting only one impersonal use: (24) I will crave but a little leave, and tell your lordship briefly how it hath fared with us since my Lord’s last coming to Durham; who, within a few dayes after, was pleased to hold his Visitacion, (which he had begun the last yeere among us) and thereat, openly before us all, to declare his great displeasure conceyved against Dr. Lindsell and my selfe, for the speech which we had with your Lordship at Fulham, and which it liked his Lordship to call (accusations and articles preferred against him). (CEECS2: 1631 Cosin, John [JCOSIN] I, 204)

Impersonal like no longer marks politeness in extract (24), but rather the contrary. If there were many examples like (24), they might amount to evidence against Allen’s (1986: 403) claim that speakers born after 1500 did not have impersonal like in their active grammar. But the overall impression is that her dating should not be postponed. If anything, impersonal like in non-fixed expressions seems to disappear a little earlier than she suggests. Whether Thomas More, born in 1478, has this option, remains unclear on the basis of the example: (25) I besech our Lord, …, it may like hym to give your father the grace, daylie to remember and praye, …even so dayly deuoutly to knele and praye it. (HC: More, Thomas, Letters [MORELET] 545)

The expression (in italics) does not seem fixed, considering that the experiencer is in third rather than second person, and that it occurs only once in the data.

Like like love 245

However, with its dummy it, it is analogous to the formulae in Late Middle English correspondence.4 Intransitivity is not required for conveying politeness. An example from the fifteenth century suggests that even personal like could mark a polite request: (26) I beseche you to have consideracion to alle the premysses and Þat ye lyke to take surete for Þe pore man Þat he may goo at large. (CEECS1: 1470S Forster, Humphrey [HFORSTER] II, 36)

5.3 Would like (to) Like is likewise associated with politeness in Present-day English. Table 4 shows that about a third of the instances of the verb-form like in the Present-day English data are preceded by would. It seems that the expression would like (to) has replaced impersonal politeness, which is probably the reason for the new increase in the frequency of like after a decrease in the Early Modern period (cf. Table 1). Table 4.Verb-form like preceded by would CORPUS (1,000,000 words each)

Like (word-form in relevant sense)

Preceded by would (or ’d)

BROWN (1961) LOB (1961) FROWN (1991) FLOB (1991)

173 293 198 216

61 [= 35%] 92 [= 31%] 66 [= 33%] 77 [= 36%]

While impersonal expressions with like tend to be directed at a superior or honourable “you”, the use of would like (to) is by no means restricted to polite questions such as: (27) “Would you like to sit down for a few minutes?” Lynne asked kindly. (FLOB: P P15: 36)

Rather, it is used in any person, singular or plural: (28) I would now like to turn to its function in the theory of information. (LOB: J: Learned, scientific J18: 11)

4.Raumolin-Brunberg (1991: 161–163) points out that More’s use of dummy subjects and objects was fairly varied. She gives three examples of impersonal verbs, one with like.

246 Heli Tissari

(29) Donald Kruger would like nothing better than to hold him as hostage, and I wouldn’t entrust a snake to his tender care. (BROWN:Adventure N07:147) (30) Now the Americans would like to imitate inimitable Ascot. (LOB: A: Press: reportage A10: 109)

Should sometimes replaces would
: (31) In concluding this section, we should like to give an example which illustrates some of the ideas of the primary decomposition theorem. (BROWN: Learned J18: 81)

Would/should like (to) has several functions related to those of impersonal expressions. Apart from marking polite inquiry (would you like to?), it marks polite suggestions (I/we would like to), and volition. In the scientific register, it precedes the introduction of new facts. Indeed, the data suggest that a more detailed study of like should consider medium and register.5 Note that would/should like (to) indeed emerges as a new phenomenon in the Present-day English data, being non-existent in the Early Modern English data. Although like to + verb does appear early on (cf. example 26), Shakespeare, for example, tends to use the verb like in simple phrases with a direct object, as in: (32) I like your work, and you shall find I like it. (Timon of Athens 1.01.160)

The earliest examples of would/should like (to) in the OED are from the nineteenth century. This development agrees with the general expansion of English modals since the Early Modern period, and the replacement of the subjunctive.6

6. Comparing like with love The OED entry for the verb love does not say anything about the analogous would love (to), but the present data suggest that it does not appear in Early Modern English, although would love with a direct object does. Even in the Present-day English data, would love (to) appears only a few times. A quick search in the BNC gives 5,705 matches for would like, which is 16 times the 344 matches for would love. The difference may be smaller in spoken English, but it is very likely to be significant.

5.Biber et al. (2000: 486, 489) point out that the modal would is “extremely common” and occurs most frequently in conversation. 6.My thanks to an anonymous reviewer for pointing this out.

Like like love 247

Love also has uses which like does not have. In the Early Modern period, it was important to claim one’s love towards patriarchal authorities, as in: (33) …this shall learne me heerafter not to be so foulish as to sende anie sutche message as shall beare anie comment but my owen, which shall be as full of love and respect as a dewtifull sone can owe to so loving a Father, thinking I shall prosper no longer then I deserve to be cald Your Mties most humble and obedient sone and servant Charles. (CEECS2: 1613 Charles I [CHARLES1] 102)

Each age has its own social norms which regulate the expression of love and contribute to the concept. These norms may explain why love seems to be much less frequent in Present-day than Early Modern English. Love has become less honour-bound and more private. Like seems to be more neutral in this respect. Its element of pleasure is less controversial, because it is not associated primarily with human relationships. Like is a good word for inquiries about the respondent’s feelings of satisfaction. Syntactically, like seems to have changed more than love in the past six hundred years, but the same hardly applies semantically. The proportions of storge, philia, eros, agape and khreia of the meaning of love seem to have changed more than the objects taken by like.

7. Conclusion This paper has compared uses of love and like since the first half of the fifteenth century. On the basis of English-language corpora, it suggested that the intensity of pleasure does not suffice to explain their different histories. To see more clearly why they diverge, one needs to take into account the syntax and semantics of these words, and even their socio-historical context. While the sub-sense ‘pleasure’ in like leads to its frequent use in polite expressions and influences its syntactic behaviour, love is tied to social norms in a different manner. It is important to know how to present a request or how to express one’s pleasure, but beliefs about the nature of human emotions and relationships regulate the very freedom of any such communication.7

7.The research reported here was supported in part by the Academy of Finland Centre of Excellence funding for the Research Unit for Variation and Change in English at the University of Helsinki.

248 Heli Tissari

Abbreviations BNC = the British National Corpus BROWN = the Brown Corpus of American English CEECS = the Corpus of Early English Correspondence Sampler (Parts 1 & 2) FLOB = the Freiburg-LOB Corpus of British English FROWN = the Freiburg-Brown Corpus of American English HC = the Early Modern English period of the Helsinki Corpus LOB = the Lancaster-Oslo/Bergen Corpus of British English OED = Oxford English Dictionary SH = Shakespeare’s plays in the OUP electronic edition of his Complete Works

References Allen, Cynthia L. 1986. “Reconsidering the history of like”. Journal of Linguistics 22.375–409. Biber, Douglas & Stig Johansson, Geoffrey Leech, Susan Conrad, Edward Finegan. 2000. Longman Grammar of Spoken and Written English. London: Longman. Burnard, Lou, ed. 1995. Users Reference Guide for the British National Corpus. Oxford: (Published for the British National Corpus Consortium by) Oxford University Computing Services. Denison, David. 1993. English Historical Syntax: Verbal Constructions. London: Longman. Diller, Hans-Jürgen. 1994. “Emotions in the English lexicon: A historical study of a lexical field”. English Historical Linguistics 1992: Papers from the 7th international conference on English historical linguistics, Valencia, 22–26 September 1992 ed. by Francisco Fernández, Miguel Fuster & Juan José Calvo, 219–233. Amsterdam & Philadelphia: John Benjamins. Jespersen, Otto. 1961 (1927). A modern English grammar on historical principles. Part III, Vol. 2. London: George Allen & Unwin. Lewis, C. S. 1968 (1960). The Four Loves. London & Glasgow: Collins Fontana Books. Mair, Christian. 1988. “The transition from the impersonal to the personal use of the verb like in late Middle English and Early Modern English — some previously neglected determinants of variation”. Historical English: On the occasion of Karl Brunner’s 100th birthday ed. by Manfred Markus, 210–218. Innsbruck: University of Innsbruck. Miller, George A. & Philip N. Johnson-Laird. 1976. Language and Perception. Cambridge, London & Melbourne: Cambridge University Press. Nevalainen, Terttu. 2001. “Continental conventions in early English correspondence”. Towards a History of English as a History of Genres ed. by Hans-Jürgen Diller & Manfred Görlach, 203–224. Heidelberg: C. Winter. OED = Oxford English Dictionary, 1989. 2nd edn. Oxford: Oxford University Press. Raumolin-Brunberg, Helena. 1991. The Noun Phrase in Early Sixteenth-Century English: A Study Based on Sir Thomas More’s Writings. (= Mémoires de la Société Néophilologique de Helsinki, L.) Helsinki: Société Néophilologique.



Like like love 249

Spevack, Marvin. 1969 & 1970. A Complete and Systematic Concordance to the Works of Shakespeare. Vols. 4 & 5. Hildesheim: Georg Olms. Tissari, Heli. 2003. LOVEscapes: Changes in Prototypical Senses and Cognitive Metaphors since 1500. (= Mémoires de la Société Néophilologique de Helsinki, LXII.) Helsinki: Société Néophilologique.



Spirantisation and despirantisation Jerzy Welna University of Warsaw

1.

Spirantisation and despirantisation

The change of a plosive into a fricative, called spirantisation, is a common process in Indo-European languages. Frequently it affects intervocalic plosives, as exemplified by the change of Latin [d] to Gallo-Roman [ð] (audire > audhire “hear”), where the fricative [ð] is represented in spelling by the digraph ·dhÒ. In the evolution of Latin very often it was the voiceless dental stop in Gallo-Roman which became a fricative when followed by the liquids [r l], as in L patrem > peðre “father”, or in word-final position, as in fidem > feit > fei[θ] “faith” (Pope 1934: 136–137). In an analogous Spanish change, [d], both original and that from voiceless [t], undergoing voicing, yielded a fricative when before the liquid [r] (cf. quadrum > cua[ð]ro “quarter”, petram > piedra > pie[ð]ra “stone”; Man´czak 1980:31–32). The same process was at work in Early Germanic, where, under a context-free rule, the voiceless plosives [p t k] became fricatives [f θ x] (the First Germanic Consonantal Shift). An opposite development, despirantisation, by which fricatives became plosives, is equally frequent in Indo-European languages. As regards dental consonants the change was common in those Germanic languages which, unlike English, have failed to preserve the dental fricatives [θ] and [ð]. Despirantisation is also documented at the earliest stage of English, where the replacement of fricatives by plosives can be exemplified by a change like PGmc *na¯¯eþlo¯- to OE na¯¯edl “needle”. Spirantisation in Germanic affected not only dentals, but also labials and velars (Blumbach 1974). Because the relevant developments influencing the last two types of consonants deserve separate treatment, the present brief study is confined to the discussion of the voiced dental plosive [d], and refers to the

252 Jerzy Welna

fates of its voiceless counterpart [t] only marginally. An effort is here made, first, to establish as precisely as possible the time and place where spirantisation of [d] to [ð] was initiated and, second, to examine the circumstances of the reverse process, i.e. despirantisation. In general, these developments are considered to have taken place in Middle and Early Modern English. Although the relevant data, especially those from medieval English, are scant, sporadic, and inconsistent, nonetheless an attempt will be made here to work out a tentative classification of items affected by both processes.

2. Previous studies Neither spirantisation nor despirantisation in Middle English belong among the favourite topics of linguistic research. As in the case of other minor phonological developments, more adequate accounts of the change are not to be sought in recent papers and studies on the subject. Since the latter are in short supply the reader must refer to standard historical grammars. Because a substantial number of words affected by both processes contain the liquid [r] after the plosive, Luick’s classic historical grammar (1940: 1009–1012) treats the change as part of r-influences (cf. OE Šeforþian > afford and OE fæder > father, both typical examples of despirantisation and spirantisation respectively). The more or less precise dates of the processes offered in standard grammars can be summarised as shown in Table 1. In order to establish the approximate dates of both despirantisation and spirantisation it is sufficient to examine evidence from the OED or other corpora, like the Middle English Dictionary (MED). This, as will be shown

Table 1.(De)spirantisation

Jespersen Wright-Wright Jordan Luick Brunner Dobson Mincoff

(1909: 209–210) (1928: 121, 1924: 112) (1925: 186, 251) (1940: 1009–1012) (1960: 378) (1968: 954–956) (1972: 246, 252)

Despirantisation

Spirantisation

ENE (?) ME 12c–15c ONhb/c1300– ME LOE/ME ME

ENE (?) LME/ENE c1400 1400–1600 15c c1400 15c

Spirantisation and despirantisation 253

below, should allow one to confirm that the change was initiated in different words at different time periods, which makes the development an ideal research area for the followers of the theory of lexical diffusion. As regards the mechanism of these processes, Jordan and Luick explain despirantisation as a typical contextual modification triggered by the sonorants [l r m n] which immediately follow the fricative. In the reverse change, i.e. spirantisation, [d] acquires features of a fricative especially when followed directly by the syllabic liquid [r]. Both hypotheses presuppose that the development chiefly occurred in inflected forms where, following the syncope of an unstressed vowel, [d] came to stand directly before the liquid, as in inflected modre (OE mo¯dor > mother) or roðres (OE roðor > rudyr “rudder”). That both developments were not consistently carried through is evidenced by the studies of sixteenth century grammarians whose statements make it clear that, after being affected by despirantisation and spirantisation, words continued to employ spellings indicating now a plosive, now a fricative articulation (Dobson 1968: 955–956). This state of things may have been caused by the interference of non-phonological, possibly semantic, factors. According to one such hypothesis, and a hardly tenable one, words with the suffix -er denoting family relationships behaved differently from others (Fetting 1972). In the present study a consistent dividing line separates two groups of words undergoing despirantisation. Words in the first group show either complete despirantisation (DC), with spirantised forms failing to survive in Standard English, while the rest, exhibiting sporadic despirantisation (DS), develop occasional spellings with ·dÒ, which are sooner or later replaced by the digraph ·thÒ or its equivalents. Analogously, words in the other group show spirantisation, either complete (SC) or sporadic (SS). Words in the latter group exhibit occasional ·thÒ spellings which apparently conceal the non-surviving voiced dental fricative.

3. Despirantisation: Complete and sporadic Complete despirantisation ([d] > [ð]) is a process which captured relatively few words. All those in which the new plosive has survived are of Germanic origin. The chronological and geographical conditioning of the change is presented in Table 2 below. The table contains (a) the contemporary and the Old or Middle English source form, the date of (b) the first and (c) the last recorded form with the original voiced dental fricative [ð], (d) the date of the first occurrence of

254 Jerzy Welna

Table 2.Despirantisation [ð > d]: complete ·ð/thÒ

·dÒ

Dialects

a

b

c

d

e

burden (OE byrðen) swaddle (OE swaðel) murder (OE morðor) fiddle(r) (OE fiðele fiðlere) rudder (OE roðer) spider (OE spı¯ðra) staddle (OE staðol)

c890 c1200 a900 a1100

1894 1615 1836 a1475

c1000– 13…– c1325– a1425–

SW, WM, EM, EM, N, Sc N : EM EM : WM : EM EM

c725 OE c825

1819 1480 1559

EM EM EM

afford (OE Šeforðian)

c1050

a1600

c1440– c1440– c1481(?)/ 1573– 1514–

(MoE)

Table 3.Early occurrences of despirantised forms byrdene “burden” suadiling “swaddling” mourdren “murder” fydelin “fiddler” rodyr “rudder” spyder “spider” stadelynges “staddlings” aforde “afford”

c1000 Laws Northumbr. Priests c55 in Schmid Gesetze 368 13… Cursor M. 1343 (Gött.) c1325 Poem Times Edw. II 430 in Pol. Songs (Camden) 343 c1440 Promp. Parv. 159/2 c1440 Promp. Parv. 38/2 c1440 Promp. Parv. 140/2 1481 City Lond. Letter Bk. 1f. 163b 1514 Barclay Cytezen & Uplandysh. (1847) 69

spelling with ·dÒ, and, whenever possible to identify, (e) the name of the dialect(s) in which the form with [d] occurs. Items are arranged chronologically according to the date of the first d-spelling for the earlier ·thÒ or its spelling equivalents, i.e. according to the chronology of (d) (Table 2). The dialectal distribution unambiguously indicates that the chronologically late forms belong almost exclusively to East Midland, from where they must have spread to other dialects. As an illustration, Table 3 above lists the earliest occurrences of despirantised forms. The listing follows chronological order. As regards chronology, the few words undergoing complete despirantisation are evidence that the process must have been at work from the fourteenth century onwards. An evident gap in the occurrence of the plosive between Old English and the fourteenth century compels one to treat ONhb byrdene “burden” as an exception. Curiously, as many as three of the eight earliest instances of words exhibiting loss of friction (fydelin “fiddler”, rodyr “rudder” and spyder “spider”) come from the Promptorium Parvulorum, a mid-fifteenth-century

Spirantisation and despirantisation 255

document from Norfolk (East Midland), whose author(s) employed a specific spelling convention. The most interesting development is that of OE byrðen “burden”, a word with a long history. Its earliest d-form byrdene, found in Old Northumbrian, is, as said above, an isolated spelling. A chronological arrangement of the earliest spellings with ·dÒ standing for the former fricative reveals a very characteristic dialectal pattern which may be a reflection of how the change diffused in dialects (see Table 4 below). Table 4.Burden (a) byrdene burdene burden burdene birdin byrdoun burdeyne byrdyngs

c1160 Hatton Gosp. Matt. xxiii 4 c1175 Lamb. Hom. 5 c1200 Trin. Coll. Hom. 65 c1250 Gen. & Ex. 1467 a1300 Cursor M. 6830 c1303 R. Brunne Handl. Synne 11959 c1440 York Myst. xxxii. 114 c1470 Henry Wallace xi. 29

(b) burden burden

c1374 Chaucer Boeth. 101 c1489 Caxton Sonnes of Aymon (1885) 131

The above sequence of d-spellings in Table 4a may indicate a correlation between time and dialect. Thus, the change may have been initiated in the Central Southwest (Hatton Gospels; c1160), spread to southern West Midland (Lambeth Homilies; c1175), western East Midland (Trinity College Homilies; c1200), eastern East Midland (Genesis and Exodus; c1250), North Midland (Handlyng Synne
; 1303) and the North (Cursor Mundi; c1300; also York Plays; c1440), ultimately reaching Scotland (Wallace
; c1470). These d-spellings must have simultaneously spread to the Southeast, especially to London (cf. samples from Chaucer and Caxton; Table 4b) where, until the nineteenth century, they competed with th-forms. A similar competition of d-spellings and th-spellings continuing almost to our times can be encountered in the history of the word murder. The list of words where the grapheme ·dÒ only sporadically replaced the original th-type graphemes is significantly longer than that containing instances of complete despirantisation. Before adducing the relevant examples it must be stated that while ·thÒ substituted for ·dÒ is an almost certain indication of spirantisation of the plosive, the grapheme ·dÒ for the earlier ·thÒ in cases when

256 Jerzy Welna

the plosive failed to survive in English may represent inverse spelling devoid of phonological significance. Because the true phonological value of the spelling ·dÒ in such words is difficult to determine it is safer to speak about graphemes rather than phonemes. Examples in Table 5 are only those words which developed d-spellings before 1500. The arrangement of items follows chronological order, according to the first occurrence of a potentially despirantised form. The table contains the following items: (a) the contemporary form of the word and, in parentheses, its source form, (b) the date of the first occurrence of the word in English, the first (c) and the last (d) form with d-spelling, and (e) the Middle English dialects in which the forms with ·dÒ occurred. Excluded are certain words, like swither 1501, furthest 1552, pother 1609, lather 1654, and bother 1745, which exhibit d-spellings only in Early Modern English, i.e. after 1500. Entries are arranged chronologically according to (c) above. Table 5.Despirantisation [ð > ð/d > ð]: sporadic ·ð/thÒ

·dÒ

Dialects

a

b

c

d

e

whether (OE hweðer) other (OE o¯ðer) fathom (OE fæðm) farthing (OE fe¯orðing) earth (OE eorðe) loathly (OE la¯ðlı¯c) brother (OE bro¯ðor) rather (OE hraðor) either (OE a¯e¯ Šðer) neither (ME neither) feather (OE feðer) worthy (ME wurði) wether (OE weðer) fother (OE fo¯ðer) fourth (OE fe¯orða) tether (ON tjóðr) farther (ME ferðer) leather (OE leðer) further (OE furðor) smother (ME smorðer) rother (OE hryðer) nether (OE neoðera)

c800– 688– c725– c950– c825– c900– a900– c850– 869– c1200– a800– c1220– 835– 900– c950– 1376– a1300– c1000– c888– c1175– 805– c825–

c800 c897 a1175 a1100 c1175 c1175 c1200 c1205 a1225 a1300 a1300 a1300 1300 1375 1375 1396 1398 c1400 14.. a1450 1474 1486

1535 1794 1785 1722 1865 c1400 1688 1422 1460 1537 1606 1804 1917 1866 1650 a1734 1652 1568 1896 1565 1518 1525

WS/N, EM, SW, Sc WS/WM, EM, SW, N, Sc WS/N, WM, EM EM Merc/WS, WM, EM, SW, N, Sc EM, SW, WM, N, K EM, N, WM, London WM EM, N N, EM N, WM, EM Sc, N, EM N, Sc, SW Sc Sc, WM N, Sc SW, WM, EM WM, SW, N (unclassified) EM (unclassified) N

Spirantisation and despirantisation 257

Table 6.D-spellings hwaeder “whether”

eordunge “earthing” noder “nother” (obs.) ferdin’ ‘farthing’ byrdene ‘burden’

800–900 in O. E. Texts 444 c1000 Ælfric Ex. xxi. 13 c1000 Ælfric Gloss. in Wr.-Wülcker 158/10 825 Vesp. Psalter cvi. 7 c897 K. Ælfred Gregory’s Past. Care li. 399 a1000 Exon Domesday fol. 356 c1000 Laws Northumbr. Priests c55 in Schmid Gesetze 368

The data show a significant spread of despirantisation at the turn of the thirteenth century. Between 1175 and 1220 the change diffused to affect several high-frequency or relatively frequent items, such as earth, fathom, brother and loathly. Although these words exhibit d-spellings in all dialects, the plosives which those d-spellings symbolise have failed to win the competition with the original fricatives. Unlike words in Table 2, none of the forms with a plosive has survived into Modern Standard English. Also, data from the corpora, especially from the OED, abolish the myth of a rather late appearance of both despirantised and spirantised forms. Those with a fricative spelt ·dÒ are found early in both southern and northern dialects. As regards regional distribution of despirantisation, complete and sporadic, Tables 7–10 below show how d-forms are scattered in dialectally important texts. A comprehensive list of examples from the OED contains words coming from the period between the date of the first occurrence of a d-spelling and the conventional end of Middle English, i.e. the year 1500. Words are adduced in their contemporary forms. The Southwest offers numerous cases of sporadic despirantisation (burden is the only word which retains the despirantised plosive in Modern English). Table 7.South Western d-forms

:

c1250 Sur (eSW) Owl and Night. (3) (earth 2, other) : c1297 Glouc (nSW) R. Glouc. (1) (earth) : 1380–88 Oxf (nSW) Wycliffe (1) (burden) : 1387–99 Gloucester (nSW) Trevisa

(5) (farther 3, leath- : er, wether)

th-forms (0) (2) (brother, other) (11) (burden 3, other 3, brother 2, neither, wether, worthy) (10) (other 3, burden 2, earth 2, feather, rather, weather)

258 Jerzy Welna

Table 8.West Midland

c1175 Staf (sWM) Lambeth Hom. c1205 Worc (sWM) Layamon A c1225 (sWM) Ancr. R. c1325 Lanc (nWM) E. E. All. Poet. c1400 Lanc (nWM) Destr. of Troy 1426 Shrop (wWM) Audelay

d-forms

:

th-forms

(8) (earth 5, other 2, burden) (5) (earth 2, other 2, rather) (3) (loathly) (2) (earth, loathly) (4) (brother, farther, leather, loathly) (3) (other 2, brother)

:

(1) (other)

:

(5) (fiddle 2, burden, brother, fathom) (1) (rather) (4) (fathom 2 fiddle, feather) (6) (murder 2, brother, earth, feather, further) (0)

: : : :

Characteristically, forms with ·thÒ prevail in Trevisa, and particularly in Wycliffe, while texts by other writers offer only scant evidence. Still, the ratio 10 to 23 confirms a significant role of despirantisation in South Western. In the Middle English period, the West Midland dialect seems to have been the leader in implementing minor phonological changes, especially in the area of consonants (Wełna, forthcoming). The material shows that instances of despirantisation are more numerous in South Midland than in North Midland. A striking prevalence of plosives over fricatives is observed in the Lambeth Homilies, an early text of c1175. The total ratio between d-spellings and th-spellings is 25 to 17. In general, fourteenth century London poets avoided using d-spellings in words belonging to our corpus. Chaucer’s and Gower’s texts supply only d-forms of murder, which, however, may reflect OF murdre rather than OE morþor. Only a century later d-spellings became relatively numerous in London, and considerable numbers are found in Caxton’s language. On the other hand, scribes from Norfolk, especially those working on the Promptorium Parvulorum MS, evidently developed fondness for d-spellings, using them instead of the regular th-digraph in the relevant words. Such forms are also prevalent in Genesis and Exodus, another literary text from Norfolk. Table 10 shows the listing of forms from the Promptorium Parvulorum together with their Latin or French correspondences. There is only one instance of a th-spelling (rothyr). The East Midland total ratio between d-forms and th-forms is 31:25 (see Table 9). The only text rich in lexical material relevant to the present study is the Cursor Mundi, whose various versions may belong to Lincolnshire (nEM),

Spirantisation and despirantisation 259

Table 9.East Midland

c1200 Hunt (cEM) Trin.Coll. Hom. c1250 Norf (eEM) Genesis & Ex. c1300 Linc (nEM) Havelok c1330 Linc (nEM) R. Manning of Brunne 1366–86 Lond (sEM) Chaucer

1440 Norf (eEM) Prompt. Parvulorum 1474–91 London (sEM) Caxton

d-forms

:

th-forms

(5) (other 2, burden, brother, earth) (4) (earth 2, burden, loathly) (1) (fathom)

:

(1) (other)

:

(0)

:

(4) (burden, farthing, other, whether) (5) (fiddle 2, earth, other, rather) (10) (brother 2, fother 2, other 2, murder, feather, loathly, whether) (1) (rudder)

(2) (burden, murder) : (3) (murder)

:

(10) (fathom 3, feather : 2, fiddle, murder, spider, rudder, brother) (6) (farther 3, burden, : swaddle, farthing)

(4) (spider 1, brother1, loathly 1, other 1)

Table 10.Promptorium Parvulorum Brodyr … germanus. (54) “brother” Eranye, or spyder, or spynnare (140/2) “spider” Fadme, or fadyme, ulna, Fadmyn (fadomyn, P.) ulno, Fademynge ulnacio (145/2) “fathom(ing)” Fedyr, penna, pluma “feather” Fedyrfu, or fedyrfoy, herbe, febriffuga (152/2) “featherfew” Fydelin, or fyielyn, vitulor (159/2) “fiddler” Rothyr, or masch-scherel, remulus, palmula, mixtorium (328/1) “rudder” Moorderyn, or prively kyllyn, sicario (342/2) “murdering” Rodyr, of a schyppe, … amplustre (438/2) “rudder”

Lancashire (Fairfax MS: North West Midland), Northumberland (extreme North) or Scotland. In agreement with the tradition, the text will be considered here Northern, as opposed to the unambiguously Scottish cycle of Legends and Barbour’s Bruce. The above material shows only insignificant overlapping of both types of spellings in the Cursor Mundi, with the ratio 21 : 19, and a striking prevalence of d-forms in Barbour’s poem (8 : 0), the total ratio for the two Scottish MSS being 11 : 2 (see Tables 11 and 12).

260 Jerzy Welna

Table 11.Northern d-forms c1300–40 Nth (Sc) Cursor (21) (earth 5, either 4, neiM. ther 3, brother 2, loathly 2, burden, swaddle, fathom, feather, other)

:

th-forms

:

(19) (other 6, brother 3, earth 3, either 3, burden 2, swaddle, whether)

Table 12.Scotland d-forms

:

th-forms

c1375 Sc. Leg. Saints

(3) (other, whether, worthy)

:

(2) (other)

1375 Barbour’s Bruce

(8) (earth 3, wether 2, fother, fourth, whither)

:

(0)

4. Spirantisation: Complete and sporadic Curiously, the instances of spirantisation in Middle English are much less frequent than those of despirantisation, although the change of a plosive into a fricative appears to be more natural than the reverse process. As in the previous section the cumulative list follows the chronology of the first instances of th-spellings replacing former d-spellings; cf. column (d) in Table 13 below. With the exception of the loanword panther, all other items showing spirantisation belong to the earliest stratum of the lexicon of English and are much earlier than those subject to despirantisation (see Table 2). Most often Table 13.Spirantisation [d > ð]: complete

a

b

whither (OE hwider) c825 panther (OF pantere) c1220 together (OE to¯ga¯¯e dere) 707 father (OE fæder) a800 mother (OE mo¯dor) c725 thither (OE þider) c888 weather (OE weder) c725 hither (OE hider) c725 gather (OE gaderian) c725

·dÒ

·ð/thÒ

Dialects

c

d

e

1556 1503 (?) 1631 1597 1589 a1600 1598 1568 1600

c900– a1000/a1550– c1050– a1175– c1200– c1205– c1205– c1340– a1382

WS/N, SW, Sc, WM, EM WS/EM WS/WM, SW, EM N, SW, WM, EM WM, N, EM WM, N, SW WM, EM N, EM, SW EM, WM

Spirantisation and despirantisation 261

found in the Midlands, both East and West, they show concentration of the new forms between mid-eleventh and early thirteenth century. Curiously, ·ðÒ in together is recorded almost 400 years earlier than in the related verb gather. The earliest forms with a spirant are shown in Table 14. Table 14.Earliest spirantised forms nethle [Ep. naeðlæ Erf. nedlæ] “needle” (c725 Corpus Gloss. (Hessels) 421) aðamans “adamant” (c885 K. Ælfred Greg. Past (1871) 270) hwiðer “whither” (9… Ælfred Gregory’s Past. C. xliv. 330 (Cotton MS)) pandher “panther” (a1000 Panther 12 (Gr.)) togæþere “together” (c1050 in Techmer Zeitschrift (1885) II. 123)

In the chronological listing of spirantised forms in Table 15, columns (c–d) contain the first and the last occurrence of th-spellings. Table 15.Spirantisation [d > d/ð > d]: sporadic ·dÒ

·ð/thÒ

a

b

c

needle (OE na¯¯e dl) bedlam (OE bedlem) wonder (OE wundor) fodder (OE fo¯dor) solder (ME soudur) cider (ME sidre) adamant (ME adamant-) under (OE under) elder a. (OE ieldra) bladder (OE blæddre) (ME consider) udder (OE u¯der)

c725– 971– c700– c1000– c1320– c1300– c1225– c825– c888– 700– c1425– a1000–

c725 1200 c1290 a1300 1300 c1315 1386 c1400 1440 1450 1454 1474

Dialects d 1702 a1400 1693 1561 a1500 1477 1450 1785 1515

e (Merc) EM, N SW, N, EM WM, SW N K, SW, EM WS/EM EM N EM EM (unclassified)

As compared with the relatively large number of words with sporadic despirantisation (see 5 above), the list of samples with th-spellings, all potential indicators of spirantisation, is much shorter. Although found over the whole territory, the th-forms are confined to one or two dialects and fail to show diffusion to other areas. These forms are most common in East Midland.

262 Jerzy Welna

The distribution of forms with complete and sporadic spirantisation compared with words whose spelling remained unchanged is presented in Tables 16–20 below. Table 16.Kent

c1315 Kent William of Shoreham

th-forms

:

d-forms

(2) (other, cider)

:

(2) (together, mother)

Table 17.South Western

1297 Glouc (nSW) R. of Gloucester

th-forms

:

d-forms

(1) (weather)

:

(4) (thither 2, together, weather)

c1380–88 Oxford (nSW) Wycliffe (7) (cider 3, father 2, : thither, together)

(32) (father 10, cider 5, whither 3, mother 3, together 3, hither 3, thither 2, gather, bedlam, weather)

1387–98 Glouc (nSW) Trevisa

(15) (mother 6, father 2, cider 2, together, thither, hither, bladder, whither)

(4) (hither2,thither2) :

As regards Kent, our evidence is confined to a modest sample from William of Shoreham, which shows a perfect balance of both types of spellings, while the sample from South Western reveals more numerous instances of spirantisation in the two last decades of the fourteenth century. Wycliffe’s and Trevisa’s manuscripts preserve d-spellings in words like father, mother, whither, together, where the stop is subject to complete spirantisation. As shown earlier, the relevant words in West Midland exhibited a tendency towards despirantisation. It does not come as a surprise that evidence of spirantisation from that region is virtually missing. As regards East Midland, its texts show on the whole a tendency to retain th-spellings only in words whose plosives ultimately undergo complete spirantisation. But surprisingly, Chaucer was very conservative in the use of spirantised forms and so was Caxton, whose manuscripts contain many words preserving the old spelling ·dÒ instead of the expected ·thÒ, even in high-frequency items like father, mother or weather. In this respect Caxton’s spelling convention agrees with that of the Promptorium Parvulorum (Norfolk) where 10 instances of the relevant words exhibit d-spellings (except

Spirantisation and despirantisation 263

Table 18.West Midland th-forms c1205 Worc (sWM) Layamon

:

d-forms

(3) (together, thither, : weather)

(10) (mother 3, thither 3, together 2, father 2)

th-forms

:

d-forms

:

(3) (together 2, father)

:

(19) (father 5, mother 4, weather 3, fodder 2, thither, gather, together, adamant, whither) (1) (weather) (7) (mother 2, whither 2, father, weather, adamant) (2) (weather, together) (0) (4) (mother 3, weather)

Table 19.East Midland

c1200 Linc (nEM) Ormulum

(1) (bedlam) c1374–86 London (sEM) Chaucer (5) (father 2, adamant 2, hither)

1387–88 London (sEM) T. Usk 1390 London (sEM) Gower c1400 Essex (sEM) T. Chestre c1400 Linc (nEM) Emare c1403–33 Suf (eEM) Lydgate 1425–78 Norf (eEM) Paston Let.

(2) (mother, hither) : (4) (father 2, together, : mother) (2) (under, wonder) : (5) (under 5) : (3) (father, weather, : gather) (3) (mother 2, gather) :

c1450–1500 London (sEM) Mirror (4) (gather 2, togeth- : of Our Lady er, bladder) c1470–85 London (sEM) Malory (2) (whither 2) : 1471–90 London (sEM) Caxton

(3) (whither, cider, panther)

:

(10) (mother 3, weather 2, gather 2, whither 2, father) (4) (weather, together, mother, gather) (5) (father 2, mother, gather, weather) (15) (father 5, mother 3, weather 2, hither, together, whither, gather, thither)

rother). In East Midland, the period of the densest concentration of forms with true or alleged spirantisation is the last quarter of the fourteenth and the first decade of the fifteenth century. The Northern dialect is represented by the Cursor Mundi (Table 20). That sample shows that d-forms were very much alive in Northern England. These data fail to confirm the hypothesis of northern influences on the shift of plosives to spirants. Again, especially striking are numerous d-spellings in high frequency words subject to complete spirantisation, such as thither, together, father, mother, hither or weather. It should be pointed out that the six examples

264 Jerzy Welna

Table 20.North th-forms a13001400 Nhb (N) Cursor Mundi,

:

(16) (thither 6, hither 2, : whither 2, wonder 2, mother, solder, bedlam, father)

d-forms (39) (thither 10, together 6, father 5, whither 5, hither 4, mother 4, weather 2, wonder 2, bedlam)

of the adverb whither with t-spellings are found relatively late, in a manuscript from around 1400.

5. Conclusions An analysis of the above data permits one to formulate several tentative conclusions: 1. Despirantisation in Middle English is documented more adequately than the reverse process of spirantisation. 2. Assuming that changes in spelling correlate with the respective changes in pronunciation, despirantisation in medieval English may be postulated to have occurred on four chronological planes: (a) up to 1000 (DC, DS), (b) 1100–1220 (DS), (c) 1300–1325 (DC, DS), (d) 1375–1400 (DS), (e) 1400 onwards (DC, DS). 3. This division is correlated with dialects as follows: (a) all dialects, (b) WM, EM, N, (c) N (DS) and EM (DC), (d) N, Sc, WM (DS), (e) EM (DC). 4. The texts where despirantisation is best reflected in spelling include the Cursor Mundi (14c, N, 21), Promptorium Parvulorum (15c., EM, 10), Lambeth Homilies (12c, WM, 8), and Barbour’s Bruce (8). 5. The distribution in time of spirantisation involves four periods: (a) up to 1050 (SC, SS), (b) 1175–1205 (SC, SS), (c) 1290–1315 (SS) and (d) 1340 onwards (SC, SS). 6. This division is correlated with dialects as follows: (a) WM, EM, SW, (b) WM, EM, (c) SW and (d) EM. 7. The texts in which spirantisation is best evidenced in spelling include the Cursor Mundi (14c, N, 16), Wycliffe (14c, SW, 7) and Chaucer (14c, EM, 5).



Spirantisation and despirantisation 265

References Blumbach, Wolfgang. 1974. Studien zur Spirantisierung und Entspirantiesierung altenglischen Konsonanten. I. Labiale und Tektale. Ph.D. dissertation, Georg-August Universität, Göttingen. Brunner, Karl. 1960. Die englische Sprache. Ihre geschichtliche Entwicklung. Vol I. Tübingen: Niemeyer. Dobson, Eric J. 1968. English Pronunciation 1500–1700. (2nd edn). Vols. I–II. Oxford: Clarendon Press. Fetting, Hans F. 1972. “Some nonphonetic aspects of a fifteenth century sound change”. From Soundstream to Discourse: Papers from the 1971 Mid-American Linguistics Conference ed. by Daniel G. Hays & Donald M. Lance, 12–19. Columbus, Miss.: University of Missouri. Jespersen, Otto. 1909. A Modern English Grammar on Historical Principles. Vol. I. Phonology. London: Allen & Unwin. Jordan, R. 1974 [1925]. Handbook of Middle English Grammar: Phonology translated and revised by E. J. Crook. The Hague: Mouton. Luick, Karl. 1940. Historische Grammatik der englischen Sprache. Vols. I–II. Stuttgart: Tauchnitz. Man´czak, Witold. 1980. Fonética y morfología histórica del Español. Kraków: Uniwersytet Jagiellon´ski. Mincoff, Marco. 1972. English Historical Grammar. 3rd edn. Sofia: Naouka i Izkoustvo. Pope, M. K. 1934. From Latin to Modern French with Especial Consideration of Anglo-Norman. Phonology and Morphology. Manchester: Manchester University Press. Wełna, Jerzy. Forthcoming. “Now you see it, now you don’t, or the fates of the Middle English voiced labial stop in homorganic clusters.” Papers from the 4th International Conference on Middle English, Vienna 4–7 July, 2002 ed. by Nikolaus Ritt et al. Frankfurt: Lang. Wright, Joseph & Elizabeth Mary Wright. 1928. An Elementary Middle English Grammar. London: Oxford University Press.



Subject index

A accommodation 86–88, 136, 177 affixation 193–195, 198, 203–205 see also compounding Afrikaans 118, 122 Aitken, Jack x, 52–55, 58–59, 63–64 Aldredian glosses 180, 181, 186, 187 Allen, Cynthia 239, 244, 248 alliteration 114 alliterative poetry 114 American English 100, 105, 126–127, 157–158, 162, 226–228, 230–236 anglicana 11–13 Anglo-Norman 110, 111 Anglo-Scandinavian 178, 179, 188 ante-date, ante-dating 80, 85, 216 Anthology of Chancery English 5–39 archilexematic 66, 67 B Beowulf 141, 142 BNC see British National Corpus borrowing 83, 84, 90, 102, 188, 190, 221 British English 100, 105, 118–121, 126, 130, 135, 157–160, 230, 236, 248 British National Corpus (BNC) 154, 235, 236, 246, 248 Brown Corpus of American English (Brown) 157, 158, 236, 238, 240–242, 245, 246 C calque 88 canting 41–47 categorisation 91, 92, 95, 103 causative 70–73, 75, 147

Caxton, William 255, 258, 259, 262, 263 Chancery Hand 5, 11–14, 29, 30, 34 Chancery writs 29, 34, 37 Chomsky, Noam 132, 136 Christ and Satan 142 classification 79, 92, 93, 95, 103, 104, 252 clauses of mainprise 30–32 Cockney 116, 119, 120, 121, 123 Cognitive Linguistics paradigm 140 collocation 68, 86, 147 comparative philology 58 compounding 195, 197 see also affixation conceptual category 65, 66, 102 conceptual domain 65 contact 84, 136 French-English 109–111, 113 Norse 178, 180, 188 Corpus of Early English Correspondence Sampler (Parts 1 & 2) (CEECS), 236, 240, 242–244 Courts of Common Law 23, 28 Craigie, William 49–54, 57–59, 214 critical discourse theory 165 D Danelaw 178, 190 Danish 139, 147, 222 definite article 111, 191 dialect 1, 2, 8, 109–119, 196, 197, 201, 208, 256–258, 260–264 Aldred 188 Northumbrian 181 onomastic 215 Scots 50–51, 166–168 dialectology 1, 122 diglossic 180

268 Subject index

distancing effect 171 distinctive feature approach 65 Dobson, E. J. 134 domain 65–68, 70, 73, 74, 91, 237, 238, 240 Dream of the Rood 142, 143, 148 Dublin English 126, 135 dummy it 244 Durham Ritual 112, 180, 190 dysphemism 92, 104 E Early Chancery Proceedings 4, 5, 18, 26, 28, 29, 32 Early Modern English (ENE) 83, 153, 204–206, 226, 236–247, 252, 256 Early Modern English period of the Helsinki Corpus 236 Early West Saxon 112 electronic tagging 89 Ellesmere Chaucer 1, 3, 36 encyclopaedia 52–58 encyclopaedist 49, 54, 55 ENE see Early Modern English English Place-Name Society 40, 213 euphemism 91, 92, 97, 104–107 euphony 194 Exchequer 5, 17, 22, 23 Exeter Book 112 Exodus 142, 255, 258 “exotic” language 81, 84 F Farman 181–184, 187 First Germanic Consonantal Shift 251 Fisher, John H. 3, 5, 8, 10–14, 16, 22–30, 33–35, 37 FLOB (Freiburg-LOB Corpus) 157, 236, 238, 240–242, 245 folk-etymology 216 Fowler, H. W. 157, 159, 193 Freiburg-Brown Corpus of American English see FROWN Freiburg-LOB Corpus of British English see FLOB

French 82, 83, 100–102, 225, 226, 230, 231 forms 153–155 Middle 86, 88 Old 60, 102 FROWN (Freiburg-Brown Corpus) 157, 158, 161, 236, 241, 242, 245 Functional-Lexematic Model (FLM) 65, 66 G Gaelic 57, 163–167, 216, 217 generative phonology 132 German 111, 126, 127, 203–204, 208 Old High 146 Germanic 27, 109, 111, 139, 147, 213, 215, 217–219, 222, 225, 233, 253 Primitive 146 stress pattern 231 glottal stop 113, 114, 119, 122 grammatical gender 188, 190 grapheme 152, 255, 256 Great Vowel Shift 125 Greek, Homeric 195 H h-dropping 109 [h]-loss 111 Halle, Morris 132 Hart’s Rules for Compositors and Readers 155–157 Helsinki Corpus (HC) 236, 241, 242, 244 historical lexicography 49, 51, 58, 213 Historical Thesaurus of English 92 homophone 151–153, 155–161 hypercorrection 109–112, 116, 120 I image schema 140, 145 impersonal uses 239–244 Indo-European 146, 251 Irish English 126, 127, 129, 135 South-west 126–127

Subject index 269

J Jespersen, Otto 239, 252 Johnson, Samuel 152, 153, 155, 225, 227–232 Jonson, Ben 231 L Labov, William 131, 134 LALME (Linguistic Atlas of Late Mediaeval English) 115 Lancaster-Oslo/Bergen Corpus of British English see LOB language change 113, 135 Lass, Roger 118, 119, 226 Latin 29–31, 34, 35, 82–83, 86–89, 110–112, 152, 251 Classical 82, 83, 86–88, 113 Late 60, 110, 113 Medieval 29, 110, 122 Latinism 102 lexical diffusion 133, 136, 253 lexical domain 65, 66, 73, 74 lexical field 76, 92 lexicon 91, 194, 198, 207, 208 lexicon vs onomasticon 218 Lindisfarne Gospels 112, 178, 180, 181, 185–187 loan 80–90, 177–189, 225–233 loanword 79–85, 111, 177–189, 225–226, 233, 260 French 225 direct 88–89 indirect 86 Norse 181–182, 184, 187 LOB (Lancaster-Oslo/Bergen Corpus) 157, 236, 242, 245–246 London English 1, 2, 17, 28, 36–37, 128, 129, 132–134 Luick, Karl ix, 110, 252, 253 M Macleod, Iseabail 55 Matheson, Lister 5, 26 mainprise clauses see clauses of mainprise maun “must” 171

mental perception 140, 147 merger 3, 125–137 metaphor 77, 103, 140 metaphorical 75, 140, 142, 147 metonymic 103–105 metonymy 92, 103 Middle Class 101, 109, 135, 163, 169, 171–175 Middle English Dictionary (MED) 9, 252 Middle English 27, 83, 86, 104, 109, 113–115, 129, 146, 177, 205–207, 209, 214, 217, 218, 226–228, 252, 253, 256–258, 260, 264 dialects 256 dictionary of 51 early 75, 102 late 239, 243, 245 Midland(s) 9, 177, 261 Central 2 Central Midland Standard 3–4, 25 East 8, 254–263 North 8, 32, 258 North-west 27, 31, 259, 263 South 258 West 27, 190, 255–262 Milroy, James 109–111, 128, 178, 180 Minkova, Donka 205, 226, 233 modal 171, 246 see also maun Modern English 194, 198, 201, 204–206, 227, 235–247 see also Present-day English morpheme 59, 86, 186, 197 morphological 80, 85, 87, 133, 181, 188, 194–198, 201, 207 morphology 34, 86, 87, 195, 207 morphosyntactic 66 Murray, James 49, 51, 54, 109, 155–156 N Nares, Robert 225, 228, 229, 232–233 naturalized 86, 87, 89 network 63, 130, 191, 198 non-Chancery forms 25 non-rhotic varieties 121, 127, 128 non-standard English 41, 42

270 Subject index

Norman 110–111, 179, 225 Norse 177–184, 183, 186–189 see also Old Norse Northumberland 178, 259 Northumbria 178, 180, 188, 189 Northumbrian 112, 180, 181, 183, 255 O OED (see Oxford English Dictionary) Old English 65–70, 73–74, 96, 102, 112–114, 139–147, 177, 180, 183, 207, 216, 217, 219–222, 239, 254 see also West Saxon late 190, 204–205 strong verbs 67 Old Icelandic 139, 147 Old Norse 177, 180, 183, 188 see also Norse Old Saxon 146, 147 orthoepic 129 orthoepists 128 orthography 114, 128, 151, 156, 159 see also spelling Owun 178, 181, 183–185, 184–188 Oxford English Dictionary (OED) ix-x, 49–51, 55–57, 59–60, 68, 73, 79–85, 92–93, 95–103, 109, 153–155, 158–159, 213, 218, 223, 235, 246, 252, 257 P paradigm 140, 194, 197, 198, 201, 205, 206 paradigmatic selection 194–197, 201, 203, 204, 208, 209 parallel borrowing 83 Parliament Roll(s) 13, 23, 24, 26, 27, 30 parliamentary petitions 22, 25, 40 PDE see Present-day English phoneme 113, 120, 121, 152 phonemic mergers 125 phonemic variants 49, 58, 61 place-name element 213, 217, 218, 221 place-name studies 213, 215 polysemous 145, 154

polysemy 140, 148, 151, 198 possessive pronoun 111 postdatings 80 pre-Conquest English 113 pre-standard language 208 predicate 66, 68–73, 76 Present-day English (PDE) 109, 112–113, 125, 226, 235–239, 245, 246 prestige 23, 109, 134, 180 low 130 Privy Seal 5, 7, 12, 13, 15–18, 22, 35, 37 Process and Paradigm 197 prototypical 75, 76, 105, 238, 249 R regional standards 37 resyllabification 200, 206 Romance 86, 110–111, 231, 233 RP (Received Pronunciation) 114, 118, 119, 126, 128, 129 Rushworth Gospels 112, 177, 178, 181, 183, 189 S Samuels, M. L. 1–5, 29, 178 Scandinavian 177–180, 182, 187–189, 217 Scandinavian belt 178, 191 schema 75, 76, 146 Scots ix, 50–57, 60, 62, 163–165, 167–169, 171–174, 213, 216–218 Middle 214, 217 Modern 217 Older 51, 57, 59 Scottish Place-Name Survey 214 Scragg, Donald 112–114, 152 secretary (hand) 11–14, 17, 31 semantic change 73, 74, 76, 91 semantic domain 91 semantic field 42, 43, 45, 140 semantic space 66, 67 Shakespeare, William 101, 228, 229, 236, 241, 243, 246 Signet 12–22, 35 Signet letter 7, 14–19, 37 Skeat, Walter 155–156



Subject index 271

socio-historical context 247 sociolinguistics 56, 109, 136, 177–178, 187–189 South African English 118–119, 121–122 Spanish 76, 82, 251 spelling 34, 58–59, 113–121, 151–153, 155–157, 254–258, 262, 264 reformers 152 variation 58 Standard English 6, 26, 29, 33–37, 133, 165, 172, 174, 193–194, 205, 253, 257 standardization 207–209 Statistical Accounts of Scotland 163, 174 Stevenson, James A. C. 53, 55 stress 120, 196, 208, 225–233 first syllable 228–233 fluctuating 225–227, 229, 230 secondary 226, 228 stress shift 226, 227, 234 stressed 225–227, 230, 231 strong ties 178, 179, 187 Survey of English Dialects 119 Swedish 111, 139, 147, 195 Old 222 Sweet, Henry 86, 120, 123 synonymy 198, 201, 205–207

T taboo 92, 104, 105 The Battle of Finnsburh 142 Thesaurus of Old English (TOE) 68, 77, 92, 141 TOE see Thesaurus of Old English toilet 91, 92, 94, 95, 97–106 transmission ix-xi, 81, 83, 90, 178, 180 U unnaturalized 87 V vulgarism 101–102, 105, 235 W Walker, John 227–232 weak ties 178–180, 188 weak verbs 67 Webster, Noah 151, 155, 158 Wells, John 226, 227, 230, 232 West Saxon 112, 113, 121, 187, 215 working class 173 would like (to) 240, 245

CURRENT ISSUES IN LINGUISTIC THEORY E. F. K. Koerner Zentrum für Allgemeine Sprachwissenschaft, Typologie und Universalienforschung, Berlin The Current Issues in Linguistic Theory (CILT) series is a theory-oriented series which welcomes contributions from scholars who have significant proposals to make towards the advancement of our understanding of language, its structure, functioning and development. CILT has been established in order to provide a forum for the presentation and discussion of linguistic opinions of scholars who do not necessarily accept the prevailing mode of thought in linguistic science. It offers an alternative outlet for meaningful contributions to the current linguistic debate, and furnishes the diversity of opinion which a healthy discipline must have. A complete list of titles in this series can be found on the publishers website, www.benjamins.com/jbp 210 DWORKIN, Steven N. and Dieter WANNER (eds.): New Approaches to Old Problems. Issues in Romance historical linguistics. 2000. xiv, 235 pp. 211 ELŠÍK, Viktor and Yaron MATRAS (eds.): Grammatical Relations in Romani. The Noun Phrase. with a Foreword by Frans Plank (Universität Konstanz). 2000. x, 244 pp. 212 REPETTI, Lori (ed.): Phonological Theory and the Dialects of Italy. 2000. x, 301 pp. 213 SORNICOLA, Rosanna, Erich POPPE and Ariel SHISHA-HALEVY (eds.): Stability, Variation and Change of Word-Order Patterns over Time. with the assistance of Paola Como. 2000. xxxii, 323 pp. 214 WEIGAND, Edda and Marcelo DASCAL (eds.): Negotiation and Power in Dialogic Interaction. 2001. viii, 303 pp. 215 BRINTON, Laurel J. (ed.): Historical Linguistics 1999. Selected papers from the 14th International Conference on Historical Linguistics, Vancouver, 9–13 August 1999. 2001. xii, 398 pp. 216 CAMPS, Joaquim and Caroline R. WILTSHIRE (eds.): Romance Syntax, Semantics and L2 Acquisition. Selected papers from the 30th Linguistic Symposium on Romance Languages, Gainesville, Florida, February 2000. 2001. xii, 246 pp. 217 WILTSHIRE, Caroline R. and Joaquim CAMPS (eds.): Romance Phonology and Variation. Selected papers from the 30th Linguistic Symposium on Romance Languages, Gainesville, Florida, February 2000. 2002. xii, 238 pp. 218 BENDJABALLAH, Sabrina, Wolfgang U. DRESSLER, Oskar E. PFEIFFER and Maria D. VOEIKOVA (eds.): Morphology 2000. Selected papers from the 9th Morphology Meeting, Vienna, 24–28 February 2000. 2002. viii, 317 pp. 219 ANDERSEN, Henning (ed.): Actualization. Linguistic Change in Progress. Papers from a workshop held at the 14th International Conference on Historical Linguistics, Vancouver, B.C., 14 August 1999. 2001. vii, 250 pp. 220 SATTERFIELD, Teresa, Christina M. TORTORA and Diana CRESTI (eds.): Current Issues in Romance Languages. Selected papers from the 29th Linguistic Symposium on Romance Languages (LSRL), Ann Arbor, 8–11 April 1999. 2002. viii, 412 pp. 221 D'HULST, Yves, Johan ROORYCK and Jan SCHROTEN (eds.): Romance Languages and Linguistic Theory 1999. Selected papers from ‘Going Romance’ 1999, Leiden, 9–11 December 1999. 2001. viii, 406 pp. 222 HERSCHENSOHN, Julia, Enrique MALLÉN and Karen ZAGONA (eds.): Features and Interfaces in Romance. Essays in honor of Heles Contreras. 2001. xiv, 302 pp. 223 FANEGO, Teresa, Javier PÉREZ-GUERRA and María José LÓPEZ-COUZO (eds.): English Historical Syntax and Morphology. Selected papers from 11 ICEHL, Santiago de Compostela, 7–11 September 2000. Volume 1. 2002. x, 306 pp. 224 FANEGO, Teresa, Belén MÉNDEZ-NAYA and Elena SEOANE (eds.): Sounds, Words, Texts and Change. Selected papers from 11 ICEHL, Santiago de Compostela, 7–11 September 2000. Volume 2. 2002. x, 310 pp. 225 SHAHIN, Kimary N.: Postvelar Harmony. 2003. viii, 344 pp. 226 LEVIN, Saul: Semitic and Indo-European. Volume II: Comparative morphology, syntax and phonetics. 2002. xviii, 592 pp. 227 FAVA, Elisabetta (ed.): Clinical Linguistics. Theory and applications in speech pathology and therapy. 2002. xxiv, 353 pp. 228 NEVIN, Bruce E. (ed.): The Legacy of Zellig Harris. Language and information into the 21st century. Volume 1: Philosophy of science, syntax and semantics. 2002. xxxvi, 323 pp.

229 NEVIN, Bruce E. and Stephen B. JOHNSON (eds.): The Legacy of Zellig Harris. Language and information into the 21st century. Volume 2: Mathematics and computability of language. 2002. xx, 312 pp. 230 PARKINSON, Dilworth B. and Elabbas BENMAMOUN (eds.): Perspectives on Arabic Linguistics. Papers from the Annual Symposium on Arabic Linguistics. Volume XIII-XIV: Stanford, 1999 and Berkeley, California 2000. 2002. xiv, 250 pp. 231 CRAVENS, Thomas D.: Comparative Historical Dialectology. Italo-Romance clues to Ibero-Romance sound change. 2002. xii, 163 pp. 232 BEYSSADE, Claire, Reineke BOK-BENNEMA, Frank DRIJKONINGEN and Paola MONACHESI (eds.): Romance Languages and Linguistic Theory 2000. Selected papers from 'Going Romance' 2000, Utrecht, 30 November–2 December. 2002. viii, 354 pp. 233 WEIJER, Jeroen van de, Vincent J. van HEUVEN and Harry van der HULST (eds.): The Phonological Spectrum. Volume I: Segmental structure. 2003. x, 308 pp. 234 WEIJER, Jeroen van de, Vincent J. van HEUVEN and Harry van der HULST (eds.): The Phonological Spectrum. Volume II: Suprasegmental structure. 2003. x, 264 pp. 235 LINN, Andrew R. and Nicola McLELLAND (eds.): Standardization. Studies from the Germanic languages. 2002. xii, 258 pp. 236 SIMON-VANDENBERGEN, Anne-Marie, Miriam TAVERNIERS and Louise J. RAVELLI (eds.): Grammatical Metaphor. Views from systemic functional linguistics. 2003. vi, 453 pp. 237 BLAKE, Barry J. and Kate BURRIDGE (eds.): Historical Linguistics 2001. Selected papers from the 15th International Conference on Historical Linguistics, Melbourne, 13–17 August 2001. Editorial Assistant: Jo Taylor. 2003. x, 444 pp. 238 NÚÑEZ-CEDEÑO, Rafael, Luis LÓPEZ and Richard CAMERON (eds.): A Romance Perspective on Language Knowledge and Use. Selected papers from the 31st Linguistic Symposium on Romance Languages (LSRL), Chicago, 19–22 April 2001. 2003. xvi, 386 pp. 239 ANDERSEN, Henning (ed.): Language Contacts in Prehistory. Studies in Stratigraphy. Papers from the Workshop on Linguistic Stratigraphy and Prehistory at the Fifteenth International Conference on Historical Linguistics, Melbourne, 17 August 2001. 2003. viii, 292 pp. 240 JANSE, Mark and Sijmen TOL (eds.): Language Death and Language Maintenance. Theoretical, practical and descriptive approaches. With the assistance of Vincent Hendriks. 2003. xviii, 244 pp. 241 LECARME, Jacqueline (ed.): Research in Afroasiatic Grammar II. Selected papers from the Fifth Conference on Afroasiatic Languages, Paris, 2000. 2003. viii, 550 pp. 242 SEUREN, Pieter A.M. and Gerard KEMPEN (eds.): Verb Constructions in German and Dutch. 2003. vi, 316 pp. 243 CUYCKENS, Hubert, Thomas BERG, René DIRVEN and Klaus-Uwe PANTHER (eds.): Motivation in Language. Studies in honor of Günter Radden. 2003. xxvi, 403 pp. 244 PÉREZ-LEROUX, Ana Teresa and Yves ROBERGE (eds.): Romance Linguistics. Theory and Acquisition. Selected papers from the 32nd Linguistic Symposium on Romance Languages (LSRL), Toronto, April 2002. 2003. viii, 388 pp. 245 QUER, Josep, Jan SCHROTEN, Mauro SCORRETTI, Petra SLEEMAN and Els VERHEUGD (eds.): Romance Languages and Linguistic Theory 2001. Selected papers from 'Going Romance', Amsterdam, 6–8 December 2001. 2003. viii, 355 pp. 246 HOLISKY, Dee Ann and Kevin TUITE (eds.): Current Trends in Caucasian, East European and Inner Asian Linguistics. Papers in honor of Howard I. Aronson. 2003. xxviii, 426 pp. 247 PARKINSON, Dilworth B. and Samira FARWANEH (eds.): Perspectives on Arabic Linguistics XV. Papers from the Fifteenth Annual Symposium on Arabic Linguistics, Salt Lake City 2001. 2003. x, 214 pp. 248 WEIGAND, Edda (ed.): Emotion in Dialogic Interaction. Advances in the complex. 2004 vi, 280 pp. + index. 249 BOWERN, Claire and Harold KOCH (eds.): Australian Languages. Classification and the comparative method. 2004. xii, 377 pp. (incl. CD-Rom). 251 KAY, Christian, Simon HOROBIN and Jeremy SMITH (eds.): New Perspectives on English Historical Linguistics. Selected papers from 12 ICEHL, Glasgow, 21–26 August 2002. Volume I: Syntax and Morphology. 2004. x, 255 pp. + index. 252 KAY, Christian, Carole HOUGH and Irené WOTHERSPOON (eds.): New Perspectives on English Historical Linguistics. Selected papers from 12 ICEHL, Glasgow, 21–26 August 2002. Volume II: Lexis and Transmission. 2004. xii, 265 pp. + index.