The Cambridge History of the English Language Volume 2 part 1 pdf

The Cambridge History of the English Language is the first multivolume work to provide a comprehensive and authoritative account of the history of English from its beginnings to its presentday world-wide use Its coverage embraces not only areas of central linguistic interest such as syntax, but also more specialised topics such as personal and place names Whereas the volumes concerned with the English language in England are organised on a chronological basis, the English of the rest of the world is treated geographically to emphasise the spread of English over the last three hundred years Volume II covers the Middle English period, approximately 1066-1476, and describes and analyses developments in the language from the Norman Conquest to the introduction of printing This period witnessed important features like the assimilation of French and emergence of a standard variety of English There are chapters on phonology and morphology, syntax, dialectology, lexis and semantics, literary language and onomastics Each chapter concludes with a section on further reading; and the volume as a whole is supported by an extensive glossary of linguistic terms and a comprehensive bibliography The chapters are written by specialists who are familiar both with the period of the volume and with modern approaches to the study of historical linguistics The volume will be welcomed by specialists and non-specialists alike and it will remain the standard account of Middle English for many years to come THE CAMBRIDGE HISTORY OF THE ENGLISH LANGUAGE GENERAL VOLUME EDITOR Richard M Hogg ii 1066-1476 THE CAMBRIDGE HISTORY OF THE ENGLISH LANGUAGE VOLUME ii 1066-1476 EDITED BY NORMAN BLAKE Professor of English Language and Linguistics, University of Sheffield CAMBRIDGE UNIVERSITY PRESS PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE The Pitt Building, Trumpington Street, Cambridge, United Kingdom CAMBRIDGE UNIVERSITY PRESS The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York, NY 10011-4211, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia Ruiz de Alarcon 13, 28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa http://www.cambridge.org © Cambridge University Press 1992 This book is in copyright Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press First published 1992 Fourth printing 2006 Printed in the United Kingdom at the University Press, Cambridge A catalogue record for this book is available from the British Library Library of Congress Cataloguing in Publication data The Cambridge history of the English language Vol edited by Norman Blake Includes bibliographical references and indexes Contents: v The beginnings to 1066 - v 1066-1476 I English language - History I Hogg, Richard M II Blake, N F (Norman Francis) PE1072.C36 1992 420'.9 91-13881 ISBN 0-521-26474-X (v 1) ISBN 0-521-26475-8 (v 2) ISBN 521 26475 hardback UP CONTENTS List of maps List of contributors General Editor's preface Acknowledgements List of abbreviations 1.1 1.2 1.3 1.4 1.5 1.6 INTRODUCTION Norman Blake Beginnings of the study of Middle English The study of Middle English since the Second World War English, French and Latin Spelling and standardisation Social and literary developments Concluding remarks Further reading page x xi xiii xvii xviii 1 15 20 21 PHONOLOGY AND MORPHOLOGY Roger Lass 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 23 Introduction Phonology, origins: the Old English input system The formation of the Middle English vowel system Consonantal developments Length and quantity Accentuation When did Middle English end? Morphology: general matters Morphology: the major syntactic classes Further reading Textual sources 23 39 42 57 67 83 Vll 90 91 103 147 154 Contents MIDDLE ENGLISH DIALECTOLOGY James Milrqy 3.1 3.2 3.3 3.4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5.1 5.2 5.3 5.4 5.5 Dialect method and the study of Middle English The study of geographical variation in Middle English Variation theory and Middle English dialectology Concluding remarks Further reading 156 156 167 192 201 204 S Y N T A X Olga Fischer 207 Introduction The noun phrase The verb phrase Questions Negation Composite sentences Agreement Word order Some grammatical processes Further reading Textual sources 207 210 233 278 280 285 364 370 383 391 398 LEXIS AND SEMANTICS David Burnley Lexis Foreign influences Word formation The structure of the lexicon Semantics Meaning, use and structure Semantic change Further reading 409 409 414 439 450 461 461 485 496 THE LITERARY LANGUAGE Norman Blake 6.1 6.2 6.3 6.4 500 Introduction 500 Early Middle English literature Later Middle English literature Special features of the literary language Further reading 508 517 Vlll 532 540 Contents 7.1 7.2 7.3 O N O M A S T I C S Cecily Clark Sources and methodology Anthroponymy Toponymy Further reading 542 542 551 587 604 Glossary of linguistic terms Bibliography Primary sources 607 629 629 Secondary sources Index IX 634 Gil Phonology and morphology emphasis will be on the regions roughly in the line of descent to the modern standards I will, however, refer to others here where their histories shed light on general trends in the evolution of English, and where they contribute to the clarification of southern (more accurately southeastern and southeast midland) developments: either by retaining older features or by being more innovative, or where features now standard originated outside the southeast midlands The shift of the political centre from Wessex to London during the post-Conquest period creates problems for the historian Because of the geographical distribution of surviving Old English texts, there is no well-attested corpus directly ancestral to the modern southern standard — which in any case did not begin to emerge until the fourteenth century The bulk of the ' classical' Old English texts are in dialects (loosely) ancestral to those of the modern south-west and southwest midlands (Dorset, Devon, Hampshire, Wiltshire, Gloucestershire, etc.); the rather small number of texts from the south-east are in the line of succession to the modern local dialects of Kent and Sussex The main contributors to the modern standard, the southeast midlands dialects, are mainly Anglian, not West Saxon This part of the country (e.g Essex, Middlesex, Surrey) is not well represented until Middle English times The English of Chaucer or of the fourteenth—fifteenth-century Chancery, which are roughly precursors of 'our' English, not have a detailed Old English ancestry Even the dialect of London itself is a hybrid, an emergent Late Middle English type combining south-west (Westminster) and southeast (Essex, City of London) sources Sitting as it does on the Thames, a crossroads where East Anglia, Kent, the West Country and the midlands meet, London might be expected to present a regionally complex picture; and this is further complicated by the immigration typical of a capital city The -(e)s marker on present third-person singular verbs, for instance, is probably northern (brought in via East Anglia), as are lexical items like hale (the northern form of whole: see 2.3.2); while the initial / v / in vat, vixen, vial are southernisms (see 2.4.1.1), and the /&/ in merry,fledge,bury is Kentish — though the < u > spelling of bury, like the / A / pronunciation of cudgel, shut, is southwestern (3.4) The Middle English continuum is conventionally divided into five major dialect areas, roughly as shown (for refinements see ch of this volume) In terms of Old English, northern = Northumbrian, east midland and the northern two-thirds of the west midlands = Mercian, 33 Roger Lass SOUTH EASTERN 2.1 The dialects of Middle English 34 Cambridge Histories Online © Cambridge University Press, 2008 Phonology and morphology southwestern and the southern third of the west midlands = West Saxon, and southeastern = Kentish Because of the dialectological complexities, and the fact that the bulk of surviving Old English lexis is West Saxon, I will normally cite the usual dictionary forms for Old English lexemes Throughout this chapter the term ' Old English' — unless further specified — will be used more or less the way we use ' English' now: to denote a ' common core' of features characterising the bulk of the dialects 2.1.5 Middle English orthography Old English (see vol I, ch 3) was written in a modified Latin alphabet; the letters < k > , < q > and < z > were used rarely, and < j > and < v > not at all On the other hand, the alphabet was augmented by the digraph , the crossed < d > or'eth' < > , and two symbols of runic origin, > 'thorn' and < p > 'wynn' Both were used for / / , and < p > , or in early texts < u > and < uu > , for /w/ Other points worth noting are the following: The same symbols were used for both palatal and velar consonants: yeldan 'yield' vs vjldan 'gild', cynn 'kin' vs cinn 'chin'; i.e /j g/ were spelled < > , and / t j k/ were spelled < c > (as the modern pronunciations suggest) The insular form of the letter < g > , which I represent here as < > , is usually changed to < g > in modern editions of Old English texts; for this section I retain < > , as its later shape played a special role in Middle English orthography distinct from that of < g > (see 5, below) The cluster < c g > represented the palato-alveolar affricate /d^(:)/ as in brycg 'bridge'; this was also written < g e > after nasals, as in sengean 'singe' (cf singan 'sing') < h > was used for [h], [x], and [c], which were allophones of one phoneme (see 2.2.2, 2.4.1.2), e.g beah 'high' [hasax] = /xaeax/ Vowel length was rarely marked: manuscript %od could be either god 'god' or god 'good' But consonant length was normally indicated (ofer /ofer/ 'over' vs offrian /of:rian/ 'offer': see 2.4.1.1, 2.5.4 for later repercussions) < y > was used only as a vowel letter, for front rounded /y(:)/; the modern use for / j / developed during the eleventh century 35 Roger Lass The establishment of Middle English orthographic norms involves both endogenous changes to the earlier system and the introduction of French usages The most important developments are as follows: Old English had no voice contrast in its fricatives: [v z 3] were allophones of / f s / (see 2.2.2, 2.4.1.1) Therefore it used < f > for [f v], < s > for [s z], and < j?/d> for [9 5] With the rise of a phonemic voice opposition, French < z > was introduced (cf %eal < OF %ele vs seal < OE segel), and < u/v > (see 2, below) began to be used for [v] in both loans and native words (French vice, virgin vs native over, wolves, formerly ofer, wulfas) Many modern editions normalise the original conventions for the use of < u > and < v > The standard medieval distribution is < v > initially for both / u / and / v / (vp, virgin) and < u > medially (haue 'have', but); though some texts use < v > initially only for / v / , and < u > initially for /u(:)/ and medially for /u(:)/ and / v / (e.g Ancrene Wisse has van 'foes' but luue 'love', ure 'our') At the very beginning of the period, < p > alternates in some texts with < u / u u > for / w / ; beginning in the thirteenth century this is replaced by northern French < w > , which is standard by the fourteenth century The / / : / / opposition has never been discriminated in English spelling (cf thigh, thy still) < > began to yield to < ]> > in the thirteenth century, though it remains sporadically through the fourteenth In general, > is the spelling for / 6/, though < th > appears in the twelfth century, and begins to take over towards the end of the period A modified version of < \ >, virtually indistinguishable from < y > , remains in use well into the Early Modern period in abbreviations likej e = pe: hence, by later misunderstanding, Ye Olde Tea Shoppe In around the twelfth century, Franco-Latin (Caroline) < g > is introduced, mainly for / g / but also for /&$/, in place of insular < > The latter, or some modification, then takes over the representation of /)/ {^ow 'you') and / x / (nyp) Later on < > for /]/ began to yield to < y > , following French practice Early texts sometimes have < i > for /)/, as in the Peterborough Chronicle's iafen ' gave' < OE $eafon beside conservative gxre ' year (dat sg.)' OE ^eare The letter < > continued in use well Phonology and morphology 10 11 12 into the fifteenth century Early printers, especially in Scotland, often substituted < z > for it, producing 'false' spellings of Scots names and other items Thus Dal^iel, Menses, capercailzie, where old = palatalised /n j P/- This led to spellingpronunciations, so that now non-Scots tend to say /menzi:z/ for native /minis/ (OSc /n j / > /rj/), and have no idea what to make of Dal^iel, which is /dijel/ In Old English /d^,/ appeared only postvocalically; in Middle English it occurs in French loans initially as well (Joy, jewel), spelled < i > or < j > These two graphs are more or less equivalent, but is rarer; in some texts it occurs for / i / in certain positions, e.g.y'«' in', and commonly in Roman numerals like iij, etc OE / x / is represented in earliest texts by < h > ; later < > and < g h > become common in non-initial positions, and < ch > is used in the north (as still in Sc bocht, nicbf) By the end of the period < g h > is the norm except for the foot-initial allophone [h], which is still spelled < h > Velars and palatals become distinct in spelling, with < c h > , < cch > , for /tJ/ and F < dg(e) > replacing both < eg > and < gg > in words like bridge < k > , rare in Old English, becomes common for /k/, substituting for older < c > before < i e l n > and postvocalically, with < ck > typical after short vowels (kiss, corn, back); < c > also appears before front vowels for / s / in French loans, e.g condicioun /kondisiu:n/ OE < s c > for llj is gradually replaced by < s h > , < s c h > , though some dialects use < s > , < s s > In some Late East Anglian texts / J / in certain contexts, e.g initially in modal auxiliaries, is spelled < x > : the Book of Margery Kempe (ca 1438) has xal ~ schal, xulde ~ schulde The digraph < ae > is lost in the thirteenth century, replaced by < a > or < e > if short (depending on dialect), and < e e > , < e a > and < e e > if long (see 16 below and 2.5.4) The latest use of appears to be in a proclamation of Henry III of 1258 After the unrounding of OE /y(:)/, leading to merger with /i(:)/ in non-western dialects (see 2.3.4), the graph < y > becomes equivalent to < i > ; hence ME /i(:)/ can be spelled < i > or < y > As early as the Peterborough Chronicle 1137 we 37 Roger Lass 13 14 15 16 find both spellings in the same word, e.g drihten ~ dryhten 'God' < OE dryhten In those dialects that retain distinct /y(:)/, it comes to be spelled after the French fashion with < u > (Ancretie Wisse has buggen 'buy' < OE bycgari) Because of various later developments, OE < e o > spellings are retained in some dialects to represent mid front rounded /&{:)/ This vowel is also spelled < oe/ue > , and in some texts < o > or < u > (see 2.3.4) In later medieval hands, the letters < i u m n > were typically written as sequences of unligatured verticals or ' minims': s, 11, 111, etc This would make something like ambiguous: it could be read as luue or lime The convention arose of writing < o > for < u > in these nasal environments: hence come(n) < cuman, loue < lufu, sonne ' sun' < stmne, etc Under French influence, < o u / o w > were increasingly used to represent /u:/, with < u > being reserved for short / u / and (in those dialects where it remained) /y(:)/; thus hows(e) ~ hous(e) < OE hits, nou < now < OE nit ~ As Middle English progressed, vowel length tended to be more consistently indicated, especially for the mid vowels /e: E: o: o:/, in the form of doubling: see /se:/ 'see' < OE seon or /se:/ 'sea' < OE set; boon /bo:n/ 'boon' < OScand bon, or /bD:n/ 'bone' < OE ban Starting in the fifteenth century, the higher and lower mid vowels were often (though not uniformly) distinguished with < a > as a diacritic for the lower of each pair: < e e > /e:/vs < e a > / s : / , < o o > / o : / v s < o a > /o:/\ F also appears for /e:/ (Thus modern beet/beat, boot/boat represent historical contrasts, merged in the first pair but not the second; for < i e > we have pierce, thief and the like.) In later Middle English, due to the loss of final unstressed vowels and vowel lengthening in open syllables (see 2.5.3.4), the discontinuous representation < VCe > became available for long vowels, with < e > a' dummy' graph or diacritic The original use in words with lengthened vowels like nose /no:z(a)/ < OE nosu is then extended to words that never had a final < e > , as in Chaucer's spelling brode for /bro:d/ 'broad' < OE brdd Consonant length is indicated by doubling (e.g bitter vs biten ' bitten'); after its loss CC spellings can be diacritics for the shortness of a preceding vowel (as already in Orm: see 2.1.3), just as a succeeding < e > can be a diacritic for length (see 2.5.4) Phonology and morphology 2.2 Phonology, origins: the Old English input systems 2.2.1 Vowels This section and the following will repeat material already presented in vol I, ch (if with some differing interpretation); this is necessary for an understanding of the major Middle English developments, and more convenient than requiring constant reference to another book I will start a little further back than may seem strictly necessary, for the sake of historical perspective For Old English of the eighth—ninth centuries, we can assume for all dialects this minimal or 'core' vowel system (on the representation of diphthongs, see below): (1) Short Monophthongs Diphthongs i y u eo o eb a; a tea Long Monophthongs Diphthongs i: y: u: e: 0: o: eo a;: a: asa These are highly symmetrical systems, with a number of features worth noting: (a) phonetic as well as phonological symmetry throughout the long and short systems: matchings like [u]/[u:], [i]/[i:] instead of the modern types [u]/[u:], [i]/[i:] — a feature that was to persist well into the Early Modern period; (b) only three contrastive heights, as opposed to the four that were to develop in the thirteenth century (2.3.2); and (c) diphthongs only of the 'height-harmonic' type, i.e with both elements of the same height, as opposed to the earlier and later closing types like /ai au/, and the much later centring types like / i a / (idea) The long/short diphthong contrast is conventionally indicated by marking the ' l o n g ' ones: beam 'tree' vs eabta 'eight', bed ' b e e ' vs heofon 'heaven' (and see the transcription / e : o / vs / e o / in vol I) I have deliberately reversed the procedure here, using notations like / e o / vs / e b / , to make a theoretical point, which helps explain why the two sets behaved as they did This is that there is nothing 'extra' about the ' l o n g ' diphthongs; they functioned, as diphthongs typically in languages with a long/short vowel contrast, as members of the long subsystem It is the 'short' diphthongs that are the 'abnormal' or ' marked' category, since they pattern with the short vowels We can see this clearly in the late Old English/Early Middle English mergers to be discussed in 2.3.1: ' l o n g ' / e o / in bed falls in with / e : / ingrene 'green', while 'short' / e b / in heofon falls in with short / e / in settan 'set' 39 Roger Lass (compare the vowels in the modern forms, / i : / vs /e/) I will return to this in more detail in 2.3.1 In later periods, the most important change was the unrounding of /o(:)/, leading to merger with /e(:)/; by the stage of Old English that can reasonably be seen as input to Middle English, we can assume for all dialects except Kentish a general vowel system of the type: (2) > y u e as a eb asa i: ye: en: u: o: a: eo aea (/y(:)/, /ae(:)/ did not occur in later Kentish, having merged with /e(:)/: Kt hedan 'hide', dene 'clean' vs WS bydan, dine.) Since reference forms in dictionaries and grammars are often based on West Saxon literary language, although many of the target dialects of Middle English have a different provenance, there is a potentially misleading etymological relation as in the two historical classes of words commonly said to ' have OE x' In West Saxon, two etymological categories have /ae:/, conventionally a^ and i2 In the usual terminology, a!j represents WGmc * / a : / as in WS seed' seed', hxr' hair' (cf G Saad, Haar); se2 is the i-umlaut of WGmc */ai/, as in hstlu ' health', Ixdan 'lead (vb)' (cf G Heil, leiten) The xx/i2 contrast is important, because some Old English dialects have / e : / rather than /as:/ for one or both of these categories, and OE /as:/ and / e : / have different Middle English reflexes (/e:/ vs / e : / : see 2.3.1) It is only in West Saxon that both have /as:/; in Anglian xx is / e : / (sed) vs a?2 with /ae:/ (laidan); in Kentish both have / e : / , and older /y:/ has become / e : / as well In terms of inputs into Middle English: West Saxon (3) &2 'lead(vb)' 'seed' *i e 'green' y 'hide' Anglian ae: ae: ae: e: y: e: e: y: Kentish e: e: e: e: By these criteria, a London poet who rhymed lead and seed would show himself to be of southwestern or Kentish origin, or would at least be perpetrating a non-Anglian rhyme Phonology and morphology 2.2.2 Consonants The late Old English consonant system was: (4) P t b d f s m w tf k d g J X t: d: f: 0: s: P: b: m: n r tj: k: d3: g: x: n: r: 1: i Again quite symmetrical, with a length contrast for all consonants except l\ j w/ While the / C / vs /C:/ contrast in final position may have been moribund in Old English (see vol I, ch 3), the medial contrast remained until late, and its loss triggered a major restructuring of the obstruent system (see 2.4.1.1) A few points with historical relevance: While stops and affricates were paired for voice, the fricatives were not; /f s/ were in most dialects voiceless except medially in the foot (see 2.1.5) Thus [v z d] appeared in native words only preceded by a stressed vowel (followed by an optional liquid or nasal) and followed by an unstressed vowel; they were always voiceless initially, finally and in clusters (see 2.4.1.1) Old English had neither an / h / : / x / contrast nor a phonemic velar nasal; [h] was the foot-initial allophone of /x/, and [r)] an allophone of / n / before velars Both features, unlike 1, remained stable throughout Middle English OE IQI had the fricative allophone [y] between back vowels as in dagas 'day (nom./acc pi.)' /dagas/, [dayas]; intervocalically if preceded by a front vowel it had the allophone [j], as in dxge 'day (dat sg.)' /daege/ [daeje] (It is not clear whether unstressed -e was a front vowel or a central [a]; this is problematic throughout the period; see 2.5.3.) Postvocalic [y j], along with postvocalic [w], played an important part in the formation of the new Middle English diphthongs (2.3.3) OE / I / was 'dark' (velarised or uvularised) in syllable codas (e.g eall 'all' [aeal:]); there is no way of telling whether it had 'clear' allophones prevocalically as in many present-day dialects Dark / I / in codas persisted into Early Modern English in most dialects, even those where it is now clear, as in Northumberland and parts of Durham OE / r / is a matter of Roger Lass controversy; in codas it was clearly back, or had some back coarticulation, and this persisted through Middle English 2.3 The formation of the Middle English vowel system 23.1 The short low vowels and the diphthongs By the eleventh century a number of changes had begun which were to lead to the restructuring of Old English vowel phonology The earliest involve the low vowels / s a / and the diphthongs; they are hard to date accurately, because of the persistence of conservative spellings (see below), but the outlines are fairly clear These developments are hard to understand without an account of the relation between long vowels and 'long' diphthongs, and some idea of what the 'short' diphthongs were probably like Let us assume that a long vowel and a diphthong in the ordinary case are equivalent structures, i.e bimoric: so if/eo/ is a vowel cluster consisting of a short [e] plus a short [o], then /e:/ is a cluster of short [e] + short [e], giving a parallelism /eo/ = /ee/ A syllable peak, that is, has temporal 'slots' of short vowel length, and may have one or two of these A short diphthong, in this framework, is also complex, but only occupies one mora or slot; it is made of two 'half-morae' of different quality, compressed into the normal temporal span of one Thus any process of monophthongisation affecting say /eo/ and /eb/ would be the same in both cases - the only difference is the time-span over which the process occurs The monophthongisations of the Old English diphthongs involved two different types of assimilation between morae In the /aea/ diphthongs the second mora assimilated completely to the first, so that /aea/ > /aeae/ (i.e /ae:/), and /aea/ > /aeae/ (i.e /as/) With the /eo/ diphthongs, the assimilation was bidirectional: the rounding of the second mora spread back to the first, and the frontness of the first forward to the second, producing a ' compromise' quality [0] for both: short /eo/ as in heofon 'heaven' came out with /©/, and long /eo/ as in bed w i t h /&&/ = /&'•/'• (5) Type (a) a; (b) eo Type ^ » a: iea: eo ** iea;(= ic:) *— »» (= rt Systematically, the (a) changes result in loss of a phonetic category and Phonology and morphology merger with an already existing one, i.e loss of an opposition; the (b) changes result in the loss of one phonetic type and the gain of another, with no change in oppositional structure At this stage two older lexical classes have fallen together with others by (a): eabta 'eight', rsett 'rat' now have /ae/, beam 'trec',glsm 'gleam' now have /ae:/; while heofon, be'o 'bee' are still distinct from anything else, though in a different way phonetically (At some time prior to the change to /&:/, some instances of long /eo/ seem to have undergone a change, often called 'accent shift'; this is a presumed shift of syllabicity to the / o / mora, which then becomes long/o:/ and participates in the later history of OE / o : / Thus choose, lose < OE ce'osan, le'osan should have come down as PDE ** cheese, **/eese with / i : / , like freeze < fre'osan; instead they have the reflex of OE / o : / like moon < mona, etc In our model here there is no need for syllabicity transfer or other complex interpretations; what we have is simply a different version of the /eo/ > / o o / change: instead of bidirectional assimilation, the first mora regressively assimilates to the second (a mirror-image of the /aea/ development), i.e /eo/ > /oo/ This version in fact competes with the other; Chaucer's chese, lese for choose, lose show /e:/ < /&:/, i.e the same development as in freeze (For details see Lass 1988.) The changes in (5) restore (if briefly in most dialects) the original system type 2, with two front rounded vowels - though with no diphthongs, or with diphthongs of the new Middle English type (see 2.3.3) just beginning to appear The total inventory of the non-Kentish dialects of the early to mid-eleventh century: (6) i y u eo o ae a i: y: u: e: o: o: £e: a: The new /o(:)/ were quickly merged with /e(:)/ (another repetition of an earlier merger) in all except the southwestern and southwest midland dialects (see 2.3.4) These mergers and phonetic changes resulted in a bewildering profusion of spellings Thus the Peterborough Chronicle for 1127 shows < e a > for OE /eo/ (heald 'held' < he'old), and for / e / (Heanri ~ Henri); as well as for /asa/ {heaued 'head' ~ hxued < he'afod); and < e o > and < e > for /eo/ (eorl 'earl' ~ erl < eorl) The twelfth century is still 'transitional'; the orthographic norms for these categories have not been stabilised (and see below on OE /as/ and / a / ) Following, or perhaps partly overlapping with, this another set of 43 Roger Lass mergers occurred; /as/ (now consisting of original /as/ plus monophthongised /aea/) fell together with / a / In most of the country except parts of the west midlands and south-east this merger resulted in a vowel spelled < a > There is a long-standing debate about this: what did graphic < a > represent at the Old/Middle English transition, and later in the period ? The obvious facts are (a) that the original categories represented by < ae > in rxtt and < a > in catte' cat' merged in Late Old English, and have been indistinguishable ever since; (b) that comparative and historical evidence suggests a low front value for the merger in later Middle English (Lass 1976: ch 4); and (c) that and < a > in Old English represented respectively front and back low vowels A merger of /ae/ and / a / could occur in principle in one of two ways: lowering of/ae/ to [a] and fronting of / a / to the same value; or, as some scholarsiiave suggested (especially those who, like Jespersen (1909-49), believed that ME < a > was also a back vowel), retraction of/as/ to [a], and fronting of the merged result at some later stage (perhaps just before the seventeenth-century change to [ae]) In outline: (7) I The second alternative is obviously less economical, since it involves two backness shifts in opposite directions; the first • is simply a convergence on a new value, perceptually perhaps somewhat 'intermediate' between the two Taking this option, the developments of the short low vowels and /asa/ are (see Lass 1977b): (8) These mergers lead to extensive spelling variation in early texts When formerly distinct categories merge phonetically, but traditional spellings are still remembered, scribes will often use the old spellings in a near-random way, since the actual pronunciations not give a clear indication of how given items should be spelled Thus in the aftermath of the /aea/-/ae/ and /ae/-/a/ mergers, the Peterborough Chronicle for 44 Phonology and morphology 1127 and 1131 shows ~ < e a > ~ < a > variation for many OE /ae/ words: hxfde ~ hafde ~ heafde for hxfde 'he had' and the like The changes in (8) radically altered the shape of the short-vowel system; it now had only one low vowel, instead of the earlier front/back symmetry It also had more vowels at the front than the back, which remained the case until well into the Early Modern period After the mergers, we are left with the following: i y u e a (9) (0) i: re: (o:) EG: u: o: a: In the eleventh or twelfth century, probably after the monophthongisations and mergers discussed above, /ae:/ raised to / e : / At around 1100, then, we would have the following systems (/©(:)/ omitted): (10) i y u o e a i: ye: E: u: o: a: Even though we have only three long unrounded front vowels at this stage, they are distributed differently in the vowel space; there is a potential 'empty slot' at low front which would allow for a new / a : / , just as there had always been one in back between / a : / and / o : / which could accommodate a new /o:/ This is obscured by the representation in (10), which can be redrawn with the empty slots boxed: (11) i: y: u: e: o: • • a: We will see in the next section how these were filled, creating the more even, less gapped systems typical of later Middle English There are empty slots in the short system too: (12) e u o • • • a Two of these, front / e / and back /o/, were indeed filled, but not until about the sixteenth century (see vol Ill, ch 1); later /o/ lowered to / D / , 45 Roger Lass filling in the lower back corner and creating a system much like that of earlier Old English, with /ae D / (cat, pot) balancing each other, and the old /e o/ slots (more or less) filled by the new centralised half-close /i u/ < older /i u/ The short / o / slot remained empty except in Scotland, where / o / never lowered to / D / in native varieties 2.3.2 Filling out the system: OE /a:/ and related matters One striking difference between modern northern and southern dialects (' southern' including historically not only Received Pronunciation and the southern regional dialects, but the extraterritorial ones as well) is the reflex of O E / a : / (e.g bone, stone, home < O E ban, stdn, ham) In non- northern modern dialects, this category has, in conservative varieties, a mid back vowel of the [o:] type; in more 'advanced' (and more widely distributed) ones, it has a diphthong, usually [ou], [au] or [ceu], derived historically from Early Modern / o : / In the 'traditional dialects' (conservative rural vernaculars: Wells 1982) of the north, the vowel is front: Sc /hem/, /sten/, Nbr /hiam/, /stian/ for home, stone To anticipate material to be treated in vol Ill, the northern front vowels in this class presuppose a Middle English low front input /a:/, and the southern back vowels and diphthongs an earlier half-open back vowel /o:/ — both qualities missing from the system in (11) It now remains to unravel the rather complex developments that produced these new vowel types As early as the twelfth century we begin, in non-northern texts, to get a scattering of < o > , < oa > spellings for OE / a : / words The sermon In diebus Dominicis (MS Lambeth 487: Hall 1920: 76ff.), written on the southern/midland border in the twelfth century shows gast 'spirit', lauerd' lord', swa ' so' < OE gast, hlaford, swd as well as on ' one' < OE an, and the variant louerd; the Peterborough Chronicle for 1134 has a few < o > forms, including noht 'nothing' < OE ndht, mor 'more' < mdra; and the Worcester Fragments (late twelfth century) have mainly < o > as in bon 'bone/leg' < ban, more < mdra, but still show some < a > , as in wa ' woe' < wd, lac ' offering' < Idc These spellings reflect a major change in progress, in which OE / a : / rounded and raised to /o:/ Since spelling normally lags behind changes in pronunciation, and changes are implemented by long-term variation weighted in particular directions (see 2.8.3 below), we can assume that the change / a : / > /o:/ was well under way by the time we have any graphic evidence for it Despite its variable implementation in different texts and different areas, 46 Phonology and morphology we can date it as coming to fruition in the late twelfth to early thirteenth century, beginning in the south-east and spreading northwards - and constituting from that time on one of the major north/south isoglosses At about the same time, OE / a : / in the north was undergoing a different change, fronting to /a:/ (though here the evidence is not from spelling, but from modern developments and indirect sources like rhymes: e.g the undoubtedly front /a:/ in French loans like dame rhymes with OE / a : / i n items like bame' home') The result of these two changes is — for a time — two quite different long vowel systems: (13) South i: u: e: o: e: o: North i: u: e: o: e: a: (I omit /y(:), o(0/> a s t n e y played, when present, no part in the basic developments treated here: see 2.3.4) The south, that is, lacks a long low vowel, and the north lacks / o : / Of the empty slots, low back / a : / was not refilled in any dialect until the nineteenth century, in words of ihe path, far classes (vol Ill, ch 1); but southern /a:/ and northern / o : / began to develop in the thirteenth century Before we consider their sources, however, it would be well to look at two early 'archetype' systems, southern and northern, with long and short vowels together in a spatial array The geometry of the paired systems is important: (14) Southern i: i u u: e: e o o: e: D: a Northern i: i u u: e: e o o: e: a: a In the thirteenth century, as a result of a complex, controversial and ill-understood change called Open-Syllable Lengthening (OSL), a set of new vowel qualities was created, producing considerably more symmetrical long and short systems On the traditional account, which I follow here (for others see below), short vowels lengthened in first syllables of disyllabic words with only one medial consonant; and if they were non-low (i.e except / a / ) they lowered by one height as well (Not all vowels were equally prone to lengthening in all areas; while the nonhigh ones lengthened generally, the high ones tended to lengthen later, 47 ... language - History I Hogg, Richard M II Blake, N F (Norman Francis) PE10 72. C36 19 92 420 ''.9 91- 138 81 ISBN 0- 5 21 -26 474-X (v 1) ISBN 0- 5 21 -26 475-8 (v 2) ISBN 5 21 26 475 hardback UP CONTENTS List of maps... account of Middle English for many years to come THE CAMBRIDGE HISTORY OF THE ENGLISH LANGUAGE GENERAL VOLUME EDITOR Richard M Hogg ii 10 66 -14 76 THE CAMBRIDGE HISTORY OF THE ENGLISH LANGUAGE VOLUME. .. xviii 1 15 20 21 PHONOLOGY AND MORPHOLOGY Roger Lass 2. 1 2. 2 2. 3 2. 4 2. 5 2. 6 2. 7 2. 8 2. 9 23 Introduction Phonology, origins: the Old English input system The formation of the Middle English vowel