SELIM 15.indb María José Carrillo, Selim 13 (2005-2006): 149—179 LEXICAL DIALECTAL ITEMS IN CURSOR MUNDI: CONTEXTS OF OCCURRENCE AND GEOGRAPHICAL DISTRIBUTION1 Abstract The analysis of the dialectal lexicon of Cursor Mundi shows in this paper the current research in ME word geography. The existence of several manuscripts of this work, which were copied in dialect areas different from the original one, makes it viable to establish the bases for historical lexical dialectal study. However, to be able to determine whether or not certain terms were used in a restricted regional scope, in a particular time-span I follow an approach combining the detailed study of the vocabulary of a particular work together with the analysis of the cumulative evidence from other sources. Furthermore, I show that word geography have some practical applications which might help to carry out a deeper analysis in the textual history of Cursor Mundi. Keywords. Medieval English Dialectology, diatopic methodology, Cursor Mundi. Resumen El análisis del léxico dialectal de Cursor Mundi muestra en este artículo el estado de la investigación sobre geografía del léxico del Inglés Medio. La existencia de varios manuscritos de la obra, copiados en áreas dialectales diferentes a la del texto original, posibilita, en cierta medida, el estudio histórico del léxico dialectal. Sin embargo, para determinar si ciertos términos se usaron en un área restringida, en un espacio de tiempo concreto, es necesario llevar a cabo una aproximación que combine el estudio del vocabulario en cada obra en particular junto al análisis de la evidencia acumulativa de otras fuentes. Así mismo, el artículo muestra que el estudio de la geografía del léxico del Inglés Medio tiene aplicaciones prácticas que posibilitan un análisis en profundidad de la historia textual de Cursor Mundi. Palabras clave: dialectología inglesa medieval, metodología diatópica, Cursor Mundi. 1. INTRODUCTION: WORD GEOGRAPHY AND CURSOR MUNDI Studies in the dialectal lexicon in Middle English are scarce. Rolf Kaiser’s Zur Geographie des mittelenglischen Wortschatzes (1937) is one of the few works published, so far, on the lexical dialectal material of individual Middle English 1 The present paper is part of an ongoing research project Edurne Garrido Anes and I are carrying out on the geographical distribution of lexical items regarded as dialectal. The analysis presented here has been carried out thanks to the funding granted for a research visit to the Institute of Historical Dialectology at the University of Edinburgh (May-August 2006) by the Ministerio de Educación y Ciencia (Ayudas para la Movilidad del Profesorado Universitario). I am grateful to Margaret Laing and Keith Williamson for their close readings of this paper and their suggestions. María José Carrillo 152 works.2 In this work, Kaiser listed around 500 words occurring in Cursor Mundi which he thought to have had a dialectal nature and which were, according to him, mainly northern words. Given the shortage of results in the investigation on lexical dialectal material, this list is an invaluable tool in the field. Nevertheless, his work does not reflect the possible distributions of the lexical items, since it is mainly based on the comparative analysis of the existing copies of Cursor Mundi, which cannot provide a full picture given the small number of dialects represented in the copies. A more thorough study is required to gain this objective. For some years now, my colleague Edurne Garrido Anes and I have been researching Middle English word geography. The methodology we have devised for the study of dialectal lexicon in Middle English, combines a detailed study of the vocabulary of a particular work together with the analysis of the cumulative evidence from other manuscript sources for each dialectal term.3 The reconstruction of the dialect areas where particular words were commonly used can only be achieved by putting together as much evidence as possible, and extracting it from as many different sources as we can collect. In principle, Cursor Mundi should offer a good opportunity for the study of dialectal lexicon. The existence of several manuscripts, which were copied in dialect areas different from the original one, makes it viable to establish, as Kaiser did, the bases for historical lexical dialectal study, that is to say, it is possible to establish whether or not certain terms were used in a restricted regional scope, in a particular time-span. However, the alterations and substitutions of the original vocabulary in Cursor Mundi, or of the vocabulary in any other work, may be caused by factors different from the diatopic 2 Other works that focus on the analysis of the dialectal vocabulary of individual texts are: McIntosh (1972), Hudson (1983), Black (1998), Horobin (2001) and Carrillo Linares & Garrido Anes (2007 and forthcoming, 2008). 3 For further information about the methodology and applications for the study of Middle English Word geography see Carrillo Linares & Garrido Anes (2007 and forthcoming, 2008). Lexical dialectal items in Cursor Mundi 153 variation.4 Additionally, very often the rhyming structure of a text makes dialectal substitutions difficult for a scribe. Furthermore, in this particular case, all the most southerly copies of the work derive from a common exemplar, which, as I show further on, affects the lexicon in relation to the localization of the existing copies. Analysing a single work, even if there are a good number of extant manuscripts, has many limitations. Kaiser never intended to establish the distribution of the lexical items he listed, but even so, in order to provide independent triangulation for lexical localizations, the analysis of more than just one work is required. To Kaiser’s list, there could be added some other lexical items, which also occur in Cursor Mundi, but were replaced by lexical equivalents in the southern version only occasionally. Nonetheless, our previous study reveals that the distributions of these items seem to be also restricted to certain areas.5 I try to demonstrate that these words are also dialectal and that the analysis of the copies of a single work can lead to conclusions which are not entirely accurate. Furthermore, I show that word geography may have some practical applications which might help to carry out a deeper analysis in the textual history of Cursor Mundi. 2. CURSOR MUNDI AND ITS SURVIVING MANUSCRIPT COPIES Cursor Mundi is an originally northern poem, composed sometime after 1325, which survives in several more or less complete forms and in various fragments. J. Thomson (1998) considers that there are nine extant manuscripts of the work and several other copies including short fragments inserted in other works related to Cursor Mundi in some way.6 Regarding the 4 Other causes of possible variation are the changes introduced in the texts as a consequence of the manuscript transmission itself, the purely individual scribal preferences, or the time factor. 5 See Carrillo Linares & Garrido Anes (2007 and forthcoming, 2008). 6 Hupè (1893) includes as copies of the Cursor Mundi, the fragments found in Cambridge, University Library, G.g 4.27 (2) and in London, BL, Additional 10036. Thomson (1998) points out that there are other texts related to Cursor Mundi such as: The Extract of the Book of Penance in London, BL, Cotton Galba E. IX, Oxford, Bodleian Library, Rawlinson poet. 175, and Wellesley Massachusetts, Wellesley College Library, MS 8, María José Carrillo 154 vocabulary of the work, the northern manuscripts of the poem retain most of Kaiser’s northern dialectal words.7 The non-northern versions, however, may show lexical variants for those terms that were presumably alien in the dialects of the copying scribes. There are three different groups of non-northern versions of the poem. (1) The family represented by the copy found in Oxford, Bodleian Library, Fairfax 14, copied in Lancashire. (2) That represented by folios 2r to 75r of Göttingen, University Library, Theo. 107, a linguistically composite text which was copied by one scribe from two different sources, one of them non-northern, and which shows features from South Lincolnshire. (3) The family that groups Cambridge, Trinity College, R.3.8 (henceforth T) and London, College of Arms, Arundel 57 (henceforth H), both of them associated with the Lichfield area in Staffordshire, and both copied after the turn of the fifteenth century; London, British Library, Additional 36983 (henceforth B) associated with Bedfordshire, and Oxford, Bodleian Library, Laud 416 (henceforth L), certainly non-northern, but whose dialect has not been yet analysed in depth. L is dated on f. 226v, where it reads: ‘Scriptus Rhodo per Johannem Newton die 25 octobris 1459.’ B is datable c. 1450. The analysis of some folios from the beginning of the work in L shows a highly standardized language, with only a few forms that are more restricted, or whose occurrences have not been recorded in many places with any frequency in A Linguistic Atlas of Late Medieval English (LALME). Nevertheless, the occurrences of the forms for the pronouns THEY, THEM and THEIR as ‘they’, ‘hem’ and ‘hir’ together with the forms found for the item IT, i.e. ‘yt’, ‘((it))’, ‘((hyt))’, and those for the 3rd person present indicative: ‘-yth’, ‘-yþ’, ‘-iþ’, ‘-ith’, ‘-eth’, seems to point towards the Midlands. also the texts found in Oxford, Bodleian Library, Additional A. 106 and Hopton Hall MS. In addition, Mooney (2003) states that there is also a fragment of Cursor Mundi in the Sutherland Collection (313/3633) on deposit in the National Library of Scotland. 7 The entirely northern copies are found in the manuscripts preserved in Edinburgh, Royal College of Physicians, ff. 37r-50v, 1r-15v London, BL, Cotton Vespasian A iii, ff. 2r-163r (henceforth C), Göttingen, University Library, Theo. 107, ff. 75r-169v, and London, BL, Additional 31042, ff. 3r-32v. Lexical dialectal items in Cursor Mundi 155 The palatalized initial in the forms for the item GIVE, i.e. ‘yaf’ (pret), ‘yeve’ (inf), also seem to restrict the area further, eliminating the most northern counties in the Midlands. The spellings for EACH ‘eche’, ANY ‘eny’, MUCH ‘muche’, ARE ‘ar’, ‘are’, ‘arn’, AGAINST ‘a-yen’, BEFORE ‘to-fore’ together with other spellings not recorded in the LALME sources such as ‘buysy’ for BUSY, when included in the fitting, seem to point toward Staffordshire as the possible area of provenance of the Laud MS, but still further analysis of different parts of the text is needed to attempt a more precise and accurate localization. All non-northern versions of this group were copied in the 15th century, but L and B are later copies. Thus, the language of L can easily fall into what M. Samuels (1981: 44) has called ‘a colourless regional writing’ as a consequence of the incipient standardization, so that ‘the dialectal traits that survive amount to only a small inventory of non-standard forms which even taken in combination, might belong to a number of widely separated districts.’ Previous studies show that none of the extant manuscripts of the last group, the one with which I am concerned in this paper, are in direct filial relationship with any other southern copy of the poem, but apparently they all have a common source.8 The dialect of the original source for all these copies is unknown, although Horrall (1986: 105) suggests that it could have been ‘made at Lichfield for the market in the south of England.’ Her claim is not based on any linguistic grounds, only on the supposition that it was written there because Lichfield was a centre for the translation and dissemination of northern texts for readers in the southern part of the country. It is indeed true that two of the extant manuscripts deriving from that version are localized there, and that a third one, as I have pointed out before, could belong into Staffordshire as well. The remaining copy is, however, localized in 8 Horrall (1978) argues that “HTLB clearly formed a closely related group” but “the relationships among the manuscripts of this group are not so obvious” T and H seem to derive from a common exemplar, and B and L from another, but they are all ultimately dependent of a common ancestor in which the scribe “consulted two manuscripts while preparing his translation” (Eldredge & Klinck 2000: 44-50). María José Carrillo 156 Bedfordshire. The consistency in the vocabulary in the different southern manuscripts suggests that most of the lexical changes were made by the original translator of the southern version and not by those of the extant copies. Therefore, the vocabulary is not that likely to have been selected - when copying from a northern original - in the areas where these manuscripts can be localized. These words, nevertheless, when changed for an alternative one, show their restricted use in the area of production of the original southern version. A detailed study of the vocabulary in the different extant texts could help to identify the place of provenance of this version. This is possible because, once the text was translated into a more southern dialect, the scribes for the subsequent copies were not likely to go back from replacement words to the words in the original northern version. 3. SOME DIALECTAL LEXICAL ITEMS IN CURSOR MUNDI The analysis of the distributions of some words which were not expected to have been translated in some areas could narrow down the localization of this southern original version of Cursor Mundi. This can be done by comparing the lexical material in the different manuscripts of this work with other material currently being analysed by M. J. Carrillo Linares and E. Garrido Anes, and whose distribution is being mapped. I have chosen three words not listed by Kaiser in order to illustrate two things: Firstly, that these words are also dialectal, and secondly, that the place of origin of this non- northern version could be, if not established, at least delimited further, on linguistic grounds. 3.1 MISTER: CONTEXTS OF OCCURRENCES IN CURSOR MUNDI AND DIALECTAL DISTRIBUTION. Lexical dialectal items in Cursor Mundi 157 The first word to be examined is MISTER.9 The meaning of the word in the context I am considering is that of ‘need or necessity’. There are 25 occurrences of this item in the lines of the northern versions of the poem comparable with parallel lines in the southern versions. As the poem is mainly written in rhyming couplets, 14 of these occurrences appear in line-final rhyming position. The rest of the occurrences lie in the middle of the line. Very frequently, the position of a lexical item within the text is significant when a scribe is copying it into his own dialect. The scribes had three possible strategies: 1.- They may leave the word as it is in their exemplar in order not to spoil the rhyme, if the word occurs in rhyming position. In any other position, they may also decide to retain it. 2.- They may rearrange the contents of the line so that the lexical equivalent can be put in non-rhyming position and then select a different rhyming word. 3.- They may choose to substitute the alien lexical item by a lexical equivalent. If this happens at the end of a line, it may imply a change in the rhyming word as well as the alien term. In any other position there may be no further requirements to keep the line structure. The motivations for leaving an item unaltered can be various. On the one hand, the reasons could be of a linguistic nature, since there might be metrical and/or semantic difficulties in finding an adequate substitute. On the other hand, the scribal behaviour might be psychologically motivated. The scribes’ attitudes towards their copy-texts can be different at different moments in the copying process. In a work like Cursor Mundi that takes up over 24,000 lines, the scribes might have changed their attitudes multiple times, or even without 9 According to OED the word comes from AN mester, mesteer, mestier, mestre, meister, mister, mistier, mystre, maestere, maistier and OF mester, mestier, mistier, maistier (MF mestier, F métier (1740)) need, necessity (c1140) < post-classical Latin misterium. According to Wright (1903: vol. IV) the word was dialectal and the sources in which he recorded it were mainly Scottish or from Northern England. María José Carrillo 158 changing it consciously, they might have been less focused on their own output at certain moments. MISTER in Cursor Mundi occurs in the northern texts both in the middle and at the end of the line. In the first five instances of the word, which occur between lines 803 and 3247,10 the position of the word is always final. In all the southern texts, the item is avoided in all instances in these lines. The word is not automatically replaced by a lexical equivalent, and the lines in the couplet suffer contents reorganization. The translator’s objective in these lines seems to be to produce a comprehensible poem for English speakers of non- northern areas. This objective loosens somehow as the copying process moves on, and he seems to relax slightly, especially when matters of metrics are involved. From line 3000 onwards, there are ten occurrences at the end of the line and nine of them are retained in all the southern copies.11 Only in one instance, in line 5144, are the contents of the lines reorganized and the word is replaced.12 A deeper analysis of the lexical choices in the different parts of the work would be required to determine whether this original southern version was in fact the product of more than one scribe; this would be another 10 It occurs in ll. 803, 1526, 1680, 2554, and 3247. E.g. þai cled þam þan in þat mister (l. 803) Wit leues brad bath o figer (C) þei hiled hem I telle hit þe With leues of a fige tre (T) þei hullud hem I telle hit þe With leues of a fige tre (H) 11 See ll. 3247, 4469, 5560, 10134, 11840, 14035, 18904 and 20124. 12 þou lighes now, eber pantener! Ne er þai noght o þat mister. 5144 (C) þou lyest he seide bi god so dere Ar þei no knyǟtis ny knyǟtis fere 5144 (T) þou lyest he seide bi god so dere Ar þei no knyǟtis ny knyǟtis fere 5144 (H) Lexical dialectal items in Cursor Mundi 159 possible cause for the changes in choice of translating strategy as the work moves on. In initial position, it occurs once in line 8589, and it is retained in all southern texts (eg. H: Mister wymmen were þer twynne) but B, where the corresponding reading is ‘comon’. This could be a misreading by the Bedfordshire scribe, or simply a conscious replacement, the term being alien for him. Moreover, the Bedford manuscript differs in a good number of readings from the other three manuscripts. In mid-position MISTER occurs six times from lines 4718 to 16277.13 In that context the word in the southern texts is always substituted by a lexical equivalent, not altering, in most of the cases, the wording or structure of the line or couplet. Finally, it occurs in mid-position in lines 19042 and 24810, and in these instances, the word is also retained in all the southern texts. Therefore, it seems that the linguistic environment conditions the substitutions, but apparently there could be also a psychological component in the process. Nevertheless, the fact that the item needs to be replaced in whatever context is, by itself, very revealing, and it suggests that the word was not common in the dialect of the original scribe(s) of the southern version. As the occurrences and avoidances of an item in any particular work can be conditioned by their textual histories as well, it is essential to compare the data extracted from one work with evidence from other sources. For the period between 1360 and 1460,14 MISTER is a well-attested word occurring in manuscripts of more than 30 works according to the sources for the MED. The distribution of the occurrences and avoidances in the manuscripts for the 13 See ll. 4718, 5281, 13141, 13468, 15661, 16277. 14 I have chosen this time-span because it covers the time from the possible date of composition of the original southern version of Cursor Mundi up to the actual date of composition of the latest extant manuscript of this version. María José Carrillo 160 works we have analysed so far are shown on map 1.15 According to the distribution on this map, it seems that the word occurs very frequently in texts with some northern, Lancashire, Cheshire or Lincolnshire connection, and only sporadically in texts with a more southern provenance.16 Lexical variants, other than those found in Cursor Mundi for this word, are found in different non-northern manuscripts of the Lay Folks’ Catechism, The Northern Homily Cycle, Mandeville’s Travels or The Siege of Troy, all of them originally also 15 In all the maps, the attested occurrences in precise localizations are marked with the symbol (T). The occurrences within a county which represent broad localizations, that is, not precisely to a particular place in that county, are grouped together and the number of texts where they occur are shown with a number in a black circle (e.g. n). If the number of these occurrences in a county is higher than five, the symbol used is (r+). The lexical variants to a precise localization are marked with (V), and the number of entries which represent broad localization of variants within a county is represented by a number in a circle with white background (e.g. c). 16 We have recorded occurrences in the following localized manuscripts containing ME works. The abbreviations for the localizations (between brackets) follow the LALME conventions. Avowing of Arthur Princeton, University Library, Dept. of rare Books and Special Collections, Taylor 9, (La). Benedictine Rule London, BL, Lansdowne 378, f. 19v (Yks); London, BL, Cotton Vespasian A.25, f. 109v, (WRY). The Prose Alexander Lincoln, Cathedral Library, 91, f. 25r (NME). Ywain and Gawain London, BL, Cotton Galba E.9, ff.13v, 18v, 21r (NME). Mandeville’s Travels London, BL, Egerton 1982, f. 95r (NRY); Oxford, Queen’s College, 383, f. 82r (SW Midlands); Oxford, Bodleian Library, e Mus. 116, ff. 22rb, 33vb (Cam). Northern Homily Cycle Edinburgh, Royal College of Physicians, ff. 16rb, 16va, 25vb, 35ra (NME); London, BL, Harley 4196 (NME). Patience and Cleanness London, BL, Cotton Nero A.10, f. 87v and f. 61v (Chs). The Siege of Troy London, BL, Egerton 2862, f. 125v (Sfk). York Plays London, BL, Additional 35290, ff. 22r, 156r (S Yk). Cursor Mundi London, BL, Additional 10036, f. 63va (Wrk); Göttingen, University Library, Theo. 107, ff. 23rb, 25va, 33va, 39vb, 46ra, 59vb, etc. (SE Li/WRY); Edinburgh, Royal College of Physicians, ff. 37rb, 39rb, 43vb (WRY); London, BL, Cotton Vespasian A.3 ff. 6rb, 10rb, 11ra, 13vb, 19va, 21ra, 27ra, etc. (WRY); Oxford, Bodleian Library, Fairfax 14, ff. 11v, 12v, 20v, 27v, 30r, 31r, etc. (La). Bevis of Hampton Cambridge, Gonville and Caius College, 175/96 (Li). Lay Folks’ Catechism York, Borthwick Institute of Historical Research R.I.11, f. 296v (WRY); London, BL, Harley 1022 (WRY); Cambridge, Trinity College, B.10.12 (WRY); London, BL, Additional 25006, f. 5v (NW Yrk); Oxford, Bodleian Library, Don. C.13 (WRY); Oxford, Corpus Christi College, 155 (NRY); Lincoln, Lincoln Cathedral, 91, leaf 216 (WRY/Li); Paris, Biblioteque Ste. Genevive 3390, f. 46v (Chs/W Dby); Nottingham, University Library, Middleton LM 9, f. 253r (Lei); Cambridge, Sidney Sussex College, 55 (Nth) Cambridge, University Library, Additional 6686, p. 363 (Nt). Lexical dialectal items in Cursor Mundi 161 from the Northern area or Lincolnshire. In most cases, the commonest lexical variant is NEDE.17 3.2 YERNEN: CONTEXTS OF OCCURRENCES IN CURSOR MUNDI AND DIALECTAL DISTRIBUTION The second lexical item I have chosen is YERNEN,18 a verb with the general meaning of ‘to desire’. It appears 33 times in lines of the northern texts with a parallel southern version, at the end of the line, initially or in the middle. All the southern copies are quite consistent in the selection of the vocabulary. Nevertheless, the scribe of L makes some further changes with respect to this item in places where the original southern translator left the word unaltered. In line 2592, where H reads: ‘þat myche þeraftir ǟerned I wis’ (f. 17ra), the corresponding reading in L is ‘longid’ instead of ‘ǟerned’ and in line 10858, where H reads: ‘he ǟerned not to haue no wyf’ (f. 63va), there is a variant in L for ‘ǟerned’ that reads ‘thoght.’ Even if the scribe of L leaves many of the occurrences unaltered as well, his substitutions may indicate that he had preferences for other items rather than ‘yernen’. In general, the tendency for this word up to around line 8500 in all southern copies is to change it in mid- 17 The localized manuscript copies in which we have recorded variants are the following: Mandeville’s Travels Oxford, Balliol College, 239, f. 119r (Wrk); variant found: ‘need’; Dublin, Trinity College, E.5.6, f. 34r (N Dby / S Yk); variant found: ‘need’; London, BL, Additional 33758, f. 34r (Dvn); variant found: ‘need’; Oxford, Bodleian Library, Rawl. D.100, f. 45r (Wor); variant found: ‘need’; Oxford, Bodleian Library, Rawl. D.101, f. 59r, (Brk); variant found: ‘need’; London, BL, Royal 17 B. xliii, f. 68v (Hrf); variant found: ‘need’; London, BL, Harley 2386, f. 111r (Dvn); word omitted; London, BL, Harley 3954, f. 36v (Nfk); word omitted; London, BL, Royal 17 C. xxxviii, f. 39va, (Glo?); word omitted. Northern Homily Cycle Cambridge, University Library, Dd.1.1 (Ely); variant found: ‘nede’. The Siege of Troy London, College of Arms, Arundel 22 (Dvn); variant found: ‘nede’. Lay Folks’ Catechism Oxford, Bodleian Library, Bodley 789 (Glo); variant found: ‘nede’; Lambeth Palace Library, 408, f. 13r (Borders Nfk, Ely and S Li); variant found: ‘neede’; Hopton Hall, f. 5v. (W Nfk, Ely); variant found: ‘nede’; Yale, University Library 317, f.33v (Nfk/Ely/S Li), word omitted; London, BL, Harley 6615, f. 284v (Nfk); variant found: ‘nede’; Oxford, Bodleian Library, Rawlinson C. 288 (Nfk); variant found: ‘nede’. 18 According to OED from OE, Northumb. ƥiorna, Mercian ƥeornan, WS ƥiernan, corresponding to OS girnean, gernean, ON girna), Gott gaírnjan, related to OE ƥeorn, Goth –gaírns. María José Carrillo 162 position and to retain it in final position19. From lines 10506 to 20142, there is no consistency at all. In some instances, the word is retained in mid- position as well as in final position20, where it occurs only in three instances, two of which show no lexical variant for it21. From line 21249 onwards, the word occurs in the northern versions 10 times22, and it is retained in the southern copies, regardless of the position it occupies within the line23. Once more, it seems that for this word the original southern scribe/translator took the trouble to translate a possible alien term far more consistently in the first half of his task, and then only occasionally to end up leaving all the instances as he found them in his exemplar. The geographical distribution of YERNEN, according to the sources analysed so far, is shown on map 2. Most of the works where it occurs are either northern or western in origin, and it is found in manuscripts localized in the northern counties of Yorkshire and Durham, in the western counties of Cheshire, and Worcestershire, in the eastern county of Lincolnshire, and also in the central Midlands in Derbyshire.24 Lexical variants, other than those 19 See ll. 1, 2592, 6188, 8205, 8298, 8448, where it is retained and ll. 2971, 3290, 3589, 5942, 7984, 8375 and 8399 where it is substituted. 20 See ll. 10506, 10758, 16167, 16185, 17608 and 19027. 21 See ll. 10513, 11475, 14847, 19317 and 20142. 22 See ll. 21249, 21771, 21779, 22340, 23458, 23539, 23542, 23543 and 23588. 23 There is only one instance in this part of the work where there is a slight transformation in the line contents. See l. 23680. 24 Occurrences have been recorded in: Richard Misyn, The Fire of Love Oxford, Corpus Christi College, 236, f. 33v (Li). Benedictine Rule London, BL, Cotton Vespasian A.25, f. 76r (WRY). The Parliament of the Three Ages London, BL, Additional 31042 (NME). Piers Plowman, C Version Dublin, Trinity College, 212 (D.4.1), f. 63v (NW Gl); London, University Library, Sir Louis Sterling Library V.17 (W Wor); London, BL, Additional 35157 (Copied by a NW Wor scribe from a SW Worcestershire exemplar); San Marino, Huntington Library, HM 143, f. 73r (SW Wor, but with some slight signs of interference typical of a London copying). San Marino, Huntington Library, HM 137 (Mon, Wales, bordering Gl). Lay Folks Mass Book London, BL, Royal 17.B.17, f. 7r (Dby); Cambridge, University Library, Gg.5.31, f. 2v (NME); Oxford, Corpus Christi College, 155, f. 253r (NRY). Off alle floures Oxford, Bodleian Library, Eng. poet. a.1 (Vernon), Lexical dialectal items in Cursor Mundi 163 found in Cursor Mundi, have been found in more southerly copies of texts with a northern origin, localized in Norfolk/Ely, Leicestershire, Northamptonshire and Gloucestershire.25 3.3 YEMEN: CONTEXTS OF OCCURRENCES IN CURSOR MUNDI AND DIALECTAL DISTRIBUTION The third lexical item chosen is YEMEN26 a verb with the meaning of ‘take care of, keep.’ It occurs 16 times in the northern lines that have a parallel southern version. The circumstances for this word are different from the previous one. In all southern texts YEMEN is always retained at the end of a f. 410v (Wor). Cursor Mundi Göttingen, University Library, Theo. 107, ff. 1va, 6rb, 13rb, 18vb, 26ra, 43vb, 45vb, 53vb, 55vb, 57rb, etc. (SE Li/WRY); Edinburgh, Royal College of Physicians, ff. 43vb, 50rb (WRY); London, BL, Cotton Vespasian A.3, ff. 16rb, 17rb, 18rb, 19vb, 33vb, 35rb, 36vb, 75vb, 80va, etc. (WRY); Oxford, Bodleian Library, Fairfax 14, ff. 72r, 75v, 83r, 93v, 99v, 100r, etc. (La). York Plays London, BL, Additional 35290, ff. 93r, 102r (S Yk). Lay Folks’ Catechism York, Borthwick Institute of Historical Research R.I.11, f. 295v (WRY); London, BL, Harley 1022 (WRY); Cambridge, Trinity College, B.10.12 (WRY); London, BL, Additional 25006, ff. 3v, 4r, 4v, 8v (NW Yrk); Oxford, Bodleian Library, Don. C.13 (NRY); Oxford, Corpus Christi College, 155 (NRY); Lincoln, Lincoln Cathedral, 91, leafs 215, 218 (WRY/Li); Paris, Biblioteque Ste. Genevive 3390, ff. 42v, 43v, 49r (Chs/W Dby); Oxford, Bodleian Library, Bodley 789 (Glo); Hopton Hall, ff. 3v, 4r, 6r (W Nfk, Ely); Nottingham, Nottingham University Library, Middleton LM 9, ff. 250v, 251r, 254r (Lei); London, BL, Harley 6615, f. 281v, 288r (Nfk); Oxford, Bodleian Library, Rawlinson c. 285, f. 62 (NME). 25 Lay Folks Mass Book Cambridge, Newnham College, 900.4 (West Midlands?); variant found: ‘desire’; Cambridge, Gonville and Caius College, 84/166 (Dby); variant found: ‘for-gyfnes’. Lay Folks’ Catechism Oxford, Bodleian Library, Douce 274 (Dby); variant found: ‘desires’; Oxford, Bodleian Library, Bodley 789 (Glo); variant found: ‘desire’; London, Lambeth Palace Library, 408, f.11, (Borders of Nfk, Ely and S Li); variants found: ‘coueyte’ and ‘desirys’; Yale, University Library 317, ff. 31r, 31v (Nfk/Ely/S Li); variants found: ‘asketh’ and ‘coueyte’; Nottingham, University Library, Middleton LM 9, f. 251r. (Lei); variant found: ‘coueyten’; London, BL, Harley 6615, f. 280v, 281v, 286v (Nfk); variant found: ‘coueyte’ and ‘askiþ’; Oxford, Bodleian Library, Rawlinson C. 288 (Nfk); variants found: ‘asketh’ and ‘coueyt’; Cambridge, Sidney Sussex College, 55 (Nth); variants found: ‘desire’ and ‘coveite’; Cambridge, University Library Ff.5.40, f. 118r (Nfk); variant found: ‘desireth’. 26 From OE ƥíeman = OS gômean to care for, guard, entertain (guests), OHG goumjan, goumôn (MHG goumen) to give heed to, observe, feast, ON geyma to heed, watch (Sw gďmma to keep, hide, Da gjemme to keep, guard, save), Goth gaumjan to perceive. María José Carrillo 164 line in a total number of twelve instances,27 and it is replaced four times by a lexical equivalent (KEPEN) or omitted in any other position28. Nevertheless, in line 9689 there are different readings in both L and B; L reads ‘lepe’ and B reads ‘sett’ in the context of the line ‘But for to kepe pees in londe’ (H: f. 56vb). These alternative readings suggest that the exemplars for both the scribe of L and that of B did not have the commonest replacement for that word, that is, KEPEN, or if these/this exemplar(s) had it, the word must have been palaeographically problematic. The reading in B makes more sense than that of L, so this was probably an emendation by its scribe, while L retains a probable misreading of an obscure initial . Additionally, in line 14638, the item YEMEN occurs in the context of the line that reads: ‘þar yee war yemed haf i ben’ (C: f. 80va), and in most southern copies the word is changed into ‘saue’ as in ‘Aboute to saue ǟow haue I bene’ (H: f. 86ra). It is only the reading in L that is different in this occasion, and it was probably caused by a misreading of the initial or, more likely, by an eye-skip provoked by the ‘haue’ that follows immediately after. The dialectal nature of YEMEN is, in this case, more difficult to envisage if we take into consideration the occurrences in Cursor Mundi only. These four replacements could be caused by different sorts of motivations, not related at all with the dialectal nature of the word. Nevertheless, if we compare the occurrences and avoidances of YEMEN in the Cursor Mundi with those in other works, it seems probable that when the scribe changed it, he did it because the word was alien in his repertoire. Map 3 shows the distribution of this lexical item and the places in which we have lexical variants in parallel texts. In this case, the occurrences of YEMEN are not strictly restricted to the north. The item seems to be also common in the western part of the country extending its scope as far south as 27 It occurs in ll. 2690, 7015, 8585, 9541, 11173, 12446, 17416, 17538, 19129, 19963, 22421 and 23136. There is a variant in line11173 in B, which seems to be a misreading of the original, for this text reads here ‘ǟeue’. 28 See ll. 9689, 9980, 14638 and 16894. Lexical dialectal items in Cursor Mundi 165 the counties of Worcestershire, Shropshire and Herefordshire.29 The variants found for this word appear in parallel manuscripts of works of northern or western origin such as the Lay Folks’ Catechism, The Siege of Troy or the 14th century version of the Ancrene Riwle.30 4. FURTHER LEXICAL EVIDENCE IN CURSOR MUNDI AND CONCLUSIONS 29 The texts in which we have checked occurrences are: Benedictine Rule London, BL, Landsowne 378, f. 17v, 30r (Yks); London, BL, Cotton Vespasian A.25, ff. 88v, 110v (WRY). The Gospel of Nicodemus London, Sion College, Arc.L.40.2/E.25, f. 19v (NME); London, BL, Additional 32578, f. 122r (La); London, BL, Cotton Galba E.9, f. 60 (NME); London, BL, Harley 4196, f. 208r (NME). Life of Saint Anne Minneapolis, University of Minnesota Library Z.822.N.81, f. 214r (NME). Metrical Version of the Old Testament Oxford, Bodleian Library, Selden Supra 52, f. 125r (WRY). Octovian Lincoln, Cathedral Library, 91, ff. 102va, 109ra (NME). Piers Plowman, C Version London, BL, Harley 2376 f. 85r (SE Hrf near Gl border). Cursor Mundi Göttingen, University Library, Theo. 107, ff. 19va, 49va, 59vb, 66ra, 67ra, 68vb, etc. (SE Li/NRY); Edinburgh, Royal College of Physicians, f. 1ra (WRY); London, BL, Cotton Vespasian A.3, ff. 16vb, 39va, 48ra, 53rb, 53vb, 55vb, 62ra, etc. (WRY); Oxford, Bodleian Library, Fairfax 14, ff. 17v, 39v, 48ra, 47r, 78r, 94r, 117r, 125v, 155r, 156r, etc. (La). Mandeville’s Travels London, BL, Egerton 1982, f. 51r (NRY). Wars of Alexander Oxford, Bodleian Library, Ashmole 44, f. 78r (Dur). Northern Homily Cycle Edinburgh, Royal College of Physicians, ff. 16rb, 18rb, 24ra, 29ra, 29ra, 29rb, 33va (NME); Oxford, Bodleian Library, Eng. poet. a.1 (Wor). Cleanness London, BL, Cotton Nero A.10, f. 67r (Chs). Havelok the Dane Cambridge, University Library, Additional 4407 (Nfk). The Siege of Troy London, BL, Egerton 2862, f. 114 (Sfk); London, Lincoln’s Inn, Hale 150 (Sal). York Plays London, BL, Additional 35290, f. 134r, 238v (S Yk). Lay Folks’ Catechism, York, Borthwick Institute of Historical Research R.I.11, f. 297r (WRY); London, BL, Harley 1022 (WRY); Cambridge, Trinity College, B.10.12 (WRY); London, BL, Additional 25006, f. 7v (NW Yrk); Oxford, Bodleian Library, Don. C.13 (NRY); Oxford, Corpus Christi College, 155 (NRY); Lincoln, Lincoln Cathedral, 91, leaf 217 (WRY/Li); Paris, Biblioteque Ste. Genevive 3390, f 49r (Chs/W Dby). 30 Ancrene Riwle Oxford, Bodleian Library, Eng. poet. a.1, f. 388ra (Wor); The Siege of Troy London, College of Arms, Arundel 22 (Dvn); variant found: ‘keped’. Lay Folks’ Catechism Oxford, Bodleian Library, Bodley 789 (Glo); variant found: ‘kepith fro’; London, Lambeth Palace Library, 408, f.16 (Borders of Nfk, Ely and S Li); variant found: ‘kepys fro’; Hopton Hall, f. 6r (W Nfk / Ely); variant found: ‘kepith fro’; Yale, University Library 317, f.33v, (Nfk / Ely / S Li), word omitted; Nottingham, University Library, Middleton LM 9, f. 254r (Lei); variant found: ‘kepeþ from’; London, BL, Harley 6615, f. 286v (Nfk); variant found: ‘kepiþ fro’; Oxford, Bodleian Library, Rawlinson C. 288 (Nfk); variant found: ‘kepeth fro’; Cambridge, Sidney Sussex College, 55 (Nth); variant found: ‘kepith’. María José Carrillo 166 All this evidence seems to show that these three lexical items were also dialectal, even if they were not included in Kaiser’s list. Nevertheless, the study of the distributions of three lexical items in isolation cannot provide sufficient support to us to be able to make accurate statements about the provenance of the original southern version. A thorough study of the dialectal lexicon as a whole is required for this purpose. In the present state of word geography studies in Middle English, this cannot be done. A pre-requisite to carry out a study of this sort is to map the distributions of the dialectal lexical items. Our present study is based on 40 lexical items and the distributions of the occurrences of all the items is not complete yet, since the potential occurrences of these items are scattered in more than 400 ME works in over 1000 manuscript copies. We have covered so far around 100 of these works and for some items the distributions are quite accurate. Other lexical items occurring also in Cursor Mundi that we consider also to be dialectal, and whose distributions have been mapped are: SERE, ALKIN, SAMEN and THOLEN31. Their possible distributions in the time-span selected (1360- 1460) are shown on maps 4 to 7. The distribution of the item SERE on map 4 seems to be almost entirely restricted to the most northern areas of the country, where it is well attested. We have recorded occurrences as far south as the eastern part of Cheshire and northern Derbyshire and also in south Lincolnshire. An original northern SERE is substituted by lexical equivalents such as MANY or DIVERSE in 31 The etymologies of these words according to the OED are: SERE: From ON: cp. OI sÕr, dat. of sik refl.pron. ALKIN: From OE ealra cynna [?] SAMEN: From OE. *samen, somen (with prep. æt somne) = OFris. samin, semin, to-semine, to saminen, OS. saman, at-samna, to samne (MLG. sam(m)ene, to samene), MDu. samen, te-samen (Du. tezamen), OHG. saman, zi samane (MLG. zesamene, mod.G. zusammen), ON. saman, til samans (Sw. samman, tilsamman(s, Da. sammen, tilsammen), Goth. Samana. THOLEN: From OE. þolian = OS. tholôn, tholian, OHG. dolôn, dolên (MHG. dolen, doln), ON. þola (Da. taale, Sw. tåla), Goth. þulan, f. OTeut. stem *þul-. Lexical dialectal items in Cursor Mundi 167 parallel non-northern texts of works where SERE occurs.32 Map 5 shows the distribution of ALKIN. Its occurrences are also well attested in the north, but its southern limits seem to be further southwards than those for SERE, and they spread out as well to south-west Midland areas. Variants or omissions in parallel texts have been found in copies scattered throughout many counties. The commonest variants found are ALL, ALL MANNER and EACH. Map 6 represents the distribution of the item SAMEN. Although this seems to be a predominantly northern term, there are several occurrences in some copies of Piers Plowman localized in the SW Midlands. The commonest variant for SAMEN is TOGETHER. The distribution of THOLEN is shown on map 7. It seems to occur consistently in the northern areas and in the west, and there are avoidances of this item in records from a larger number of counties. The lexical variant found for THOLEN is almost always SUFFER. By displaying together on maps the distribution for some of the items we have analysed, and that have been substituted in the southern version, we obtain a picture that provides the global areas in which the lexical equivalents would not be the first option for a scribe, and therefore, a replacement should not be expected in those areas. The remaining areas would be either those for which we have found lexical variants, or the areas for which we have no evidence. If this process cannot provide the possible area of composition of the southern version of Cursor Mundi, it could, at least, eliminate the areas of occurrences of the items that have a lexical equivalent in the Cursor Mundi manuscripts. Unfortunately, we have not yet recorded instances for occurrence or avoidance of any of the items in Staffordshire. The only evidence for this county is that extracted from the Cursor Mundi manuscripts, and this is very uncertain, as I have pointed out before. Map 8 represents the areas of coexistence of the lexical items analysed. The counties of Durham, Yorkshire, 32 Most of the records of occurrences are from texts that seem to have been originally composed in northern areas. The avoidances we have recorded here are found in the most southern copies of originally northern works. María José Carrillo 168 Nottinghamshire, Derbyshire, Cheshire, and probably Lincolnshire, in which SERE is commonly used, would not be likely to have been the original place of composition of the southern version, since the item is substituted. Other replaced terms analysed can also eliminate with more certainty some areas of Lincolnshire and Worcestershire (ALKIN); the western counties of Lancashire, Shropshire and Herefordshire can be disregarded as well (YEMEN), and likewise the county of Gloucestershire (THOLEN). The more items we add, the closer we can get and the narrower the limits are going to be. As stated earlier on, the present state of research in Middle English word geography does not allow for a much deeper study. After putting together the evidence for dialectal terms we have collected, we cannot be very precise about the place of origin of the version from which the extant southern manuscripts of Cursor Mundi derive. Horrall’s claim about Lichfield being the place of composition of the original southern version cannot be disregarded taking into consideration the linguistic data. Nevertheless, having no evidence for Staffordshire, the closest we can get is by looking at the evidence in the surrounding counties. Some of these items certainly occurred in most of the counties surrounding Staffordshire to the north, west and south. Lichfield is located in the south-eastern part of the county, and replacements and avoidances of several items have been recorded in the nearby counties of Warwickshire and Leicestershire, to the east and south of Lichfield. It is feasible that the scribe of the original southern version was copying it, or was himself from that area or further south or east. The more data we accumulate for each lexical item, the more precise we could be about the localization of the place of composition of the southern version of Cursor Mundi. The analysis carried out in these pages shows a practical application of the research on Middle English word geography. With this example, I would like to emphasize the importance of carrying on a research line on the dialectal lexical material in Middle English, which as well as being a topic worthy in its own Lexical dialectal items in Cursor Mundi 169 right, can additionally have many other more specific or practical applications of great interest for other kind of studies. Mª José Carrillo Linares University of Huelva REFERENCES Black, M. 2000: Putting Words in their Place: An approach to Middle English Word Geography. R. Bermudez-Otero, D. Denison, R. M. Hogg & C. B. McCully (Eds). Generative Theory and Corpus Studies: A Dialogue from 10 ICEHL. (pp. 455-479). Mounton de Gruyter, Berlin, New York. Carrillo Linares, M. J. & Garrido Anes, E. 2007: Middle English Lexical Distributions: Two Instances from the Lay Folks’ Catechism. G. Mazzon (ed.), Studies in Middle English forms and Meanings. (pp. 85- 100). Peter Lang, Frankfurt. Carrillo Linares, M. J. & Garrido Anes, E. Forthcoming, 2008: Middle English Word Geography: Methodology and Applications Illustrated. Proceedings of the Fourteenth International Conference on English Historical Linguistics (14 ICEHL), University of Bergamo, 21-25 August 2006. Eldredge, L. M. & Klinck, A. L. 2000: The Southern Version of Cursor Mundi. Vol. V. University of Ottawa Press, Ottawa. Fisiak, J. 2004: Some Remarks on Middle English Word Geography. Poetica 62: 1-15. Fowler, R. 1990: The Southern Version of Cursor Mundi. Vol. II. University of Ottawa Press, Ottawa. María José Carrillo 170 Horrall, Sarah M. 1978: The Southern Version of Cursor Mundi. Vol. I. University of Ottawa Press, Ottawa. Horobin, S. 2001: J. R. R. Tolkien as a Philologist: A Reconsideration of the Northernisms in Chaucer’s Reeve’s Tale. English Studies 82:2: 97-105. Horobin, S. 2004: Southern copies of the Prick of Conscience and the Study of Middle English Word Geography. Poetica 62: 89-101. Hudson, A. 1983: Observations on a Northerner’s Vocabulary. In E.G. Stanley & D. Gray (eds.). Five Hundred Years of Words and Sounds. D.S. Brewer, Cambridge. 74-83. Hupe, H. 1893: Essay on the Manuscripts and Dialect. In R. Morris Cursor Mundi. EETS 101. Trübner, London. Kaiser, R. 1937: Zur Geographie des mittelenglischen Wortchatzes. Palestra 205. Leipzig: Mayer & Müller. MED = Kurath, H. & al. 1952-2001: Middle English Dictionary. University of Michigan Press, Ann Arbor. Lewis, R. 1992: Sources and Techniques for the study of Middle English Word Geography. In M. Laing & K. Williamson. Speaking in our Tongues. (pp. 205-214). D.S. Brewer, Cambridge. McIntosh, A. 1972: Some words in the Northern Homily Collection. Neophilologische Mitteilungen lxxiii: 196-208. McIntosh, A. 1973: Word Geography in the Lexicography of Medieval English. Annals of the New York Academy of Sciences ccxi: 55-66. Reprinted in M. Laing (1989) pp. 86-97. McIntosh, A. 1978: Middle English Word Geography: Its Potential Role in the Study of the Long-Term Impact of the Scandinavian Settlements upon English. The Vikings, Proceedings of the Symposium of the Faculty of Arts of Uppsala University, June 6-9. Reprinted in M. Laing (1989) pp. 98-105. Lexical dialectal items in Cursor Mundi 171 LALME = McIntosh, A., Samuels, M. L. & Benskin, M. [with the assistance of M. Laing and K. Williamson]. 1986: Linguistic Atlas of Late Medieval English. Aberdeen University Press, Aberdeen. Morris, R. 1874, 1875, 1876, 1877, 1878, 1893: Cursor Mundi. EETS 57, 59, 62, 66, 68, 99, 101. Trübner, London. Mooney, L. 2003: A Fragment of the Cursor Mundi in the Sutherland Collection on Deposit in the National Library of Scotland. Journal of the Early Book Society 6: 143-47. Mous, Peter H. J. 1986: The Southern Version of Cursor Mundi. Vol. IV. University of Ottawa Press, Ottawa. OED = Murray, J. & alii 1989: The Oxford English Dictionary. Clarendon Press, Oxford. Samuels, M. L. 1981: Spelling and Dialect in the Late and Post-Middle English Periods. In M. Benskin & M. L. Samuels (Eds.), So Meny People, Longages and Tonges: Philological Essays in Scots and Medieval English Presented to Angus McIntosh. Middle English Dialect Project, Edinburgh. 43-54. Stauffenberg, H. J. 1985: The Southern Version of Cursor Mundi. Vol. III. University of Ottawa Press, Ottawa. Thompson, J. J. 1998: The Cursor Mundi: Poem, Texts and Contexts. The Society for the Study of Medieval Languages and Literature. Medium Aevum Monographs, Oxford. Wright, J. 1898-1905: The English Dialect Dictionary. Oxford University Press, London. María José Carrillo 172 Lexical dialectal items in Cursor Mundi 173 María José Carrillo 174 Lexical dialectal items in Cursor Mundi 175 María José Carrillo 176 Lexical dialectal items in Cursor Mundi 177 María José Carrillo 178 Lexical dialectal items in Cursor Mundi 179 * † *