iiili iiiijiiiijl !!! li i ! I iilillil i I iiiii!i!!iiniiiiiiiiiiiiniiiuiiiiiiiiiiiiiiiiiiitiiMniuijHitniiiuiiiui!iliiiiiMiiiii!!iifiiii!ji!iiiiii!ii(iiiniir 'i'"""""""""''Hinuiiiiiii!(mimi(iii(!i ii'liiiiiiiiiiHHiiiitfrmJiiiifiuiiitiiiimiiiiiiiiiiininHiiijii- JIIMi!!lini!ilinillil!!iiiliiinniiiniininHii iiiiiiiiiiiiniiiiMMiMMiii.. I,.....,.. :..h....:. liiii'i iilllli 11! I i SfC-NRLF HHIilillli ^"*»ffl"uiiiniii| ^ ^ Oflfl 7c^3 f^H^' COIUMBIA; 1893 I ijNlVERSlly £ Press K^SSS^j ii p j i, 11 , M r ' ill ' ' j Pll'^ II Ml 1 i 1 I ili*'^ M ! n s .iiiiijiiij ! i I iii Mi i lull i Ifiiilil iiiiiiniiiiii:iii iiii.,!^ ,.. ■ ■ ••'•> iitll!illti|l!ii!(!ii!iillilil!lll||(l|lil|jj|j|i||!i!ii|ii|iJ!lilll!il!n nii!)|il!itll||||jlillillli r;i!iiiiniiiiiii!'iiiitiiiiiiiiii.'!itniiitii!Mii!iijii(!(inii!iiii!;ii lUKiKiUiiiiiiiiiiiiiiiu !llUi!iJi;ii!H!tiiHIMU!(lliil(i ii i I i i I i 1 1 II i 1 n 1 1 1 1 i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 II I r . 1 1 1 ; > 1 1 1 1 : ^ LIBRARY OF THE University of California. Class *M» Columbia Sant&ersitg STUDIES IN ROMANCE PHILOLOGY AND LITERATUBE THE ANGLO-NORMAN DIALECT 'T)^^^C^^ THE ANGLO -NORMAN DIALECT A MANUAL OF ITS PHONOLOGY AND MORPHOLOGY WITH ILLUSTRATIVE SPECIMENS OF THE LITERATURE BY LOUIS EMIL MENGER, Ph.D. LATE PEOFESSOa OF ROMANCE PHILOLOGY IN BRYN MA WE COLLEGE OF THE THE COLUMBIA UNIVERSITY PRESS THE MACMILLAN COMPANY, Agents LONDON: MACMILLAN & CO., Ltd. 1904 All rights reserved COPYEIGHT, 1904, bt the macmillan company. Set up and electrotyped. Published July, 1904. Nortoooti i^tcBB J. S. Gushing & (.'o. — Berwick & Smith Co, Norwood, Mass., U.S.A. PREFATORY NOTE In view of the circumstances attending the appear- ance of this work, it has been thought appropriate to give it in some degree the form of a memorial volume. One of the author's close friends and associates has been asked to prepare a brief biographical notice, and a likeness of Dr. Monger, introduced as a frontispiece, will enhance the interest of the book to his many friends. A large part of the work had received the benefit of the author's revision in the proofs; but it remained for the editor a very special duty of love and friendship to exercise the patient vigilance requisite to seeing accurately through the final stages of the press a work so bristling with technicalities and minute details. In particular it devolved upon the editor to seek out, insert, and verify the array of cross-references called for in blank, but not indicated in the main in the author's manuscript nor in the early proofs. Owing to the peculiar nature of the presentation, it was not found to be feasible or desirable to provide an index in the ordinary sense of the word, but it is thought that the detailed analytical table of contents will serve every necessary purpose. It was not a part of Pro- fessor Monger's plan to supply a vocabulary and notes to the illustrative extracts. The exposition of these passages, so far as called for, was intended to be given by the professor in charge of the course of study. V 164116 vi PREFATORY NOTE In the belief that the amount of highly specialized and skilful labor bestowed by Dr. Menger on the task of his predilection will be found abundantly justified by the degree of its usefulness to all students of Anglo- Norman, — even to those most thoroughly versed in this broadly important field, — the book is confidently offered as the partially fulfilled promise of an under- taking and of a career prematurely cut off in the fresh- ness of youthful vigor and fruition. H. A. TODD. "V OF THE UNIVERSITY OF LOUIS EMIL MENGER. June 21», 1871 — August 4, 1903. IN MEMORIAM A SAD accident has cut off the author of the present work from activity and effort before he could see the fruit of his years of earnest study. Since 1897 he had been collecting and digesting the bibliography of the Old French dialects, and during the winter of 1902 he decided to publish a section of the work he had in mind, that he might see whether the plan of the series met with approval, and might profit in the later parts by the criticisms of the first volume. This is no longer possible ; but could the author have foreseen the untimely ending of his life, he would no doubt have asked that the book still be judged with the calm, scientific spirit that knows and seeks nothing but truth ; and its appearing as a memorial volume implies in no wise a disposition to alter in this regard what would have been his wish. The material for subsequent volumes in the Manual of Old French Dialects, projected by Dr. Menger, is limited to card references, and unfortunately is not in a form sufficiently advanced to be utilized for publica- tion. An indication of the scope which might have been given to it is furnished by an article which appeared in Modern Language Notes^ vol. XVIII, pp. 106-111. vii Viii IN MEMORIAM Louis Emil Menger was born June 29, 1871, at Clinton, Mississippi. After graduating from Missis- sippi College, and teaching two years at Vicksburg, he entered the Johns Hopkins University. In the third year of his stay he was Fellow in Romance Languages, and in June, 1893, received the degree of Doctor of Philosophy. For a year he was Instructor, and for three years Associate at the Johns Hopkins ; in 1897 he was appointed Associate Professor at Bryn Mawr College, and three years later was made Professor of Romance Philology and Italian. In December, 1900, he was married to Miss Elizabeth Buckley. During the summer of 1903 he travelled with his wife in Italy, and on August 4 was drowned while bathing in the Lago Maggiore at Ghiffa. A fuller sketch of his life with a list of the monographs and reviews he had pub- lished is given by his friend and colleague, Professor F. De Haan, in Modern Language Notes^ vol. XVIII, pp. 225-226. Dr. Menger's straightforward, manly character and his kindly disposition are familiar to all who were associated with him. He was an able teacher and a zealous investigator, but that which gave the greatest promise of a life of usefulness and success was the steadiness of development in his intelligence and knowl- edge. Thus Part I of the 3Ianual of Old French Dialects is not a measure of the best he could have done, but an indication of the far greater attainment w^iich might have been reached had not his life come to a sudden end just as he was approaching the period of full maturity and power. E. C. A. CONTENTS PAGE Prefatory Note v Frontispiece facing vii In Memoriam vii Introduction ^ . . . . xvii THE ANGLO-NORMAN DIALECT I. General Considerations 1 II. Texts 6 Philippe de Thalin 8 Cumpoz 9 Bestiaire 9 Alexis 10 Laws of William the Conqueror .... 11 Oxford Psalter 12 Cambridge Psalter 13 Arundel Psalter 14 Quatre Livres des Rois 15 Roland 16 Brandan 17 Gaimar 18 Adgar 19 Fantosme 20 Angler 21 Chardri 22 Guillaume de Berneville 23 Modwenne 24 St. Thomas 24 Adam 25 Bevis of Hampton . . . . . . .26 Amadas and Ydoine ...... 26 Chevalier, Dame et Clerc 27 ix X CONTENTS PAGE St. Auban .... . 27 Sardenai .... 28 Aspremont 29 Madeleine .... 30 Fabliau du H^ron . 30 Le Donnei des Amants . 31 P61erinage de Cliarlemague 31 Miscellaneous Poems . 32 Bible Translation . 33 Apocalypse 33 St. Paul .... 34 ( Bozon ..... . 36 Tabulation of Texts . 36 III. Phonology and Morphology . 37 PHONOLOGY 38-109 Latin A 38-50 1. Free A > E. 1. Quality of the e < a . 38 2. Orthographic variants of e < a 39-43 le, 39-40 ; ei, 40 ; oi, 41 ; ee, 41- -42 ; ae or oe, 42 ; ff, 42 ; i, 42-43 ; ai, 43 ; eo, 43. 2. A before a palatal • 43 46 1. EI and E . . . . • 43-45 2. Orthographic variants of ai, ei, E 45 46 ee, 45 ; «, 45 ; ae, 45 ; o^■, 45- -46; aeU 46; ea, 46. 3. A before L mouill^e • 46 4. A before a nasal .... • 46-48 1. Free position , 46 47 ain and ein, 46 ; ain and ame, 47 ; en, 47 ; 5. 6. an, 47. 2. Checked position aun, 47 ; ain, 48. A before L or R A pretonic 1. Loss of A in hiatus .... 2. A replaced by other letters . e for «, 49 ; ai for «, 50 ; o, au for «, 50 3. Pretonic a + palatal .... 4. Pretonic a + nasal .... 47-48 48-49 49-50 . 49 49-50 . 50 . 50 CONTENTS XI Latin E . 7. Free E 1. EI, AI, oi £ 8. 9. 10. PAGE 50-65 50-53 ••.... 50—51 52 3. Orthographic variants of e, ei, oi . . 52-53 Of e (ee, ie, i, oe), 52; ei, oi (eai, aei, ui), 52-53. Checked E . 1. Orthographic variants (ee, ei, eo) E before L mouill^e . 11. Latin 12. E before a nasal 1. Free position ain, ein, en, 53 2. Checked position E before L or R • * E . . . . . Free E . . . variants (oe^■, eie, 13. 14. 15. 16. 17. 1. IE and E . . . 2. Quality of e < ie 3. Nature of the diphthong 4. Orthographic variants of ie (e) ee, 57 ; ei, 57 ; eie, 57 ; i, 57. Checked E 57 E before a palatal 57-58 E before a nasal 58 1. Free position 58 Orthographic variants (ee, ei, i, eie, iei), 58. 2. Checked position 58 Variants (ei, eie}, 58. E (and E ) before L 58-59 1. Variants 59_60 eus, us, 59 ; als, aus, 59 ; eols, eous, ols, ous, 59-60 ; oels, ouls, euls, 60. Pretonic E (close or open) 60 1. Fall of pretonic e 60-61 Before a vowel, 60 ; before a consonant, 61. 2. Parasitic pretonic e 61-62 3. Pretonic e before a nasal 62 4. Variants of pretonic e . . . . 62-63 Phonetic (a, i, 0, u), 62-63; orthographic, (eo, ae, oe), 63. i), 54 . 53 . 53 . 53 53-54 53-54 . 54 . 54 55-65 55-57 55-56 . 56 . 56 . 57 Xll CONTENTS PAGE 18. Post-tonic E 63 1. Fall of post-tonic e 63-65 After vowels, 63-64 ; after consonants, 64. 2. Parasitic post-tonic e 64-65 3. Orthographic variants (i, u, a, oe, ae) . 65 Latin ' [ 65-69 19. I, Tonic . 65 20. I, Atonic . 65 Latin . 66 21. Tonic 66-68 1. u and o . 66 u for 0, 66 ; u for o rhyming with w< u, 66-67. 2. ou ........ 67-68 3. EU ....... • . 68 4. Orthographic variant of u . . 68 22. before a palatal . 68 1. oi and o . 68 2. ui and u . 68 23. before a nasal . 69 1. ON and UN ..... . . 69 and 11, 69 ; u « o + nasal) rhyming with ?^ < ii -1- nasal, 69. 2. GUN 69 Latin . 70 24. 0, UE, EU, U, and E 1. o 2. UE 3. EU 4. u 5. E 6. Orthographic variants 70-76 . 70 70-72 . 72 72-73 73-74 74-76 oe, 74; oi (id), 74-75; eo, 75-76; on, 76; 0, 76 ; b, 76. 25. Checked "^"^ In 1. OU . 77 2. o : o . 77 26. Locum, focum, jocum 77-78 27. before a palatal . 78 28. O before L mouill^e . . 78 29. O before a nasal . 78 30. O before N mouill^e . . 79 31. Pretonic . 79 CONTENTS Xlll Latin U 32. 1 U : and U : I 2. Orthographic variants (ui, o) 33. U (and 0) before a palatal 1. UI, u, and i 2. oi 3. UE Consonants M and N 1. 2. 3. 4. 5. B Interchange N after r M and N final M and N and the nasalization of vowels Miscellaneous Confusion of n and I, 85 ; confusion in or thography of n and gn, 85 ; loss of etymologi cal ?i, 85 ; insertion of an inorganic n, 85. PAGE 79 79-80 80 80-82 80 80-81 81 81 82-85 , 82 82-83 . 83 83-85 . 85 35. L (before a consonant) 1. L>U . 2. L lost . 3. Final l > u (or lost) 4. Miscellaneous Confusion of I and r, 87 ; inorganic Z, 88. 36. L and N mouill^es 1. Loss of palatilization 2. Orthographic variants 1. 2. 3. 4. 6. R and RR . . . R before a consonant . R + I . R after a consonant e inserted, 90 ; r lost, 90 Metathesis of r . -re > -er, 91 ; -er > -re, 91 Inorganic r 7. Interchange of l and r T, D . 39. T and D intervocalic 40. T and D final . 85-89 85-88 . 86 . 87 . 87 87-88 Z+j>r, 88 88-89 . 88 88-89 89-91 . 89 . 90 . 90 . 90 90-91 . 91 . 91 91-92 92-97 02-93 94-96 xiv CONTENTS PAQB 41. T and D miscellaneous 96-97 1. Confusion of t and d in groups .... 96 2. Final x and d 96-97 t, dd, 97; final t > A;, 97 ; final t lost, 97 ; inorganic final t, 97. C . . 98-99 1. c pronounced as k 98 2. c pronounced as ts 98 3. c pronounced as ch 98-99 G 99 V 99-100 f 1. Loss of V in groups and intervocalic . . 99-100 2. Confusion of v and f 100 W 100 J 100 S 101-109 47. S before a consonant 101-102 48. S before L and N 102-105 1. General remarks on substitute letters . , 102 2. DL 103 3. LL 103-104 4. DN 104 5. GN 104-105 49. S before T 105-107 1. Early retention of s 105 2. Fall of 8 105-107 Time, 105-106 ; stages, 106-107. 3. English words with st 107 4. -STR > -ISTR 107 5. Inorganic s + t, 107. Summary 107-108 50. S intervocalic 108 51. S final 108 1. Miscellaneous 108-109 MORPHOLOGY 110-129 Definite Article 110-111 52. 1. General 110 2. Accusative for nominative 110 3. Confusion of masculine and feminine articles . 110 4. Lu 110-111 CONTENTS XV 6. 6. Adjectives Nouns . 1. 2. 3. LUI Miscellaneous Accusative for nominative NEFS, SACS, COLPS, CtC. Miscellaneous Personal Pronouns 55. First person 1. JE and JOE . 2. MI 56. Second person . 1. Tu and vous 2. ous 57. Third person 1. Tonic instead of atonic forms 2. Confusion of forms in the singular Masculine, lu = lui^ lu = lo, 116 ; 116-117. 3. Confusion of forms in the plural is = il, 117 ; es = eZes, 117 ; il = els = eles, 117 ; les = Us, 117. Demonstrative Pronouns Possessive Pronouns 1. Feminine . 2. Miscellaneous forms Eelative Pronouns Verbs 61. Interchange of conjugations 1. -EIR>-ER . PAGE . Ill . Ill . 112 112-114 112-113 113-114 . 114 114-117 114-115 114-115 . 115 . 115 . 115 . 115 115-117 . 116 116-117 feminine, 117 eles, 117 ; 2. 3. 4. 5. 6. -RE, -IR > -ER 62. -ER > -RE -ER > -IR -lER > -IR . -OIR and -DRE Present indicative 1. First person singular -e, 120 ; -c, 120-121 2. Second person singular -z, 121 ; -is, 121. 3. Third person singular (, 121 ', st = t, 121. . 117-118 . 118 . 118 . .118 . 118 . 119 LS . . . . 119-120 . 119 . 119 . 119 . 120 . 120 . 120 . 120-122 . 120-121 IT . . . 121 V . 121 XVI CONTENTS 4. First person plural 5. Second person plural 6. Third person plural 63. Present subjunctive . 1. Forms in -ge 2. First and second person singular 3. Third person singular 64. Imperfect indicative 1. Oldest forms 2. -EIE, -DIE . 3. -EINT . 65. Imperfect subjunctive 66. Future 67. Preterite . 1. -DEDI . 2. -s preterite . 3. -ui preterite 4. -lERENT > -IRENT 5. -EUMES > -UMES 6. -RENT . 68. Past participle . 69. Miscellaneous irregular or peculiar forms avoir, connaitre, creire, dire, ester, 127 ; 128; faire, getter, plaire, pouvoir, 128 savoir, trouver, 129. TEXT ^ELECTIONS Philippe de Thatin : Cumpoz Oxford Psalter Cambridge Psalter Quatre Livres des Kois Brandan Gaimar . Adgar . Fantosme Angier . Chardri . Guillaume de Berneville Vie de St. Thomas Le Chevalier, La Dame et Le Clerc P61erinage de Charlemagne Bozon .... PAGE . 122 . 122 . 122 122-123 . 122 122-123 . 123 123-125 . 123 123-124 124-125 . 125 . 125 125-126 125-12G 126 126 126 126 126 126-127 127-129 etre, 127- ; prendre, 130-167 130-135 135-137 137-138 138-140 140-143 143-146 147-148 148-149 150-154 154-157 157-160 160-103 163-164 164-165 165-167 ^ OF THE UNIVERSITY OF INTRODUCTION I NOW offer the first part of a work, the advantages, diffi- culties, and necessary deficiencies of which students of Old French dialects will at once appreciate. We have, for the present, several grammars of Old French which are, doubt- less, as good as they can be made at this time ; additions to our knowledge of that subject must come, for the most part, from two sources : A more intimate acquaintance with the individual part that the separate Old French dialects played in the history of various phonological or morphological changes, and an observation of the bearing which modern dialects and sub-dialects have on the developments of the older language. Up to a comparatively few years ago we heard little of scientific investigation of the modern dialects. Then, almost suddenly, attention was diverted to them, and scholars, in their enthusiasm for this new study, neglected a most essential feature of work done by many minds, as was the work on the old dialects : They omitted a gather- ing of their results, the synthesis that follows analysis and that renders conclusions accessible to all. Certain it is that we have no one book to which the student may go when he wishes to learn, for example, what texts belong to a given dialect, the relative dates of originals and manuscripts, or even the exact titles of printed (and especially early) editions of the texts ; nor is there a place where he may find the characteristics of the dialects succinctly stated, or a guide to the history and results of discussion on these characteristics. A work of the nature just indicated is lacking, and in all these years no one has even promised to satisfy our needs in this regard. I therefore present now the first cast of xvii xviii MANUAL OF OLD FRENCH DIALECTS the first book of reference on Old French dialects, in the hope that it may be of help to my colleagues who are special- izing on other points, and to beginners who wish to learn something of Old French dialects but have no guide of the kind through the labyrinth of information which, up to this time, has been permitted to remain scattered. My work may be dubbed by some a mere " compilation," though scholars of intimate acquaintance with studies on dialects will, per- chance, give me credit for little additions to our knowledge of that subject, interspersed here and there in the pages that follow. It were presumptuous in me to cite as my models some very noteworthy compilations, the constant companions of workers in Old French philology. Besides, does not a certain and even a large portion of original work consist in setting old facts in a new light ? 1 intend that my compilation shall lend to each dialect as presented a definite setting; the student can derive from it definite ideas and information with regard to each dialect, its texts and peculiarities. If my work is the only one to which he may go for both general and detailed information, or else for references, on these points, it needs no apology for its existence. I have reserved any discussion of the general questions of the origin, development, or boundaries of dialects until I shall have completed my treatment of the individual dia- lects. A satisfactory resume of the points of the subject is to be found in an article by Horning.^ To my mind, " Norman," ^' Picard," " Burgundian," " Lorraine," and the like, whatever else the terms may imply as names of dialects, will always be used to indicate given points, movable if you will, around which are to be grouped certain phenomena which are marked in those regions; the fact that these dialect districts are not to be bounded as speech regions as definitely as they are, or were, politically and geographically 1 Zt. Bom. Phil XVII, 160<=-187 ; cf. Bomania, XXII, G04-607. INTRODUCTION xix does not lessen their value as points d^appui for the student. In that sense I shall use the names. The reader of this book will probably wonder at times whether the work is meant as a grammatical treatise or as a collection of references — a bibliography. In a way it is meant as both. In the grammatical part I have tried to write, at greater or less length, on the more important points, so that the student who, presumably, has not been over the ground as I have, may know at once what these points are. In the bibliographical portions I have given the essential references I have found in the monographs, journals, and books that are in the average library of insti- tutions where studies like the present are likely to be pursued. I have endeavored to select, in the first place, representative texts (and not all the texts) of the begin- ning, middle, and end of the centuries in which the given dialect literature may be said to have flourished (using the word " literature," of course, in a restricted sense) ; again, I have chosen those texts to be consulted in collec- tions like, for example, the Altfranzosische Bibliothek or the Anciens Textes Frangais, or else those published in journals, especially in the Romania and the Zeitschrift fUr Romanische Fhilologie ;^ my object has been to use those texts to which the student is most likely to have access. My book is intended, for the most part, as a guide to be in the hands of the student when he is working for himself on the points to which I refer. For example, I do not indicate the loca- tion of manuscripts further than to say they are of London or Paris, as it may be ; I make only what I consider neces- sary remarks about some of the manuscripts. Any one wish- ing to look up any especial manuscript mentioned can easily find his way by following my references. I hope that these 1 To which I shall usually refer as "i?owi. " and "Z^.," for the sake of brevity. The third most commonly cited journal, the Litera- turhlatt fur Germanische und Bomanische Fhilologie, I refer to as ''LitblV XX MANUAL OF OLD FRENCH DIALECTS and all othor references will snggest sufficient information to put the right kind of a student on the right track for investigations of his own. When, in his study of my work, such a student arrives at the stage where he can detect deficiencies in the work, I shall feel that my labor has not been in vain, I begin with the Anglo-Norman dialect as being probably the most important for the general student of Old French philology. This importance is derived from the following fact : The first half-dozen texts in our list offer us speech characteristics of an Old French dialect at a time for which manuscripts in the lle-de-France dialect are unknown. Many phenomena recorded in our early texts become, at a later date, jjart of the history of the lle-de-France speech. The study of the first appearances of such phenomena must, then, be of the greatest importance. My order of treatment for each dialect will be the fol- lowing : — I. Any necessary general remarks on the given dialect as a whole, its especial significance, its difficulties, or any question of importance raised by students of the dialect. II. Bibliography of the most important representative texts in the given dialect, and detailed information with regard to these texts. This information will be comprised under the following headings, and always in the order here indicated : — 1. General notes on the text or its author. 2. The published edition used by me. 3. Notes on the manuscripts followed in these editions. 4. Date of the original composition of the text, or date of the manuscript on which the edition is based.^ 5. References to detailed studies on the language of each text. III. Phonology and morphology of the given dialect. 1 It may sometimes be more convenient to give the date of the Ms. in connection with the remarks under § 3. THE ANGLO-NORMAN DIALECT ff ^ OFT Of / THE ANGLO-NOKMAN DIALECT I. GENEEAL CONSIDEEATIONS What is Anglo-Norman? The name indicates that Anglo- Norman is the speech carried into England by the Normans and there so modified as to need a new name to distinguish it from Norman spoken on the continent. We must not suppose, however, that we have to do with an idiom which, brought over by the Norman colony, — transplanted on new soil, so to speak, — underwent a regular course of gradual growth followed by gradual decay ; such was not the case. The one great distinguishing characteristic of Anglo-Nor- man is its irregularity. To explain. The first question of the inquiring student will be : What was the influence of French on English, and what was the influence of English on French ? Was the modification of the French language which took place on English territory due to the contact and the mutual influ- ence of the two languages? The limited extent of the influence of any such contact on the French becomes evident when we note the very restricted progress that French made in England, and the consequent lack of probability of modi- fications arising from a defective use of French by the inhab- itants of England. French never became the language of the populace ; many facts indicate that it was employed mostly by the higher, perhaps to some degree by the mid- dle classes, and there is no reason for supposing that it may not have been at least understood by people outside of those classes ; writings like those of Bozon (first half of the four- teenth century, cf . p. 35) were surely not addressed to court circles. It was, nevertheless, essentially the language of B 1 2 MANUAL OF OLD FRENCH DIALECTS the higher circles, and even among these its cultivation was not of long continuance. Toward the end of the thirteenth century we find that a knight, Gautier de Bibbysworth,^ compiled an elementary treatise for the benefit of children of the aristocracy who might wish to learn French. This fact points toward English and not French as the mother tongue of the English aristocracy of the time. Toward the latter part of the fourteenth century French was not even taught regularly in the schools, nor to children of the nobles, nor used in court. At this time begin English trans- lations of French originals, and Langland and Chaucer com- mence to write. Furthermore, though this seems but natural, French did not impress itself except in those sections of England where Frenchmen actually settled. In more than one province it remained either entirely unintelligible or else little known. In any possibility of organic development the transplanted language evidently lacked, therefore, an indispensable ele- ment of organic growth, — becoming a part of the life of the great masses of the people ; it had no new blood in- fused into it, so to speak, and was doomed to early decay. The first evidences of this decay took the form of simplifi- cation; such as, in phonology, the monophthongization of diphthongs; in morphology, the reduction of cases. Our Anglo-Norman writers had no vigorous living tongue to draw on ; they were far removed from the mother French, and they were hearing English all about them. Thus it was that they forgot many niceties of speech observed on the conti- nent, and thus it is that we find examjDles of simplifications in Anglo-Norman before we note them on the continent. (This explains, incidentally, why illustrations of develop- ments in Anglo-Norman figure so largely in our Old French grammars.) If, then, Anglo-Norman did not enjoy a regular organic development, our writers had no stable usage of the language 1 Cf . Bom. XXXII, 44. ANGLO-NORMAN DIALECT 3 in England from which to draw ; the only check on licenses they took with the idiom was, therefore, their greater or less acquaintance with French of the continent, and we must expect the usage of each writer to vary according to his knowledge of that French, and his communication with France. This personal equation, this independence of each individual author, goes far toward explaining irregularities in Anglo-Norman, and is typical of it. We need not be surprised to find a writer, even in the last days of Anglo- Norman, using French that is quite correct, like, for exam- ple, that found in the Bible translation spoken of on page 33. While it is most important to recognize the fact of variations in the language incidental to the vary- ing degree of culture of each writer, we must not forget, either, that there do exist numbers of traits common to Anglo-Norman authors and scribes as a whole ; in no case, however, do these traits represent stages of consistent devel- opment. They are but the decay, the simplification already referred to, illustrations of carelessness as to distinctions that. were being observed on the continent. That most of these simplifications should later take place in French proper makes a study of the circumstances attendant upon their history in Anglo-Norman the more interesting. Such changes on the mainland were, as a rule, consequent upon the action of some general principle, and their regular course can be traced in texts of the continent. In Anglo-Norman the alterations were, for the most part, fortuitous ; they are not necessarily general among the writers of any one epoch, and fixed time limits can hardly be assigned for them. We have still another reason for irregularities in this dia- lect, though it is rather a further illustration of the one just referred to (individual prerogative in Anglo-Norman). We need not think that the term " Anglo-Norman " refers exclu- sively to the dialect of the province of Normandy as used in England. History indicates that men from many parts of France took part in the expedition of William the Conqueror. 4 MANUAL OF OLD FRENCH DIALECTS Moreover, several phenomena may be cited in Anglo-Nor- man which, are not to be found in Norman. It seems but natural to suppose, however, that the essential basis of the original French in England was Norman, and, for that matter, Norman influence had been at work in England twenty-four years before the advent of the Conqueror ; that is, during the reign of Edward the Confessor. The prepon- derating political influence in England was that of Norman leaders, and the literary men of France most likely to be attracted to England were Norman men of letters, — friends, it may be, of the political chiefs. Any exclusive Norman influence must have waned, however, at least after the end of the twelfth century, since in 1204 the individuality of Normandy itself was merged into that of the lle-de-France ; furthermore, during the reign of Henry III (1216-1272), who married Eleanor of Provence, the court colony received additions from the southern provinces of France. We may say then that "Anglo-Norman" seems to designate more aptly the early period of the dialect, while " Anglo-French," as used by some scholars, would better apply to the latter part of it ; neither defines accurately the entire period. We have now seen that our dialect underwent no pro- gressive, organic development ; the history of the language cannot be divided into successive stages, either according to dates or according to leading lines of growth ; the dialect is, to a great extent, the product of the individual caprices of writers ; it has not even a fitting name. The only gen- eral definition of Anglo-Norman possible is that it is bad French as used in England (during the Middle Ages) ; even at this point we have to modify our characterization by saying that it was often " bad " only as regards the date of its use, since what was " bad " French in Anglo-Norman terri- tory, afterwards might become good French on the continent, where Anglo-Norman simplifications often became the rule. I have already manifested my want of confidence in the feasibility of dividing Anglo - Norman phenomena into ANGLO-NORMAN DIALECT 5 periods. Perhaps I should say something of the so-called " sub-dialects,'' supposed by some to exist within the Anglo- Norman. If what I have said with regard to the lack of any stable, organic principle of life in Anglo-Norman be true, surely little stress is to be laid upon any divisions of the language within that dialect. Indications of such divisions have been traced, at least between the north and the south, as based, for example, on the rhyming or not rhyming of u with il. It is almost an axiom that a language which is being learned by an individual, or collectively by a town, a province, or a nation, becomes modified in proportion to the difficulties of the learners in imitating the sounds of the new language they hear about them. Now if French made little headway among the masses of English-speaking people, variations in one of our texts as compared with another hardly arose from the greater or less ease with which the writer, or the people of the section from which the writer came, learned and spoke French. The average Anglo- Norman scribe represented only himself ; he availed himself of French as a literary exercise in most cases, and the peculiarities in his text reflect mainly his individual caprice. From what sources, then, can we derive any justification for supposing sufficient cohesion or crystallization in any one section of the country to dignify that section with the appellation of a ^' sub-dialect " ? ^ 1 Bibliography on the foregoing pages, and general Anglo-Norman bibliography : The above ideas are derived from many sources, — notes Jotted down in the course of much reading. I append a few refer- ences : — For a general introduction to the study of Anglo-Norman, see Vising : Etude sur le Dialecte Anglo-Normand du XIF Steele, Upsala, 1882, pp. 5-15 ; Behrens in Paul's Grundriss der Germanischen Philologie, I^, 799 ss., and in Franzosische Studien, V, heft 2 ; Meyer in the introduction to the edition of Bozon (cf. below, p. 35), pp. lii-lvii. For expressions concerning the effect of the influence of individual authors in Anglo-Norman, see Vising, ^tude, pp. 13 and 14 ; Mall, MANUAL OF OLD FRENCH DIALECTS II. TEXTS I have endeavored to arrange the texts, now to be enumer- ated, in chronological order. In doing this several methods occur to one. For example, we might adopt an arrangement according to the time in which the original was actually Cumpoz (cf. below, p. 9), pp. 39, 40, and 63; Bomania, I, 71; XII, 201 ; Zt. Mom. Phil. VI, 485 ; XXI, 575. For the term ''Anglo-French" and distinctions to be observed be- tween Norman and general French influence in Anglo-Norman, see Suchier, Fran^ais et Frovengalj p. 163 ; Grober's Grundriss, I, 572 ; Franzos. Stud. V, 2, 2 ; Bom, XXVIII, 151 ; KriL Jhrsbrcht. Fort- schritte Bom. Phil. II, 1, 248 ; Paul's Grundriss, I^, 808, Some notes on sub-dialects may be seen in Vising, £tudej pp. 11-12, and in Litblt. Germ. u. Bom. Phil. IX, 176. There are two further points to which reference should be made, detailed treatment of which is forbidden by the nature of my Manual as a whole. These points are the use of accent marks in Anglo-Nor- man manuscripts and peculiarities of Anglo-Norman versification. On the accents we have the monograph of K, Lincke : Die Accente im Oxforder und im Camhridger Psalter sowie in anderen altfranzosischen Handschriften, Erlangen, 1886. The following occasional references may be added : Zt. Bom. Phil. Ill, 161 ; X, 299 ; Bomajiia, XII, 208, 434 ; Suchier, Grammatik, p. 8 ; Plahn, Quatre Livres des Bets (full title below, p. 16), p. 5 (here are many useful references); Cloran, Dialogues of Gregory (see below, p. 21), p. 72. — For the study of Anglo-Norman versification the best starting-point is the review by P. Meyer (Bomania, XV, 144), of a dissertation on the subject by Vising. Important remarks by Koschwitz and Mus- safia are to be found in Zt. Bom. Phil. II, 339 ; III, 597. The following books are almost indispensable in a study of Anglo- Norman, and I shall take it for granted that they are in the hands of every student using my work : Suchier, Grammatik., and Vie de Seint Auban (cf . below, p. 28) ; Stiirzinger, Orthographia Gallica {Altfranz. Bibl. VIII), Heilbronn, 1884 ; Behrens (work just re- ferred to); Stimming, Boeve (cf. below, p. 26). This last volume mentioned (Boeve) comes near combining most of the others cited. I hope I may facilitate the use of it for the beginner by the headings I have adopted in the treatment of ray subject, and by the frequent references to Stimming. ANGLO-NORMAN DIALECT 7 composed. This time may be approximated in three ways : First, from a consideration of the established facts as to dates in the lives of the authors. But in the case of most of our texts, even when the name of the author is known, which is rarely the case, our biographical notices of him are most meagre. Second, from internal evidence. Such evi- dence is obtainable, for the most part, only in works dealing with historical or legendary material ; in dealing with histori- cal evidence we are too often thwarted by the inaccuracy of historical allusions, and there exist few legends out of the great mass of mediaeval material, the date of whose appear- ance or episodes can be or have been accurately stated. Third, from a study of the language. Here we have to dis- tinguish between the language of the author and that of the writer or copyist of the manuscript. (In only one Anglo- Norman text are we sure that author and scribe were one and the same person, viz., in that of Frere Angier ; cf. p. 21.) Such a distinction is possible to any important extent only in poetical compositions, where the rhymes furnish a com- paratively stable element ; and even then deductions have to be weighed most carefully (the prose text of the Lois de Guillaume is one of the exceptions here; cf. p. 11). The majority of our texts have not been studied with this espe- cial point in view. Or again, we may arrange our texts by following the dates of the manuscripts at our command. These dates are arrived at with fair accuracy by the science of paleography. This science affords the only criteria that may be applied to all our texts alike, because it follows in every case certain definite and fixed principles of discrimination. The manu- script represents, as a rule, as far as anything we have can represent it, the language used by the scribe and his con- temporaries (though even at best the scribe's mode of repre- senting the language will be modified by previous training and tradition; by the character of the original or copy which he is transcribing, and by other circumstances). In 8 MANUAL OF OLD FRENCH DIALECTS the study of individual authors, it may be of prime impor- tance to fix the date of the compositions of the author. But in an investigation such as the present, where we are concerned with dialects as a whole, we are more interested in the monuments of that dialect arranged so as to show the consecutive history of the language. I have, therefore, fol- lowed the order of the time of the manuscripts of my texts as nearly as I could. For the sake of consistency I have adopted this arrangement even when the date of the original composition is accurately known (in which cases I have indi- cated such dates as well as those of the manuscripts). As a matter of fact, the order of most of the earliest (and more important) texts would not be seriously affected by rear- ranging them according to the known or supposed dates of their original forms. We must bear in mind with regard to all of our earlier texts, those, viz., of the middle and last part of the twelfth century, that for many reasons they might almost equally well be classed as Norman, — because of the comparatively slight variations in their forms as compared with those of Norman texts of the continent ; because, in some cases, the author or copyist was born in Normandy ; because the date of the original compositions fell so shortly after that of the Norman Conquest. PHILIPPE DE THAiJN 1. Philippe is our first real Anglo-Norman author. He may be called so, however, only because his language begins to show traces of peculiarities which later became charac- teristic of Anglo-Norman (such as ie : e, o : q, and the loss of inflection). Otherwise he might be classed as Norman.^ 1 Cf. Mall, Cumpoz, pp. 19, 36, 40, 45, 08, 100 ; G. Paris, Vie St. Gilles, p. XV (full title below, p. 23) ; Grober, Grundriss, II, 1, 483, For the few facts we have of Philippe's life, cf. Wright, Biographia Britannica Literaria. Anglo-Norman period, London, 1846, p. 86 ; Walberg, Bestiaire, p. xvii. Thaiin (Taiin, Thaon, Than) is near Caen, in Normandy. Ai^TGLO-NORMAN DIALECT CUMPOZ 2. Li Cumpoz FhiUpe de Thaiin, herausgegeben von E. Mall, Strassburg, 1873. The introduction (111 pages) to this edition is one of the important early contributions to the general science of Old French philology, and remains a standard work to this day. 3. Mall used four principal manuscripts, all Anglo-Nor- man, besides notes on three others from the Vatican. These are all described at length (pp. 1-19). Mall fol- lows the S[loane] Ms. of the beginning of the thirteenth century, when supported by one of the older Mss., of which the most authoritative is the imperfect A[rundel] Ms. of the twelfth century. The L[incoln] and the C[otton] Mss. are of about the middle of the twelfth century. 4. The Cumpoz is older than the Bestiaire; this becomes evident from a comparison of speech differences and from internal evidence ; the Cumpoz was written before the end of February, 1120.^ The year generally given is 1119. 5. For the language see Mall's introduction. There is likewise a separate study (of 70 pages) by L. Fenge : Sprach- liche Untersuchung der Reime des Computus, in Stengel's Aus- gahen und Abhandlungen, LV, Marburg, 1886. This work comprises a Rimarium, a Grammar of the rhymes (pp. 18-34), and a Glossary of the Cumpoz. BESTIAIRE 1, 2. Le Bestiaire de Philippe de Thaun, E. Walberg, Paris, 1900.2 3. For the Mss. add to the information in the edition the note referred to in Romania, XXXI, 175. Walberg iCf. Mall, pp. 20, 24; Walberg, Bestiaire, V- Ixxx ; G. Paris, Lit Fr. Moij. Age, p. 246. 2Cf. the additional notes by Walberg in Zt. Bom. Phil. XXV, 697 and the careful reviews of Paris, Bom. XXIX, 589 ; Herzog, Zt. XXVI, 248, aad Tobler, Herrig's Archiv, CV, 194. OF THE UNIVERSITY OF J 10 MANUAL OF OLD FRENCH DIALECTS used as a basis the London Ms. of the second half of the twelfth century, written in the same dialect as the original. 4. The usual date assigned to this text is about 1130; Walberg thinks it before, rather than after, that year. 5. For the language see Wal berg's introduction. ALEXIS 1. This text finds a place in our present bibliography in so far as the two oldest Mss., and probably one other, were written in England.^ 2. La Vie de Saint Alexis, G. Paris et L. Pannier, Paris, 1872. This edition (together with Mall's Cumpoz, already referred to) marks an epoch in the science of Old French philology. 3. The Anglo-Norman Mss. are known as Mss. L and A. The first is so designated from the Abbey of Lambspringen to which it belonged originally. This Ms. has been edited separately four different times, first by Mueller, in 1845. Its date is the middle of the twelfth century.^ Paris, in his edition (pp. 171-176), gives a detailed collation of it. Ms. A is of the Ashburnham collection, quite inferior to L, but hardly any later. It suffered from a " corrector " of a little later date than the copyist; the "corrector" tried to im- prove A by such means as changing the assonances to rhymes ; to this end he sacrificed sense, grammar, and metre, and often rendered his original unintelligible. Paris gives variants from this Ms. at the foot of each page of his text.^ Ms. P, now in Paris, was probably written in Eng- ^ The Alexis (meaning thereby the Alexis of the Anglo-Norman Mss.) is often referred to as an Anglo-Norman text. Cf., for example, Meyer-Liibke, Oram. Lang. Bom. II, 396 ; Zt. Bom. Phil. IV, 544 ; Such. Gram. p. 5. 2 Cf. Paris ed. pp. 2, 3, 28, and 137. « Cf, the edition, pp. 2, 4, and 137. \ ANGLO-NORMAN DIALECT 11 land ; it is of the latter part of the thirteenth century. Paris gives variants from this Ms. too.' 4. The date has already been referred to as the middle of the twelfth century. 5. I know of no separate study on the language of the Anglo-Norman Mss. alone. 'O^ LAWS OF WILLIAM THE CONQUEROR 1. This text forms one of a numerous collection of laws, charts, decisions, and the like, of the time of the French occupation of England. William probably had nothing to do with the compilation, but it was attributed to him for the sake of the authority of his name, as were others to Edward the Confessor or to Henry I. 2. The text has been edited a number of times, the first edition bearing the early date of 1623 ; the last edition is that of J. E. Matzke : Lois de GiuUaume le Conquerant en Francais et en Latin. Textes et Etude Critique, avec une preface historiqne par C. Bemont, Paris, 1899,- The Latin text referred to in the title (and printed in this edition in columns parallel to the French laws) is one that is proved by Matzke to have been copied from the French (though not from the French Ms. preserved to us). 3. The single extant Ms. of Holkham (Norfolk) was writ- ten about 1230. The other Mss. (at least six) were lost, but not before being utilized in editions previous to that of Matzke. 1 Cf. p. 3. The exact text of the Mss. L, A, and P is given in Foerster und Koschwitz, Uebungsbuch, second ed. Leipzig, 1902. Paris himself gives this reference in his latest edition of the text of the Alexis, Paris, Bouillon, 1903, p. 8. Cf. Bom. XXXI, 401. 2 In connection with this edition, the detailed review of Suchier, Litblt. Germ. u. Bom. Phil. XXII, 119, must be taken into account ; also that of Paris, Bom. XXIX, 153. There is a searching study on the general subject of the Lois in Arch. Stud. Neu. Spr. u. Lit. CVI, 113-138 (Liebermann). Cf. ibid. CVII, 134. 12 MANUAL OF OLD FKENCH DIALECTS 4. The state of the language in the Holkham Ms. is much at variance with what we might expect to find, judging from the date of the Ms. In fact, the phonology is in many cases more consistently representative of earliest Anglo-Norman than is that of the Psalters or Philippe even. For this rea- son I place this text in my earliest group. Matzke supposed the composition of the Lois to have been between 1150 and 1170. Paris and Suchier (in their reviews) favor an earlier date, the reign of Henry I (1100-1135). 5. Matzke studies at length the language of the text (pp. xli-lii), comparing it with that of the oldest Anglo-Norman monuments. OXFORD PSALTER 1. This text shows some characteristic Anglo-Norman traits, though not to such an extent as does the Cambridge Psalter} 2. The first and only complete edition is that of Fran- cisque Michel : Libri Psalmorum Versio Antiqua Oallica, Oxonii, 1860, 376 pp. 3. The Ms. used as a basis by Michel was that in the Bodleian Library ; hence the name, " Oxford " Psalter. Since it is probable that the Oxford Ms. was written in the " Monasterii Montisburgi " (that is, Montebourg in Nor- mandy), the name " Montebourg " Psalter is really more appropriate than the time-honored " Oxford " Psalter.- The date of this Ms. is toward the end of the twelfth century.^ 4. As for the date of the original we may say only that the Psalter was probably translated in the course of the first half of the twelfth century, and that it is certainly 1 Cf. Zt. Bom. Phil. I, 569. Much interesting and valuable infor- mation on Anglo-Norman versions of various parts of the Bible is given by Berger in his Bible Fran<^aise au 3Ioyen Age, Paris, 1884. See Meyer's review of the same, Bom. XVII, 121, and cf. Grober, Orund- riss, II, 1, 484. 2 Cf. Zt. I, 669; Bom. IX, 626. » guch. Gram. p. 4. ANGLO-NORMAN DIALECT 13 older than the Cambridge Psalter, as is shown by certain peculiarities of its language ; for example, medial e is not dropped, I does not vocalize, the III declension nouns her, jugierre, etc., do not take s in the nominative, -abat and -EBAT endings are kept distinct, — the opposite of the state of affairs for all these points in the Cambridge Psalter} 5. On the phonology of the Oxford Psalter we have the work of F. Harseim, Vokalismus und Consonantismus im Oxforder Psalter, Romanische Studien, IV, 273-327, — a somewhat diffuse and unequally proportioned article, though easy to refer to on account of its many subdivi- sions.^ For the morphology, there is the work of J. H. Meister, Die Flexion im Oxforder Psalter, Halle, 1877, 121 pp., which is to be controlled by the lengthy review, amount- ing to a new article, of Koschwitz in Zt. Rom. Phil. II, 480- 489. A useful portion of Meister's work consists of his three pages of corrections (118-121) of Michel's text, the result of his own collation of the Oxford Ms. CAMBRIDGE PSALTER 1. The evident Anglo-Norman provenance of this text has already been referred to (p. 12). The study of speech dif- ferences within the text shows that the Psalter, as we have it, is the work of three different copyists or translators. The first part, psalms I to CXXIV, is the only one that may be attributed to the old translator of the Versio Hehraica. From CXXIV to CXXXI there is a break. The second part, CXXXI to CXLVIII, had as its translator the copyist of the first part. The third part. Canticles, etc., is from a translation of the Galilean Psalter.^ That the Oxford and 1 Zt. I, 560-570. 2 cf. Zt. Rom. Phil. IV, 464. 3 This is tlie summary of Schumann's results (cf. Litblt. V, 392), though the idea of such divisions was not original with Schumann. He gives a detailed history of the subject on the first page of his arti- cle (for title see above, § 5). 14 MANUAL OF OLD FRENCH DIALECTS Cambridge psalters are not the work of the sarae translator has been definitely proved.^ 2. The edition is again one of F. Michel: Le Livre des Psaumes, Paris, 1876.^ 3. The relations of the two Mss., the Cambridge and the Paris, are discussed by Schumann (pp. 4-6). The Ms. at Cambridge represents the labor of the copyist, Eadwin of Canterbury, and for this reason the collection is at times spoken of as " Ead win's Psalter, ^^ again as the " Canterbury Psalter^ The Paris Ms. is later than the Cambridge and often shows more correct readings ; both copyists doubtless had the original before them. 4. Our text is older than the Quatre Livres des Rois, and probably falls in the reign of Henry II (1154-1189).^ The date usually assigned is about 1160. 5. On the phonology we have W. Schumann : Volcalismus und Konsonantismus des Cambridger Psalters, Heilbronn, 1883 {Franzos. Studien, IV, 4), 69 pp. — an excellent work.'' On the morphology we have E. Fichte : Die Flexion im Cam- bridger Psalter, Halle, 1879, 96 pp. This is to be controlled by the lengthy revision of it by Schumann at the end of the latter's work just noted (pp. 51-69). ARUNDEL PSALTER 1. This Psalter may be so called from the name of the Ms. (Arundel, 230) in which it is found in the British Museum. It is the only known interlinear version of the Galilean Psalter and is peculiar in that the French words are placed above the corresponding Latin ones without any re- gard to the proper order of the French sentence. It may 1 Cf. Zl XI, 513. 2 cf. Zt. I, 568, for review. 8 Cf . Paris, Vie St. Gilles, p. xxii ; Such. Gram. p. 5 ; Zt. I, 569. 4 Cf. Eom. XVI, 608 ; Litblt. V, 392. There is a dissertation by K. Dreyer : Der Lautstand im Camb. Psalt., Greifswald, 1882, which I have not seen. ANGLO-NORMAN DIALFXT 15 represent the original form of the Gallican Psalter;^ it is truer to the Latin than is the Oxford Psalter. 2. Only the first 53 psalms of this Ms. have been pub- lished up to the present, — these by Beyer, in Zt. Pom. Phil XI, 513-534 ; XII, 1-56. 3. The Ms. spoken of above is the only one mentioned. 4. The Ms. is of the twelfth century, and that is as definite a date as we may assign for the present. 5. No study of the language has been published.' 2 QUATRE LIVRES DES ROTS 1. As long as the edition mentioned below remains the only one, we seem to be justified in placing this text in our list, though its Anglo-Norman provenance has been called into question.' 2. The edition referred to is that of Le Eoux de Lincy : Pes Quatre Licres des Pois, Paris, 1841. In consulting this text one should note the revision of it by Ollerich at the end of Schlosser's dissertation (mentioned below). This revision has the following basis : on the Ms. of the Pois are written many corrections of the main body of the text ; the corrections were used inconsistently and indiscriminately by De Lincy. Ollerich studied and divided them into four sets." The first set was made by the copyist of the Ms. The 1 Cf. Jhrsbrcht Rom. Phil. I, 375. 2 I have inserted a number of examples from this text (in my pho- nology and morphology of Anglo-Norman), some of which are very in- teresting ; great stress is not to be laid on them, however, until the Ms. shall have been made the subject of further study. To judge from the printed edition, a number of peculiarities in the Ms. may be due to a very indifferent scribe. [A complete edition of the Psalter is now being prepared.] « Cf. Zt. I, 669 ; Rom. VII, 346, XV, 641 ; Meyer-Liibke, Gram. Lang. Rom. I, 190, 196. * He calls these « « Correcturen der ersten Hand," " Correcturen der zweiten Hand," " Correcturen der jungeren Hand," and " Correcturen der modernen Hand." 16 MANUAL OF OLD FRENCH DIALECTS second may have been so too ; in any case it is old enough, to be taken into account; the others are worthless for a study of the language. 3. De Lincy's edition rests on the only Ms. known for a long time, the Mazarine, of the second half of the twelfth century, undoubtedly by an Anglo-Norman scribe ; but Berger and Me3'er have directed attention to other Mss., and the original of the text may be continental and not Anglo- Norman.^ 4. The Bois is to be placed a little later than the Cam- bridge Psalter, probabl}^ about 1170.- 5. For the phonology we have P. Schlosser : Die Laut- verhdltnisse der Q. L. R., Bonn, 1886 (dissertation) ; also the first few pages of a work concerned with the general mor- phology of the Rois : K,. Plahn, Les Quatre Livres des Reis, Gottingen, 1888 (diss.). For the verbs alone we have K. Mer- wart, Die Verhalflexion in den Q. L. R., Wien, 1880, 19 pp. Schlosser's work is based on Ollerich's corrections (referred to above) ; Merwart's has the advantage of giving the actual count of all phenomena treated. I have not seen a disserta- tion by W. Bartels : Wortstellung in den Q. L. R., Hannover, 1886. ROLAND 1. This text, like the Alexis, may be included in our bibliography ; the relatively best Ms. of it is an Anglo- Norman one. 2. The standard edition of the Ms. is that of Stengel: Das Altfranzosische Rolandslied, Heilbronn, 1878. This edition was especially valuable because it was a diplomatic one, and because accompanying it was a photographic repro- duction of the entire Ms. (published separately from the 1 Cf. Bom. XVII, 125 ; XXV, 180 ; G. Paris, La LiUerature Nor- mande avant V Annexion^ Paris, 1899, p. 30, f.-n. 3. 2 Cf. Zt. I, 509 ; Schlosser, ox. p. 4, f.-n. 1 ; Such. Beimpredigt (Bihl. Norman. I), p. xviil, 13. ANGLO-NORMAN DIALECT 17 text, however). Stengel's work on this edition is now prac- tically incorporated in his new critical edition of the Roland (Vol. I, Leipzig, 1900), the basis of which remains the Oxford Ms. 3. The Ms. was adequately described, for the first time, by Stengel in the introduction to the edition of 1878. 4. The date of the Oxford Ms. is about 1170. 5. We have no separate study on the subject of the general Anglo-Norman characteristics of the Ms. Kefer- ences for works on the language of the Eoland are given by Seelmann in his Bibliographie des Altfy-anz. MolandsUedey Heilbronn, 1888, p. 59 ss. BRAND AN 1. The Anglo-Norman version of the legend of St. Brandan is the oldest one known in a vulgar speech of the Middle Ages.^ 2. The poem was published by Suchier, in Momanische Studien, I (1875), 553-588, under the title "Brandan's Seefahrt.'' 3. This edition was based on the London Ms., Cotton Yesp., of the end of the twelfth century. It and other Mss. are commented on by Suchier in his Introduction.^ 4. Internal evidence shows that the date of the original composition was 1121.^ 5. For the study of the language we have to note sev- eral works ; for the phonology, that of Vising, Etude, pp. 67-91, and of Hammer, Zt. Rom. Phil. IX, 75-115 ; for the 1 Cf. Such. ed. pp. 553 and 555 ; Grober, Grundriss, II, 1, 479. 2 Details as to Mss. and editions are given by Vising, l^tude, pp. 18-24. Tliere is another work on the Mss. of the Brandan which I have not seen ; namely, that of M. Wien : D. Verhdltniss d. Hss. d. angloiiormann. Brandanlegende, Halle, 1886. 3 Such. ed. p. 553 ; Vising, Etude, p. 24. Mall alvs^ays spoke of the Brandan as contemporaneous with the Cumpoz (cf. pp. 80 and 82). See a note by Walberg, Bestiaire, p. Ixxxi, f.-n., and cf. Rom. XXIX, 590, f.-n. 1. c 18 MANUAL OF OLD FRENCH DIALECTS morphology that of Vising, Etude, pp. 06-100, and of Brekke, Etude sur la Flexion dans le Vouage de Saint Bran- dan, Paris, 1884, 77 pp. The last few pages of Hammer's article are devoted to morphology, and Brekke treats the phonology incidentall3\ Neither one of these two seems entirely authoritative; Hammer's effort is satisfactory in its treatment of versification, but otherwise is not so good as that of Brekke. Many of the ideas of the latter are dis- puted in the reviews, which must be taken account of.^ A resume of the marked linguistic traits of the Brandan is given by Walberg in his Bestiaire, p. Ixxxi. I have not seen a work by E. Birkenhoff : Ueber Metrum u. Beim d. altfranz. Brandan, Marburg, 1884. GAIMAR 1. " Geoffrey Oaimar was a distinguished trouvere of the reign of Stephen (1135-1154). . . . He was the first who published an Anglo-Norman version of the history of the British kings by Geoffrey of Monmouth." ^ 2. " Lestorie des Engles solum la translacion Maistre Geffrei Gaimar.'^ Edited by Hardy and Martin. Two volumes, liii + 404 and xlii + 294. London, 1888 (in Rerum Britan- nicarum Medii Aevi Scriptores). This edition has been un- favorably reviewed.^ 3. A discussion of the four Mss. of the Estorie is given by Vising, Etude, pp. 25-28. He here criticises the study of the Mss. by Kupferschmidt. 4. The date of the composition of the Estorie has been 1 Cf. Bom. XV, 628 ; Zt. IX, 158 ; Litbtl VI, 370. 2 Wright, Biogr. Brit. Lit.., pp. 151-154, speaks of Gaimar. Cf. Meyer-Liibke, Gram. I, 661 ; Litbtl. IV, 311 ; Roman. Stud. IV, 417. The portion of Gaimar's history taken from Geoffrey of Monmouth is lost. Cf. also Grober, Grtindriss, II, 472. 3 Cf. Meyer in Bom. XVIII, 314-318. In this review, as well as in Vising, ^tude, p. 25, will be found references to other editions of the Estorie, in whole or in part. ANGLO-NORMAN DIALECT 19 established, from internal evidence, as falling between 1145 and 1151.^ 5. On the language the most careful work is that of Vising, Etude, pp. 80-91, for phonology ; pp. 100-103 for morphology. Kupferschmidt, in the course of an investi- gation 2 into the relations of the lay of Havelok with Gai- mar's chronicle, gives seven pages (417-423) of remarks on Gaimar's language and versification. ADGAR 1. All we know of this poet is the little that remains of what he tells us ; and the first part of the Ms. containing his work, the part in which the mediaeval poet generally intro- duces himself, is lost. His name is Adgar, though, he says, most people call him " Willame." Wright ^ calls him " William the Trouvere." The legends he relates are not original with him, but translated from a Latin book which he says he took from the library of St. Paul's, London.* 2. The complete edition of his works is that of C. Neu- haus: Adgar's Marienlecjenden, Heilbronn, 1886 (Altfran- zosische BibUothek, IX). Adgar's legend of Theophile (1080 lines) had already been published in Zeit. Rom. Phil I, 531-540, but, apparently, rather carelessly.^ 3. The legends are preserved in a London Ms. fully described by Neuhaus in his introduction. 4. The date of this Ms. is either the end of the twelfth or beginning of the thirteenth century. Kolfs (see below), from a comparison of the language of Gaimar, Fantosme, and Adgar, judged that Adgar was nearer Gaimar, that is, about 1150. 1 Cf. Wright, o,c. p. 154 (Wright specifies between 1147 and 1151) ; Vising, ^tude, p. 3.3 ; Bom. XVIII, 314. ^ Roman. Stud. IV, 411-4-30; cf. Bom. IX, 480. 3 Biogr. Brit. Lit. p. 4(34. * On Adgar, cf. Grober, Grundriss, II, 1, 050. 5 Cf. Zt. II, 81 ; Bom. VII, 343. 20 MANUAL OF OLD FRENCH DIALECTS 5. Neuhaus makes no study of the language in his edi- tion, since that had already been done quite thoroughly by W. Rolfs in Bomanische Studien, I, 179-236.^ FANTOSME 1. Jordan Fantosme is mentioned in several places, so that we are sure of a few facts of his life; he was a master in the schools at Winchester, and was present in the north of England when that district was invaded by the Scots under William the Lion in 1173 and 1174. It was this in- vasion which he afterward described in an Anglo-Norman Chronique in verse.^ 2. The Chronique was published by F. Michel, as an ap- pendix (pp. 531-613) to Vol. Ill of the Chronique des Dues de Normandie, Paris, 1844. This was the second time that Michel published the poem, the first being in 1839 for the Surtees Society of Durham.^ The next edition was that by Howlett * (Rolls Series, 1886). A new edition by Barbier is announced in JRom. XXX, 468. 3. The two Mss. of the Chronique, one of Durham and one of Lincoln, both of the thirteenth century, are described and compared by Vising.^ 4. The date is determined, from internal evidence, to have been between 1174 and 1183.^ 5. For Fantosme's language we have to refer to Vising's Etude, pp. 91-95 for the phonology, pp. 103-104 for the morphology.^ 1 See review by Vising, Lithlt. IV, 180. 2 Wright, Biogr. Brit. Lit. p. 22L Cf. Grober, Grundriss, II, 1, 638 (in the Begister, p. 1270, this reference is incorrectly given as 636). 8 Cf. Chron. Ill, 613. 4 Not Haslitt, as cited in Litblt. XIII, 416. ^ :^tnde, pp. 34-38. 6 Cf. Vising, I^tude, pp. 41, 42 ; Lithlt. Ill, 17, IV, 31L 7 On the metre of the Chronique there is a dissertation by H. Rose in Boman. Stud. V, 2, 301-382 (cf. Lithlt. Ill, 352). OFTHi- UNIVERSi I t OF ANGLO-NORMAN DIALECT 21 ANGIER 1. In the translation from Latin into Anglo-Norman of the Life and of the Dialogues of Gregory the Great we have one of the most valuable aids to the study of our dialect ; this is because the translator gives his name, then the place and time of the manuscript, which he writes himself, and he shows care and correctness rarely to be found among Anglo-Norman scribes. We owe our knowledge of the work of this translator — a certain Frere Angier of Sainte Fri- deswide — to P. Meyer, who was the first ever to mention him. 2. Meyer edited 242 verses of the Dialogues in his Eecneil, pp. 340-343. Later, in Rom. XII, 145-208, he published the Vie de Saint Gregoire le Grand in full (2954 lines) with notes on the Ms., author, and language, and a vocabulary. Selections consisting of the Prologues of the Dialogues, a prayer to the Holy Spirit and to the Trinity are given on pages 5-30 of the Dissertation of Timothy Cloran: The Dia- logues of Gregory the Great, Strassburg, 1901.^ Cloran prom- ises a complete edition of the Dialogues, which are much longer then the Vita, since they contain 19,367 verses. 3. 4. The Ms. containing Angler's translations is in the Bibliotheque Nationale, Paris. Meyer, while not perfectly sure, considers it probable, and gives good reasons for his opinion, that we have in this Ms. Angler's autograph copy. Meyer's idea is that Angier first made a sketch (" brouillon," " minute ") of his translation of the Dialogues, then copied them in full and dated them November 29, 1212. About a year and a half later he added the Vie, which he finished April 30, 1214. There is no paleographic argument against attributing the Ms. to these years ; the writing is that of the first half of the thirteenth century. 5. Meyer gives an extended study of the phonology and 1 For corrections of details in Cloran's work see Bom. XXXI, 174 ; Mod. Lang. Notes, XVI, 241 (Sheldon). 22 MANUAL OF OLD FRENCH DIALECTS morpholog}^ of the Vie in his edition, pp. 193-201. Cloran studies the same in the Dialogues, pp. 40-53, following Meyer in the main, though giving some supplementary de- tails (cf. also p. 4). CHARDRI 1. About all we know of this poet is his name, and also that he lived in England. The three poems of his which we have are of some literary importance on account of the legends which they embody.^ The Petit Plet is particularly interesting.^ 2. John Koch, Chardry^s Josapliaz, Set Dormanz und Petit Plet, Heilbronn, 1879 (Altfranz. Bibl. I). This edition was much criticised by Suchier^ and Mussafia,'* and is hardly what might be termed a definitive edition. Koch spells the name " Chardry," which is against the authority of the Mss. 3. The Mss. are three, the oldest in the British Museum and of the first half of the thirteenth century ; this one served as the basis of Koch's text; the second Ms. is in Oxford, and of the middle or else second half of the thir- teenth century ; the third is in the Vatican and probably the work of two scribes, of the thirteenth and fourteenth centuries. The first two Mss. mentioned contain all three poems, the last only the Petit Plet. 4. Since the London Ms. was written before 1216, the original must have been composed in the course of the twelfth century.^ 5. The language is studied by Koch, pp. xxv-xl, and these pages are carefully examined by Mussafia in his review.*^ 1 In addition to Koch's study of these, there is a dissertation hj A. Reinbrecht, Die Legende von den sieben Schlafern und der Anglo- Normannische Dichter Chardri, Gottingen, 1880. Cf. Koch, Litblt. II, 290 ; Such. ih. 30:3, and Varnhagen, Zt. V, 162. One chapter of Reinbrecht's dissertation is devoted to the biography of Chardri, but offers nothing new. Cf. also Grober, Grundriss, II, 1, 643, 609. 2 Cf. Horn. IX, 171. 3 Lithlt. II, 359. * Zt. Ill, 591. 6 Cf. Koch, I.e. XL VI ; Suchier, I.e. p. 361. 6 pp. 502-597. ANGLO-NORMAN DIALECT 23 GUILLAUME UE BERNEVILLE 1. A commune of the Manche bears the name of Berneville, and a text of the thirteenth century presents the name under the Latin form Bernevilla. The family of our poet prob- ably moved from tliere to England, as did that of Philippe from Than. Certain contrarieties in the language of Guil- laume are to be reconciled only when we consider him an Anglo-Norman and not a continental Norman. His languacre is superior to that of Fantosme, for example, and he holds honorable rank beside Philippe, Gaimar, or the author of the Brandan} 2. La Vie de Saint Gilles, par Guillaume de Berneville, poeme public par G. Paris et A. Bos, Paris, 1881 (Societe des Anciens Textes Frangais).^ 3. The one Ms. of the poem is that of the Laurentian library in Florence ; it was, without any doubt, written in England, because it shows all the characteristics of Anglo- Norman writing of the thirteenth century.^ 4. The date of this Ms. is the first half of the thirteenth century.** The original is older : from evidence based on the use of the names of the three Magi (Melchior, Gaspar, and Balthasar), which were not introduced into the West till after the pretended discovery of their bodies in Milan in 1158, we cannot date the original earlier than about 1170. The study of the grammatical forms, too, shows that the Vie was written after 1150.^ 5. The language is studied at length, pp. xxvii-xxxiv ; the contrariety already referred to is considered, pp. xvii- xxi. It consists in the fact that the style and the pho- netics (with the exception of one trait, the fall of pretonic 1 Cf. the edition, pp. xv, xxi, and xxxv ; Meyer-Liibke, Gram. I, 561 ; Grober, Grundriss, II, 1, 642. 2 In the use of this text account should be taken of the important corrections of Introduction, Text, and Glossary by Mussafia, Rom. XI, 594-598. 8 Ed, p^ x.vii. * Ed. p. xiv. ^ Ed, pp, ^xv and xxvii. 24 MANUAL OF OLD FRENCH DIALECTS e in hiatus: jeu>ju) are archaic, while the declension approaches very nearly the later Anglo-Norman inflection. MODWENNE 1, 2. Suchier published 112 verses of the life of St. Modwenne in his Vie cle St. Auban, pp. 54-58. It was taken from a Latin original of the first half of the twelfth century, for the lack of a manuscript copy of which Suchier postponed the complete edition of the Anglo-Norman poem. 3. The verses occur in an Oxford Ms. 4. The date of the Ms. is the first half of the thirteenth century.^ 5. Suchier cites examples from this text in his Auhaiiy but there is no detailed study of the forms. ST, THOMAS 1. The long conflict between Thomas Becket and Henry II, and, more particularly, Becket's assassination and consequent repute as a martyr, furnished material for many " Vies " of the latter, both in Latin and in French. The oldest, and most important, French life of Thomas is that of Gamier de Pont Sainte-Maxence, composed between 1174 and 1176. Of this "Life" there are six Mss., all executed in England; it has been published twice, --by Bekker, in 1838, and by Hippeau, in 1859. A second life of Thomas is by a certain frere Benet, of probably the first quarter of the thirteenth century ,2 of which several Mss. exist. This is the biography published by F. Michel in his Chronique des Dues de Normandie, III, 461-509 (same volume as that containing the Chronique of Fantosme). 1 Cf. Auhan, pp. 4 and 33, and the general remarks in Grober, Griindriss, II, 1, 647. 2 According to Meyer, St. TJiomas, p. ii, from whom all the above information is derived. Paris (Vie St. Gilles, p. xxii) gives the date as about 1172. Cf. also Grober, Grundriss, II, 1, 645. ANGLO-NORMAN DIALECT 25 Michel used a Ms. (and a poor one) of the Bibliotheque Nationale, and gives (pp. 615-630) variants from a British Museum Ms. 2. The edition of a life of Thomas most useful for our purposes is a third one, as given in the volume of P. Meyer : Fragments cVune Vie de Saint TJiomas de Cantorberyf Paris, 1885 (Soc. Anc. Textes Fr.). 3. These fragments are from a Ms. in Courtrai consisting of four leaves ; on each side of each leaf is a miniature and a fragment (all photographed and printed in the edition). 4. The writing of the Ms. indicates that it is of a date not later than the middle of the thirteenth century. The original may be placed betvt^een 1198 and the time of the Ms., it may be the year 1220. The author is unknown.^ 5. Meyer mentions (pp. xxviii-xxxi) the characteristics of the text, dwelling particularly upon the point of the elision of e before a vowel. ADAM 1. This is a very important text in connection with the history of the Mysteres.^ 2. Karl Grass : Das Adamsspiel, Halle, 1891 {Romanische Bibliothek, VI). This edition called forth several impor- tant reviews.^ The text had already been published twice. Grass edits, as an appendix (p. 53), the " Quinze Signes '^ (360 lines), which follow directly upon the Adayn in the Ms., but really have no connection otherwise with it, and do not belong to the Anglo-Norman dialect.'* 3. 4. The single Ms., that of Tours, gave rise to a discus- sion as to whether or not it was written at two different epochs. Porster holds to this idea. He even assigns the 1 Meyer ed. pp. iil, v, and xxvii. 2 Cf. Grober, Grundriss, II, 1, 712. 3 Notably those of Tobler, Litblt. XII, 341 ; Such. G'otting. Gelehrt. Anzeig. 1891, p. 685, and Mussafia, Zeit. Oestr. Gymn. XL VI, C7. ^ Cf. Horn. XXI, 280, and Jhrsbrcht. Rom. Phil II, 1, 250. 26 MANUAL OF OLD FRENCH DIALECTS first part, that in wMcli our poem occurs, to the twelfth century.^ This idea is not universally accepted, and the middle of the thirteenth century seems a safer date. 5. Grass offers a long study (pp. 111-142) of the lan- guage of the poem, comparing it with that of the Oxford and Cambridge Psalters and the Qaatre Livres des Rois. Care must be exercised in consulting these pages, since, curiously enough, Grass apparently takes no notice of the ten pages (69-78) of Forster's corrections of his readings. BEVIS OF HAMPTON 1, 2. Der Anglonormannische Boeve de Haumtone, zum er- sten male herausgegeben von Albert Stimming, Halle, 1899 {Bibliotheca Normannica, VII). Suchier says (p. cxcv of the Introduction) that this edition is made with a " Griind- lichkeit " and " Sachkenntniss " not to be noted in any other Anglo-Norman text. 3. The Mss., fully described by Stimming, pp. iii-viii, are two, both in Paris, one of the thirteenth and one of the fourteenth century. 4. The time of the original composition, if we judge from the state of the language, must have been the first half of the thirteenth century.^ 5. The language is treated in the Introduction (pp. viii- xxxii) and again in the Appendix (pp. 171-240). The Ap- pendix is a mine of information on Anglo-Norman phonology in generaL A MAD AS AND YDOINE 1. The poem of this subject was very popular in England,^ and the primitive form of it is the Anglo-Norman redaction.^ 1 A description of this Ms, had already been given, Bom. II, 91-95. Delisle here says he does not think the Ms. is to be placed before the middle of the thirteenth century. 2 For general remarks on Boeve, of. Grober, Gnindriss, II, 1, 572. 8 Cf. Bom. XVIII, 627, f.-n. * Of. Bom. XXX, 633. ANGLO-NORMAN DIALECT 27 2. Andresen edited two fragments of the poem, the one of 140, the other of 146 lines, in Zi. Mom. Phil. XIII, 84-97. 3. These fragments he found on two pages of Ms. in tlie Gottingen library. There is another Ms. (Picard ?) in Paris, of the date 1288, edited by Hippeau : Amadas et Ydoiyie, Po^me cV Aventures, Paris, 1863. (This edition 1 have not seen.) 4. Andresen supposes his Ms. to be of the first half of the thirteenth century.^ Meyer, in his review,- questions so early a date, but goes into no detail to disprove it. 5. The editor makes a note (pp. 85-87) of the Anglo- Norman peculiarities of his text. CHEVALIER, DAME ET CLERC 1, 2. This fabliau, of 586 verses, was edited by Meyer in Rom. I, 69-87, under the title " Romanz de un Chivaler et de sa Dame e de un Clerk.^^ S, 4. The Ms. is in Cambridge and was written in England toward the middle of the thirteenth century. 5. There is no detailed study of the language. ST. AUBAN 1. The poem of this name treats of the life of the first martyr in England; in his honor t^ie abbey of St. Auban was named, and there, in the thirteenth century many Mss. were written ; our present one, along with others, has been attributed to Matthew^ Paris, a celebrity of the abbey (died 1259). The editor of the poem (Atkinson) was of this opinion, doubtless an incorrect one.^ 2. The first edition of the poem was that of E. Atkinson, Vie de Seint Auban, London, 1876. Many emendations of his text were made by reviewers.'* 1 Cf. Zt. XXI, 576. 2 jiom. XVIII, 626. 3 Cf. Suchier, Auban, pp. 2 and 3 ; Meyer, Bom. XV, 146 ; Grober, Gi'undriss, II, 1, 647. * A list of these reviews is given by Uhlemann, lioman. Stud. IV, 623. 28 MANUAL OF OLD FRENCH DIALECTS 3, 4. The Ms. is in Trinity College, Dublin, and of the middle of the thirteenth century. The original is of about the same date.^ 5. On the language of the poem we have two works. First, the very important one of Suchier : XJeher die Mat- tliaeus Paris ziigeschriebene Vie de Seint Auban, Halle, 1876. This little volume developed out of what Suchier intended to be a review of Atkinson's edition. It deals with the authorship of the Auban, Anglo-Norman versification,^ and some difficult points of Anglo-Norman phonetics. The other work is the rather prolix one of Uhlemann : Ueber die Vie de Seint Auban in Beziig auf Quelle, Lautverhciltnisse und Flexion^ in Roinan. Stud. IV, 543-626.^ SAKDENAI 1, 2. This little poem of 452 lines was published by G. Eaynaud in Rom. XI, 531-537, under the title [Le Miracle de Sain]te Marie de Sard[enai]. 3. There are four Mss. of the poem. The one from which Raynaud published his text was of Tours, the same as that containing the Mysth^e d^Adam} Raynaud thought that the original poem was Picard, though the scribe had evidently had access to a model executed in England ; later, our editor made the acquaintance of two other Mss. (of London and Oxford), which were by Anglo-Norman scribes, and in Rom. XIV, 88-93, he adds variant readings from these two Mss. Later still (Rom. XV, 354) Mej^er describes a Ms. of Cambridge containing the Miracle and expresses 1 Cf. Rom. V, 384. 2 On this point, cf. the remarks of Meyer in Bom. XV, 146. 3 Cf. the reviews in Zt. VI, 485 ; Lithlt. Ill, 15. * Cf. p. 25. Grass, p. vi, mentions the Miracle in enumerating the contents of this Ms. ; but speaks of it as inedited, although Vol. XI of the Romania was in existence nine years before the date of Grass's publication. ANGLO-NORMAN DIALECT 29 his assurance that England was the original place of the composition.^ 4. The date of the Tours Ms. is given as the middle of the thirteenth century ; ^ that of London, the thirteenth ; ^ that of Oxford the beginning of the fourteenth ; ^ the date of the Cambridge Ms. is not mentioned. 5. Raynaud makes a few remarks on the language in Rom. XI, 530, and XIV, 87. ASPREMONT 1. Many manuscripts of this important chansoyi de geste were written in England. In fact, the only complete manu- scripts of it seem to have been executed either in England or in Italy. ^ 2. The text we refer to here is that edited by Langlois in Horn. XII, 446-458 (preceded by a fragment of Otiyiel from the same Ms.). The continuation of this Aspremoyit is given by Meyer, Rom. XIX, 205-216. 3. The Ms. used by Langlois is of the Bibliotheque Nationale, though originally from the archives of Lozere. Curiously enough, the fragment edited by Meyer, and now in Clermont, formerly belonged to and was a part of the Lozere Ms.^ 4. The date of this Ms. is about the middle of the thir- teenth century.^ 1 Raynaud, Bom. XIV, 87, corrects his own mistake with regard to li eule being Picard. (He had supposed that eule was feminine.) There are, however, several phenomena to be noted in the Tours Ms. he published which are not usually associated with Anglo-Norman, and which may be with Picard. For example, we may cite viaut (volet) 308, doiipeiiple, 21, do Temple, 363. The Oxford Ms. shows volt, but neither it nor the London Ms. offers a variant for the doii. The Tours Ms. has iaiis {oculum) constantly where the London and Oxford Mss. present oilz. 2 j^^m. XI, 519. ^ jjom. XIV, 82. 4 Bom. XIX, 201 ; here a list of the Mss. is given by Meyer ; of these, five were from England. ^ Rom. XIX, 204. 6 Rom. XII, 434, and XIX, 203. [There are two foot-notes numbered "3 " on this page.] 30 MANUAL OF OLD FRENCH DL\LECTS 5. Laiiglois calls attention (p. 431) to a few peculiarities of the language and orthography, particularly the use of accent marks in the Ms. MADELEINE 1. This little fragment of seventy-eight lines has an interesting bit of literary history connected with it; it was announced by its first editor, Kauffer, to be a composition of E-ichard the Lion-hearted and to vindicate for the French the right to the title of originators of the " terzine." This and other mistakes of Kauffer were soon corrected.^ The fragment records part of a legend of St. Madeleine. 2. The text is published by Suchier (who profited by Forster's emendations of the edition of Kauffer) in Zi, Rom. Phil IV, 362-363. 3. The Ms. is of the town library of Treves. 4. The poet cannot be placed before the beginning of the thirteenth century. The writing of the Ms. seems that of the latter part of the same century. 5. Schmidt, Roman. Stud. IV, 540, notes four characteris- tics of the language: the rhyming of e : ie\ -ebat ; -abat; q : u (luis : pluis from locos, plus) ; ai : e. FABLIAU DU H^RON 1. This piece, of 172 lines, is important because it offers, in all probability, one of the few instances of fabliaux of English origin. 2. It is edited by Meyer in Rom. XXVI, 88-91: Le Fabliau du Heron ou de la Fille mal gardee. 3. This fabliaa occurs in the same IMs. as the Aspremont fragment published by Meyer (cf. p. 29), and follows directly upon it. 4. The copy of the fabliau is not, however, by the same hand that wrote the Aspremont, but by one apparently a 1 For the whole story, cf. Bom. IX, 49 L ANGLO-NORMAN DIALECT 31 little later — Meyer suggests the last years of the thirteenth century. 5. The language, too, shows some variations from that of Aspremont, and is considered by Meyer (p. 88). LE DONNEI DES AMANTS 1. This is an anonymous poem that cannot be attributed to any of our well-known Anglo-Norman poets, although the author resembles Huon de Rotelande and Chardri in several points. Indeed, the latter seems to have known the Donnei, and to have developed an idea from it in his Petit Plet. The word "donnei" is derived from Provencal domnei, which was formed from domneiar, a term meaning to " pay court to the ladies." In our poem, and generally, donnei indicates an amorous conversation.^ 2. The poem and a study of it were published by Paris in Rom. XXV, 497-541 (the text itself, of 1244 lines, occupies pp. 500-522). 3. The Ms. is from the Phillipps library, Cheltenham, and was written at the end of the thirteenth or beginning of the fourteenth century.^ 4. Paris thinks the original was written during the last years of the twelfth century.-^ 5. The poem reveals several characteristic Anglo-Norman traits, such as lenz = laenz; aver: aler; joe = ego, etc."* PilLERINAGE DE CHARLEMAGNE 1. This text is included in ray list, as are the Poland, and the Alexis, because of the Anglo-Norman Ms. of it, which, as in the case of the two monuments cited, is easily acces- 1 Eom. XXV, 522 and 634. The Donnei appears after Chardri in my list because of the difference in the dates of the Mss. of the two. 2 Cf. pp. 497-500. 3 Cf. p. 5.34 ; he is sustained in this by Vising, Jhrsbrcht. Bom. Phil. IV, 1, 297, though Grober, Zt. XXI, 575, prefers the first quarter of the thirteenth. * pp. 531-532, and cf. Vising, Jhrsbrcht. IV, 1, 298. 32 MANUAL OF OLD FRENCH DIALECTS sible to the student, since it is reproduced in the critical edition of the poem. 2. This edition is that of E. Koschwitz, Karls des Grossen Relse nach Jerusalem und Constantinopely Leipzig, 1900. Fourth ed. (Altfrayiz. Bibl. II). 3. The Anglo-Norman Ms. (of the British Museum), the only one known to contain the Pderinage, was lost in 1879. Before this date, however, Koch had photographed it, and Wlilcker and Nicoi had collated Michel's edition (the first) of the poem (1836) founded on this Ms. Koschwitz prints Koch's facsimile in its entirety opposite the critical text, availing himself of variants from the two collations re- ferred to. 4. The Ms., a faulty one, by a scribe but ill acquainted with French, belongs to the end of the thirteenth or the beginning of the fourteenth century. 5. The language of this text was made the subject of an especial study by Koschwitz in his book, Ueberlieferung und Sprache der Chaiison du Voyage de Charlemagne, Heilbronn, 1876. Here we may find many points of interest for the student of Anglo-Norman. MISCELLANEOUS POEMS 1, 2. I refer here to the scattered pieces published by Meyer, Rom. IV, 370-397, with the title : Melanges de Poesie Anglo-Normande. 3, 4. The names of the pieces, Mss., and dates are the following : — a (p. 370). Missus Gabriel; British Museum Ms. of end of thirteenth century. Here we may include the " Chanson a hoire " from the same Ms., published by Paris, Rom. XXI, 260. It is worth noting that this was the Ms. (now lost) containing the PUerinage (cf. above). h (p. 373). Pri^re; Trinity College Ms., of the end of the thirteenth century. ANGLO-NORMAN DIALECT 33 c (p. 374). Chanson; Bodleian Library, of the second half of the thirteenth century. d (p. 385). Etats du Monde; Cambridge Ms., middle of the thirteenth century. This is a most interesting little poem from several points of view. Meyer devotes ten pages to a consideration of it. e (p. 395). Plainte; British Museum, end of thirteenth century. Of the fourteenth century we have : — / (p. 375). Three Chansons from a Cambridge Ms. g (p. 383). Definition de V Amour from a Bodleian Ms. In addition, several pieces are printed whose dates are not indicated (pp. 372, 380, and 384). BIBLE TRANSLATION 1, 2. A fragment of a translation into verse is given by Bonnardot, Bom. XVI, 177-213. This fragment consists of 1013 verses. 3, 4. The editor used a Ms. of Treves, appending variants from two other Mss. of the Bibliotheque Nationale, all three being of the fourteenth century, and all written in England.^ Meyer, in a note at the end of Bonnardot's article (p. 212), mentions two other Mss. (of Oxford and Cheltenham) con- taining the translation. APOCALYPSE 1. This text is important in several ways: similar versions of the Apocalypse were very popular ; of the present one in verse there are seven Mss. ; of that in prose ^ we know of six- teen; again, it represents a change in taste of the public of the time (end of thirteenth century) in so far as this public no longer cared for imaginative works in themselves, but rather 1 The language, however, is of a purity that indicates a much earlier date, or else that the copyist had before him an origmal from the continent. 2 Published by Meyer and Delisle : r Apocalypse en Fran E. 1. Quality of the E < A. Here we enter upon one of the fundamental questions of Old French philology. Early in the history of our subject it was seen that we have to deal with three different e's : e < a, e < e, and e < e or i. The distinction among them is undeniable, but as to the exact value of each of the three no definite result has been at- tained. In brief, e < a assonances only with itself (or with an e < E which has not diphthongized in free position, as deiis, eret), so that this e < a is neither open nor close in the sense that 6 < e or e was close or open.^ In Anglo-Norman this e < A (and here we have to in- clude the e of the reduced diphthong^ ie, cf. p. 55) is to be considered close, for the earlier monuments at least. The bibliography on this point is quite one-sided.^ The ulti- mate confusion in Anglo-Norman rhyme of close e < A with an open e rejflects the similar condition of affairs on the 1 A general idea of the points of the question may be had from consulting the following references : Paris, Bom. IV, 499, VII, 122 ; Such. Zt. Ill, 137 ; Meyer-Lubke, Gram. I, 213, § 225 ; Nyrop, Gram. I, 153 ; Such. Gram. p. 24. 2 We note the first definite statements for Anglo-Norman by Suchier, Zt. II, 293, III, 140. Cf. Zt. Ill, 593 (Mussafia) ; Uhlemann, in his work on Auban, Boman. Stud. IV, 563, questions the close e ; he suggests that its value may have varied at different periods of Anglo- Norman, or according to accented or unaccented position. Suchier, lAthlt. Ill, 15, disproves this, and adheres to his idea that e was close in every position up to the end of the twelfth century. Cf. Vising, £!tude, p. 68. 38 PHONOLOGY 39 Continent. The tendency to confusion was especially, and maybe first, manifested in the case of e + ? or r} 2. Orthographic Variants of E < A. The two most important and frequently recurring variants are ie and ei, both found in a number of the earliest texts and appearing, with greater or less frequency, throughout the entire Anglo- Norman period. They are to be noted, too, because it has been claimed that they represented, in some cases at least, real phonetic values. a. le. A few examples for this are the following : CmniJoz, diers (cLARUs) 300G, jyiert (paret) 2509 (both Ms. L) ; Camb. Psalt. rememhriere (LXXIII, 18, 22), aUtiere (XXI, 3) and the like.^ Instances in the Camb. Psalt. seem confined to derivates of -ator. Chardri, Petit Plet, tiel, 216, 451 ; Boeve, tiel, 3564, 3830 (Ms. D). Full lists of examples are given by Sttirzinger ^ and Stimming.* The first explanation we note here is that of MalP to the effect that since Continental ie early lost its value as a diph- thong in Anglo-Norman (cf. below, p. 55), the scribes did not have a definite knowledge as to the proper use of the ie; in their efforts to write a correct French, they replaced their Anglo-Norman e by ie, but gave no thought to the ety- mological background of their e; consequently we find in Anglo-Norman this ie representing not only the regular de- velopment of Popular Latin open e, but also substituted for e < A and e < e. Schumann (p. 14), with an eye to the class of words (<-ator) peculiar to his text {Camb. Psalt.), thought that words of the type of remembriere were analo- 1 Cf. p. 48, " a + Z or r " for references and examples. Cf. also Meyer-Lubke, Gram. I, 214, § 226. 2 Schumann, p. 14. (For the full titles of works referred to, as here, by a mere name, I expect the student to refer to my text bibli- ography above. In this way he will become familiar with the monu- ments and works and workers on them. For example, here he will look under Camb. Psalt. p, 14, § 5, studies on the language.) 3 Orth. Gall. pp. 39-40. * Boeve, p. 176. ^ Cumpoz, p. 69. 40 MANUAL OF OLD FRENCH DIALECTS gous to others like jugierre, in which ie was a regular phonetic development. The explanation of Mall seems more in accord with what we should expect of the average Anglo-Norman scribe.^ The suggestion that the ie might represent a phonetic change was Suchier's ; ^ he based it on English friar < frdre (fratrem). Stimming, however,^ claims that friar, briar, and the like originated in South England, where close e regularly becomes i and el, and that the a was a glide sound between these and the r. On the relative ages of ie and ee, see below, p. 41. b. ei. Examples for ei are very numerous, beginning with the earliest texts ; for instance, Oxf. Psalt} seit (sapit) LXXII, 11, LXXXVIII, 15, seis, LXVIII, 4 ; Camb. FsalL' espeie, XLIII, 3, 6, incurveie, XXXIV, 15; Gaimar, leveiz, 1383 ; Dialog. Greg, aleir, remaneir, acheveir, and the like.^ Verb forms, like those just cited from Angier and Gaimar, swelled the number of examples of this phenomenon; since -eir verbs so frequently become -er in Anglo-Norman (cf. p. 119), scribes were confused as to the correct usage of -eir and -er. In examples not drawn from verb forms we have to do with an uncertainty of the scribes, like that in the case of ie; that is, since ei (< close Popular Latin e) was re- duced to e in Anglo-Norman (cf. p. 52), the scribes, in their efforts to restore ei, substituted it for every e, whatever the origin of e. Suchier '^ considered this ei as a graphic variant, but here again Stimming^ disagrees with him, saying there is evi- dence that ei was a diphthong.^ 1 Cf. Meyer-Lubke, Oram. I, 173, § 179 ; Stiirzinger, OrtJi. Gall. pp. 38-39, 41 (here we find a bibliography for the history of the ie). 2 Gram. p. 23. s jjoeve, pp. 176, 181. ^ Harseim, p. 277. ^ Schumann, p. 15. 6 Cloran, p. 40. ' Gram. p. 24. ^ Boeve, p. 175. ^ Rhymes from Dialog. Greg, may indicate the same. Cf . Clorau, p. 40. PHONOLOGY 41 c. oi. This orthography may be counted as a variant of the ei just treated; it is used for words in which ei replaced e < A, and is an extension of the oi for ei < e (cf. p. 50). The only examples noted are for the Dialog. Greg} d. ee. A few examples here are Campoz, ^wert (Ms. C, 2519) ; Bestiaire, seet, eet, p/e?i/ee (Walberg, LXXXIV) ; Gaimar, leez, 665-, Chardri, Pet. Plet, seet (Ms. V, 172).^ Here again the first explanation we note is that of Mall/ who speaks of the use of ee or ee as a variation of ie ; that is, the scribe desired to indicate a sound other than that of e, but did not want to use ie. This remark is quoted by Suchier * in refutation of Uhlemann, who, in upholding his claim for an open pronunciation of e < a, cited the orthog- raphy ee (of the Auhan) as indicating an open e. Mall thought of it as pointing to close, lengthened e. Sttirzinger ^ makes some interesting remarks upon the relative ages of ee and ie, though what he says may appear somewhat contra- dictory. In his discussion of ee < free open e, he says that ee was not a stage preceding ie, but was so used to denote the length of the vowel {i.e. of e which Anglo-Norman used for ie), after the fashion of English orthography of the fourteenth century. Then he says that ee = e = a fell in phonetically with ee = ie = e, and by false analogy to the latter, the former ee was likewise written ie. Thus here (in the case of ee = e = a) ee must have preceded ie ; in the other case (ee = e), ee must have followed ie. If, however, ee was fashioned after English orthography of the fourteenth century to indicate the length of e in one instance, it seems reasonable to suppose that it was used at this time to denote the length of all e's, regardless of their provenance. If, again, ee was thus late in its diffusion, it probably had little to do with the origin of fe = e = a, a better explanation of 1 Cloran, p. 40. 2 Texts and examples, Stimming, Boeve, p. 175 ; Stiirzinger, Orth. Gall pp. 40^1. 8 Ciimpoz, p. 60. 4 Litblt. Ill, 16 (cf. Zt. Ill, 477). ^ Orth. Gall. p. 40. 42 MANUAL OF OLD FRENCH DIALECTS which has already been suggested (p. 39). Difficulty will be experienced in establishing comparative dates for such phenomena, since we have to base our arguments on the usage of scribes who so easily confused spellings. The simplest statement to be made is doubtless this : ee was used to indicate close e; this close e was that < a and that < ie < k ; the ee was independent of ie < e (and, of course, of ie as used for e < a) ; there is no evidence point- ing to any progression in the phonetic development of the sounds denoted by ee and ie, or in the use of these signs by the scribes. e. ae or oe. Examples for ae have been noted in Auban : saet (sapit),^ V^ie Greg. : aeve (acqua) ; ^ Camb. Fsalt. paerre, XXXVIII, 14. Uhlemann^ adduced ae as a further proof of the open value he wished to assign to the e < a. Suchier^ again refuses to accept this idea, saying that ae is only an ortho- graphical ornament used as an alternative for e, just as y is used for i (cf. p. 65) — both without any especial phonetic value. He reverts to this again in his Grammatik,'^ refer- ring there to oe (as well as ae) as an alternative for e. Meyer,^ however, seems to understand ae as indicating the open sound of e in his text. (See further ae = ai under a -f- palatal, p. 45, and ae = e < ei < e, p. 52.) f. a. Examples for this are noted by Suchier ; ^ as, Camb. Psalt. avortad; Quatre Liv. R.: strae (for estree); he says such words are either scribal errors, or else to be ex- plained each by itself. Stimming^ notes estat in Boeve. g. i. With the exception of til = tel, Denis Pyramus, Vie St. Edmond, 654,^ examples seem to be confined to the pret- 1 Uhlemann, p. 5G2. 2 jj^^n. XII, 194. 8 LUhlt. Ill, 16. * p. 19. 6 Gram. p. 23. « Boeve, p. 176. 7 My examples from this text are drawn from the edition of Arnold in Memorials of St. Edmond's Abbey, London, 1892. Vol. II, pp. 137 ss. [A new edition of the Vie is now being prepared.] PHONOLOGY 43 erite and past participle 7'emis or mist (rtmasus, masit), obviously analogical. Stimming gives texts and references.^ h. ai. ai really does not belong here, as it is not a vari- ant of e in orthography, but of an a which has remained, for whatever reason. It has nothing in common with the a noted above (under f). It seems to be a peculiarity of four- teenth-century texts, though in atonic position it occurs at a much earlier date (cf. p. 50). Stimming^ gives examples like lerrai for lerra, ai for a (habet). i. eo. eo for close e corresponds to ea for open e. It is comparatively late ; cf . below, p. 57, § 13. 2. A BEFORE A PALATAL. The history of the developments here is interesting be- cause some of them are limited to Anglo-Norman territory and because several of the most important are to be observed in Anglo-Norman before they make their appearance in continental French. We find in our texts that ai has the value of qi, or else that of ^ in pronunciation and that ei and e occur with great frequency in orthography, though not to the exclusion of ai which is found particularly in the older prose texts, as the Psalters and Lois Quilkiume, and doubt- less there represents the original pronunciation (di). Anglo- Norman poets, however, without regard to the value of ai as di, ei, or e (cf. below), used it, at times, in rhyme with i (pais : dis, Denis Pyramus). 1. EI AND E. We note the former in Bestiaire,^ Lois GuiUaume, Oxf. Psalt., Quatre Liv. R., Branclan, Gaimar, Angier, and Boeve ; the latter in Cumpoz, Bestiaire,^ Domes- day Book,^ Lois Guillaiime, Camb. Psalt., Q.L.B., Brandan, Gaimar, Angier, Chardri, Gme. de Berneville, Boeve, Sar- denai (leisse : confesse, 241), Donnei (p. 532). Some of the discussion here is worthy of attention. Mall ^ 1 Boeve, p. 176 ; cf. below, § m, 2. 2 Ibid. p. 172. 8 Walberg, pp. xlviii, Ixxxv, xUv. * Zt. VIII, 358. 6 Cumpoz, p. 59. 44 MANUAL OF OLD FRENCH DIALECTS mentions tliree cases from liis text and four from the Bes- tiaire in which a + a palatal is represented by e, all the ex- amples showing ai-\-str','^ he says the e shows that the character of the ai as a diphthong was being modified at the time of these texts and was approaching the value of q. Six years after Mall's statement had been made, G-rober - seems to wish to combat a tendency then prevalent to place the general reduction of ai at too early a date. He says the monophthongization took place in the works of Philippe de Thaiin on account of the phonetically difficult (and long) consonant group following the ai (maistre, paistre) and not from a general tendency in that direction. The nature and date of the development attracted much attention about the time of Grober's note,^ and after this time we observe that editors and grammarians study the ai with reference to its position before consonant, or vowel, or consonant group, or final, and the like. For example, Paris establishes the fol- lowing for Gme. de Berne ville : ^ ai is distinct from ei, both final and before a consonant. Ait :^eit only once (3307) nor does ai7i : ein ; but mais : pres, lairme : terme, fairs : terre. The value of the latter ai is ^. If we draw up a table formu- lated from Pariy's remarks, we shall have the following: — e < ton. check. ^ : ai-{- Cons. e < ton. check. ^ does not rhyme with e < e, i. e < ton. check. ^ and e < e, i do not rhyme with e < a. (Incidentally we note the three e's in his text.) In Vising, Etude, pp. 75, 84, and 93, we note the following : in Brandan, Gaimar and Fantosme, ai : ei before nasals. Under other circumstances there are variations : Brandan ; ai not : ei. 1 Cf. Walberg, p. xliv. 2 zt. Ill, 451, f.-n. 3 Keferences to several important articles are given by Neumann, Lithlt. IV, 18. Add to these bis own clear statement, Zt. XIV, 569. * p. xxvii. ^ I use this sign [:] to indicate " rhyming with," "rhymes with," etc. PHONOLOGY 45 ai : e (once, termes : lermes, 891). Gaimar ; ei not : ai ; ^ ei not : e ; ai : e. Eantosme -, ai:e often ; ^ ei : e seldom ; ei : ai ; ^ ei : e. From this state of affairs Vising concludes that ai became a monophthong (e) before ei did. [This to explain ai not : ei, and ei not : e.] When we do find ai: ei, it is after both had become monophthongs. If Vising means that ai and ei went different ways in developing into e, he may not be right; it seems more likely that ai passed through the ei stage before simplification to e.'^ Suchier mentions as char- acteristic of Anglo-Norman the contraction in a free sylla- ble ; likewise the endings ai, aie, continue as diphthongs in this dialect. We note an example of the latter usage in Boeve,^ where ai rhymes with ei (< e) ; as sai : mei, etc. 2. Orthogkaphic Variants of AI, EI, E. a. ee. This variant is spoken of at length by Stlirzinger,^ who says it is phonetically different from the ee < e and the ee < e (< a), and that it is never written ie. He gives exam- ples from Gaimar, Chardri, Langtoft, and others. Stim- ming^ adds a few : ^jees (pacem), fees (fascem), and the like. b. a. This occurs with some frequency; many of the examples are those of verb terminations where -ai has been replaced by -a ^ (fra =ferai, sa = sai, etc.), also fare, fates, plase, and the like. Stimming ^ refers to texts. Cloran ^ adds three examples from his text. c. ae. This has been noted only for Angier,^" and exam- ples seem confined to closed syllables {paestre, maestre, etc.). d. oi. This occurs often in Angier ^ {foit, soie, porroi, 1 Cf. here Kupferschmidt, p. 417. 2 Chardri's usage is similar to that of Fantosme ; cf. Vising, J^tude, p. 75, Such. Lithlt. Ill, 17. 3 In aine^ eine, aire, eire, and not when final ; Such. Auban, p. 4 ; Litblt. Ill, 17. * For detailed statement cf. Such. Gram. pp. 37-39. s Stimming, p. viii. « Orth. Gall. p. 41. ^ Boeve, p. 195. 8 For the opposite process (a > ai), cf. p. 43. 9 Dialog. Greg. p. 42. 1° Meyer, p. 193; Cloran, p. 41. 46 MANUAL OF OLD FRENCH DIALECTS etc.), and is not limited to words in which the ai follows a labial.^ An instance occurs in Melanges^ (cf. p. 32), soi~ SAPIO. e. aei. This seems very rare. There is an example in the Arundel Psalt. paeis/ and Cloran* notes two (traeit, vaeis) in the Dialog. Greg. Here the aei = ei (which in Angier is more frequent than ai), the ae being equivalent to the e (cf. p. 42). f. ea. Here I have noted only eat = ait < habet in Articuli Willelmi^ (cf. p. 57, ea for checked e). 3. A BEFORE L MOUILLEE. Anglo-Norman belongs to the dialects which change aVe > eVe. This occurs in Q. L. R. before the accent only ; later in accented syllables.^ 4. A BEFORE A NASAL. 1. Free Position. a. ain and ein. These two do not rhyme in Cumpoz and Bestiaire, nor in the Vie St. Gilles.'^ In the other poetical texts they do.^ In prose texts ei and ai interchange from the beginning. In Oxf. Psalt., for example, ai occurs even for etymological ei (fain, rain). In Lois Guillaume ei oc- curs but three times (two of these in atonic syllable).^ In Anglo-Norman the early confusion of ain and ein is charac- teristic ; the ei became ai in pronunciation in the middle of the twelfth century, and after that time the two were used promiscuously in orthography. 10 1 Cf. Such. Gram. p. 39, e. ^ jjo,,^. iv, 377. 8 Zt. XII, 24. * Dialog. Greg. p. 42. & Zt. XIX, 81, f.-u. 10. 6 Meyer-Lubke, Gram. I, 219, § 232. 7 Mall, p. 59 ; Walberg, p. xlviii ; Paris, p. xxvii ; Vising, Lithlt. IV, 311. 8 References in Stimming, Boeve, p. 106. ' Matzke, p. xliii. 10 Such. Gram. p. 72 ; (in) Grober's Gruiidriss, I, 672 ; Franqais et Provenqal, p. 23. PHONOLOGY 47 b. am and aine. It is worth noting that these two rhyme in late Anglo-Norman (as fontaine : lendemain)} c. en. Examples of en for aiyi are rare in earlier Anglo- Norman. In Q. L. R. we see enz ; in Roland, marrenes. A few others, including some from Boeve, are given by Stim- ming.- In the Aj^ocalypse, ain, ein, and en have but one sound, and in Bozon en is even written for ai {greyidre) and ei (mendre).^ d. an. This seems more frequent for ain than does en, and occurs in rhyme several times, as aJian : payi (panem), Adain; also in proper names, as Jolian : Abraam.^ 2. Checked Position. The most interesting phenome- non we encounter here is the appearance of au for a -\- nasal -j- consonant : quaunt, davaunt, graunt, and the like. The first examples accurately dated are given by Stlirzinger,^ the earliest being of 1266. Still earlier, however, may be those given by Koch ^ from Ms. L of Chardri (first half of the thirteenth century). We may say, in a general way, that the a + nasal + consonant became au during the second part of the thirteenth century. Note, for example, the remark of Meyer ^ in connection with the Fabliau du Heron, to the effect that it is not surprising to find in the latter part of a manuscript (written at the end of the thirteenth century), au7i and oun which do not occur in the poems of the earlier 1 Examples and references, Meyer, Bom. XXV, 255. 2 p. 197 ; cf. Such. Gram. p. 71. 3 Bom. XXV, 256 ; Bozon, p. lix. ^ Such. Gram, p- 71 ; Stimmmg, Boeve, p. 197 ; Walberg, p. xli ; Cloran, p. 43. ^ Orth. Gall. p. xxxix. 6 pp. vi and xxx. Meyer-Ltibke, Gram. I, 227, § 245, speaks of examples from Ms. O, which is later than L. Koch specifically cites aun of jVIs. L, while avowing that it is rare. Cf. also Such. Gram. p. 67 ; Stimming, Boeve, p. 173 ; and Uhlemann, Auban, p. 559. Stim- ming probably errs in including Angler's works among texts showing aun, since Meyer, p. 193, says aun does not occur in Vie Greg., nor does Cloran (p. 43) cite an instance for the Dialogues. 7 Bom. XXVI, 88. 48 MANUAL OF OLD FRENCH DIALECTS part of tlie same Ms. (written in the middle of the thirteenth century). The origin and nature of this aim have not been definitely- established. Koschwitz first suggested ^ that the diphthong arose after a + Z + consonant had become au, — it was equiva- lent to a + a T(-glide ("nachklang"). We note no further discussion till twenty years later, when Sheldon touches upon the point,^ proposing a phonetical explanation based on the result of the contact on English territory of the French and English pronunciations of a and of nasal consonants. Many questions are to be raised in a study of this problem, and the whole demands extended treatment.^ ain. I have noted one example of ai in checked position, — that in Arundel PsalL, caimp} 5. A BEFORE L OR R. For the endings -alem, -alum we note hesitation in Anglo- Norman between a and e. Examples are given by Suchier ^ and Walberg:^ tal, mal, mortal, etc. We find tal in Vie GregJ and al in general in Adam.^ This tendency for -al is Norman and Anglo-Norman.^ In cases where al does become el, poets may treat this el as fj ^^ (this forming the exception to the general e < A in Anglo-Norman, cf. p. 38). The same confusion exists for e and ^ before r. Notable exam- ples are to be seen in the Donnei des A^nants : " manere : contrefere; chanter : quer; identical rhymes are not cited outside of Anglo-Norman. Even the e < ai -j- r may be 1 Ueberlieferung (full reference above, p. 32), p. 21. 2 Child Memorial Volume, Boston, 1896, pp. 69-76. 3 Cf. the remarks on Sheldon's paper by Paris, Bom. XXVII, 320, and by Vising, Jhrsbrcht. Bom. Phil. V, 2, p. 289. 4 Zt. XI, 520 ; in the same text occurs cam ; Zt. XII, 50. 6 Gram. p. 22. ® Bestiaire, p. xli. 7 Lines 2699, 2767, for example. 8 Grass, p. 111. ^ Bom. XXI, 261. 10 Such. Gram. p. 25 ; Stimming, Boeve, pp. liv, Ivii. 11 Bom. XXV, 532. PHONOLOGY 49 treated as close.^ Obviously, Anglo-Norman did not demand an open vowel before r ; phonetic princij)les seem to play no part here. Cases of rhymes (or assonances) in -er came up frequently, particularly because of numbers of infinitive forms (in original -er and in -er < -eir). This -er was treated by the poet as -er or -^r according to the needs of his verse. 6. A PRETONIC. 1. Loss OF A IN Hiatus. The most noteworthy point in connection with pretonic a is the possibility of its loss in hiatus ; examples are very few, but in any case they have not been found in texts other than Anglo-Korman, according to Paris,^ who cites lenz = laenz for the Donnei, 695, Adgar and Chardri. We note that in the latter's Josaphaz, line 2621, Ms. L has lens by the side of Ms. leyns.^ In Vie Gi'eg., line 437, an i is inserted to break the hiatus, laienz. Is not clieles (< cJiaeles, quid velles'*) Brandan, 343, another instance of the contraction? Cf. a similar contraction of pretonic e ( leale > lele) on p. 60. 2. A REPLACED BY Other Letters. a. 6 for a. The use of e in cases like essalt, rechata, seems much favored in Anglo-Norman ; ^ we note, too, that before r + consonant, pretonic a > e with great frequency : merclier, herneis, etc.^ It seems, nevertheless, that cases in which a remains are the more frequent, and instances of phonetic e > a are not lack- ing ; cf., for example, chaval of the Cambridge and Arundel Psalters, and in the latter text e + s + consonant often becomes a; as, asperunt, aster, astrainge ^ (cf. p. 62, pre- tonic e). 1 Stimming, p. 193. 2 j^om. XXV, 53L 8 Koch, Variants, p. 186. ^ Such., Zt. I, 428. 6 Citations, Stimming, p. 172. We may add to his list Such. Auban, p. 37 ; Schlosser, ^. Z. i?. p. 7. 6 Pretonic e + r -\- consonant > -ar is not constant in general French either. Cf . Nyrop, Gram. I, 206. "^ Zt. XII, 28, 35. OF TKE UNIVERSITY OF 50 MANUAL OF OLD FRENCH DL-iLECTS b. ai for a, vaillet, maUiuet, etc. For citations here, cf. Stimming, Boeve, p. 172. c. 0, cm for a, ovesques, chaustel, etc. Cf. Stimming, p. 173. 3. Pretonic a + Palatal. Here the favorite orthography seems to be ei', reisini, teisez, etc.^ I have noted one exam- ple of ae as variant of the e of ei (cf. pp. 42 and 46) in Sardenai (L), line 98 : traeisist. P)esides ei we find a {esmaez, aez, etc.), and of course the regular ai and e. 4. Pretonic a -\- Nasal. As variants of a here we find ain (eiii), e and o.^ The mm of accented nasal a is to be seen quite as often in the pretonic syllable (demaunder, saunte, etc.).^ 7. FREE E. A proper classification of the developments here would be that of ei, ai, e, on the one hand, and oi on the other, be- cause ei, ai, e, represent the real Anglo-Norman products of free e, while oi was an imitation of continental usage. Then, too, ei, ai, e, denote a phonetic series : ei became ai (especially in the endings -eis, -eise, -eire, -eide, -eit, -eite) before ai (< a + palatal) had become q ; then ai (< ei) fell in with ai {oz, p. 70 ; Such. Anban, p. 3 ; Kupferschmidt, Gainiar, p. 417 ; Grass, Adam, p. 141 ; Vising, Etude, p. 09 {Brandan), p. 81 (Gairaar), p. 92 (Fantosme). 4 Koch, p. xxxi. ^ Vising, Etude, p. 14. •J Slimming, p. Iv. Stimming seems to indicate that Boeve is the only poetical text showing the confusion in rhyme in Anglo-Norman ; exs. cited by Koch for his text were not called into question by his reviewers, however, and Rottiger, Der Tristran des Thomas, Got- tingen, 1883, p. 30, cites two examples. Rottiger here gives a general bibliography on the point. Suchier, in his lieimpredigt, Halle, 1879 {Jilhl. Norman, I), pp. 09-71, takes up separate exceptions. Cf. his Grammatik, p. 09. In Lithlt. I, 25, he included Fantosme's Chronique among the texts showing the confusion, but he corrects the error in Eeimpredigt, p. 70, f .-n. Stimming, p. 185, gives reference to examples not in rhyme words. ' Cf. Such. Gram. p. 82. PHONOLOGY 55 12. FREE E. A number of interesting points arise in the study of e in Anglo-Norman ; we shall consider the following : (1) the reduction of ie to e, and the confusion of the two in rhyme ; (2) the quality of the e < ie ; (3) was ie a rising or a falling diphthong ? (4) different orthographies found in our texts. 1. IE AND E. We shall include here a similar reduction of ie < palatal -f a. The reduction was cited early in the history of Anglo-Norman studies, as characteristic of the dialect.^ As to the time of the change, we have to dis- tinguish between works in prose and those in poetry ; for the latter, again, we must separate the instances in which e for ie occurs in rhyme words from those in which it appears in the interior of the verse ; finally, we should observe the presence and proportion of the rhymes e : e, ie : ie, and ie : e. Only the last illustrates conclusively the Anglo- Norman peculiarity ; e : e would not prove it, for though the ie may have been simplified to e, it was done before the use of the e in rhyme, and might have been caused by reason of analogy or a similar principle. Actual statistics appar- ently indicate that our poets were not so careless in their use of ie and e as we might suspect. In prose texts we find e from the beginning of the Anglo- Norman period. Examples have been given for the Domes- day Book^ {Oliver, Cheure < CJiieure) ; Oxf. Psalt:' {perre, requer) ; Camh. Psalt} (seldom, eel, sede), Q. L. R.,^ etc. Eor poetry we note the following : Suchier ^ indicates that 1 Cf. Such. Zt. I, 5G9 ; Franrais et Proven<^aU P- 23, (in) Grober's Grimdriss, I, 572 ; Gram. pp. 5 and 47 ; Meyer-Lubke, Gram. I, 1G8, § 173 ; 173, § 179 ; 237, § 200 ; Vising, Zt. VI, 381. 2 zt. VIII, 358-359. 3 Harseim, p. 281 ; Zt. I, 5G9. * Schumann, p. 24. 5 Schlosser, pp. 3, 22 ; Plahn, p. 5 ; the latter gives the proportion- ate use of ie and e for Oxf. Psalt. , Camh. PsaU. and Q. L. B. as follows : ie : e, Oxf. Psalt. 100 : 43 ; Camh. Psalt. 100 : 14 ; ^. L. B. 100 : 26. 6 Auhan^ p. 3 ; cf. Uhlemann, p. 588. 56 MANUAL OF. OLD FRENCH DIALECTS ie and e were kept separate till after the middle of the twelfth century. About the same date is specified by Meyer/ who says that the rhyme ie : e became more and more frequent in England from the end of the twelfth century. Suchier, in another place,^ cites examples of the confusion in Philippe de Thaiin's Cumpoz, though the editor of the Bestiaire ^ claims no confusion in his text and restores ie in all cases. Philippe's contemporary, the author of the Brandan, keeps the two separate.'* Gaimar, too, observes continental rules,^ though one may detect a tendency toward the reduction. From the time of Fantosme,^ in any case, instances of the confusion become abundant, though we may find occasionally a poet who evinces a preference for e : e, or ie : ie. We note ie : e in Angier ^ (here, however, the propor- tion is in favor of the regular rhymes), in Guillaume de Berneville (examples * are given only for ie < palatal + a), Boeve,^ and Amaclas}'^ 2. Quality of E < IE. We refer, in the first place, to the remarks on the e < a, p. 38, When Meyer first speaks of e < ie,^^ he gives it as close. Next, Suchier^^ classes this e with that < A, as e. Paris marks the e as e in the Vie St. Oilles.^^ Meyer ^^ seems doubtful of the quality of the e and ie in his edition of Angier, giving ie (e), but offering i^ (^) as a substitute, saying there is doubt about it ; he attempts no discussion. Stimming ^^ specifies e. 3. Nature of the Diphthong. As to whether ie is a rising or a falling diphthong, the point can be made clearer by comparison with the diphthong iie (< o). The two are considered together, p. 71. 1 Bom. I, 72. 2 Qram. p. 47. ^ Walberg, p. Ixxxiv. 4 Such. Gram. p. 47. ^ Vising, J^tude, p. 86. 6 Ibid. p. 92. 7 Meyer, p. 194 ; Cloran, p. 47. ^ p. xxix ; cf. here the corrections of Mussafia, Bom. XI, 594, who reduces the number, already small. ^ Stimming, p. ix. 1*^ Andresen, p. 85. " Bom. I, 72. i'- Zt. II, 293. 13 p. xxix. 1* Bom. XII, 194. ^^ Boeve, p. ix. PHONOLOGY 57 4. Orthographic Variants of IE (E). a. ee. The essential facts with regard to this ee have already been stated, p. 41. A few references may be given here.^ b. ei. Examples for this have been noted for the Bestiaire ^ and later texts.^ In the Apocalypse I find teirz, 405, ceil, 515. c. eie. This I have found only in Arundel Fsalt.,* peiez (of. preiere, creiendrums). The first e seems the parasitic, pretonic e described on page 61. d. ^. i for ie occurs frequently : milz, arire volentirs, etc. Stimming^ gives a list of texts showing it. I find it often in Otinel and Aspremont, both in the interior and at the end of verses : tirz, 128 (cf. Olivirs, 58, chevalirs, 72, premir, 74, etc.), also in the Fabliau du Heron, mestir, rivire,^ etc. 13. CHECKED E. The confusion of ^ with e in checked position has already been commented upon, cf. p. 53. The variant ei occurs for ^, too, in seit (septem), Gaimar.^ In addition we note ae and ea which appear as variants of ^ alone ; ae occurs often in the Dialogues of Gregory^ {apraes, engraes, daestre), Aiiban, saet (septem), PUerinage, bael, 216. In Ms. L. of the Alexis we find seat (septem), which is a borrowing from middle English.^ 14. E BEFORE A PALATAL. Here we find for the most part i, though not infrequently ei (preise for pme). With regard to the ei there is a ques- tion whether it is a mere graphic variant of i, or whether it represents a sound other than i. Stimming gives examples ; ^^ 1 Mall, Cumpoz, p. 68 ; Meyer-Liibke, Gram. I, 173, § 179 ; Such. Gram. p. 48 ; Stimming, Boeve, p. 202. 2 Walberg, p. Ixxxiv. 3 Stimming, Boeve, p. 202. * Zt. XI, 530 ; XII, 6, 45. 6 Boeve, p. 202. ^ Bom. XXVI, 88. 7 Stimming, Boeve, p. 175. ^ Cloran, p. 47. 9 Such. Zt. XIX, 81, f.-n. 10, Gram. p. 42. See also under a + palatal, p. 46. ^^ Boeve, p. 187. 68 MANUAL OF OLD FRENCH DIALECTS we may add those taken from the Oxf. Psalt. (neie), Camb. Psalt. and Vie St. Gilles} In the latter text words in which E + palatal occur are not found in rhymes in i, so that Paris hesitates to say if De Berneville used iei, ie, ei, e, or i. Vising, ill his review,^ says such a state of affairs is the result of a coincidence, and that the regular Anglo-Norman product was i. 15. E BEFORE A NASAL. 1. Free Position. Here we find le and e as in the case of E + simple oral consonant, but we must remember that in Anglo-Norman the continental ie before a nasal does not suffer the reduction to e as frequently as it does when before an oral consonant. a. Orthographic variants (ee, ei, i, eie, iei). Stimming^ gives texts showing these. In Otinel we note bin {hien) 123, and veint (vient) 227, the latter also in the Apocalypse, 83, 137. The Brandan veient^ is another example of the para- sitic pretonic e (cf. p. 61). An example of iei is furnished by Camb. Psalt. mieins, XVII, 34. 2. Checked Position. a. Variants {ei, eie). The variant ei occurs for both e and ^ in checked position (cf. p. 57), dedeins, leins, veint. A further variant of the ei arises from the addition to it of the parasitic post-tonic e, as seyens (ecce intus). Cf. eyens (antea) and ley ens (illac intus). ^ 16. E (AND E) before L. We have already referred (p. 54) to the confusion of e with ^ when before I, the I having a tendency to keep an open vowel (cf. a -{-I, p. 48). The history of the vowel before the I must, of course, go hand in hand with that of the I itself, which remains, for the most part, in our earliest texts, though the tendency to vocalization {> u) is to be 1 Harseim, p. 282 ; Schumann, p. 25 ; Paris, p. xxx. 2 Lithlt. IV, 311. 8 Stimming, Boeve, p. 203. * Hammer, p. 91. ^ Stimming, Boeve, pp. 180 and 185. PHONOLOGY 59 noted at a very early date (cf. p. 87). Again, where the I is kept, we have to distinguish between the cases in which we find the glide a {-eal) or simple -el. Philippe de Thaiin did not know of -eal < -el (according to his two editors '), though Suchier ^ cites eals, ceals from the Cumpoz (L.). The Oxf. Psalt. has, as a rule, -el (oisels), -eal occurring exception- ally ^ {chalemeals). In Camb. Psalt^ -eal is not infrequent, and -eols occurs too. Oxf. EolancV has -els, except in one instance (healmes). The Q. L. R. ^ has -els almost without exception (beaus once). In the Branclan we find (by the side of older forms) oiseiis, beus, etc.^ This -eus (without the glide) is found in Norman and Anglo-Norman texts, seldom elsewhere in Northern French.^ After the time of the early texts the combination e + / -}- consonant is represented in many different ways, older and later forms being used with- out apparent discrimination. As a rule we may say that the normal Anglo-Norman form was -eals, -eaus. 1. Variants. We have to record the following variants. a. eus, us. eus has just been cited from Brandan; exam- ples occur in Vie St. Thomas (as II, 47, beu sire) and Boeve.^ The -us is probably a development from -eus (cf. -eol > -ol, eal > al), and occurs only in later texts. To Stimming's list ^^ we may add Ajjocali/j^se chevus, 52, 452. b. als, aus. These may be reductions from -eals, -eaus. In Camb. Psalt. we find imissals,^^ neiial, oal; in Q. L. R., halme.^^ We note haume in Lois Ouillaume,^^ oisaus in Bes- tiaire and Angier." c. eols, eous, ols, oiis. Iceols is found in Camb. Psalt. and Brandan. The o in iceols is doubtless a glide (like the a in -eals) ; -ols is a development from eols ; in the Camb. Psalt. 1 Mall, p. 65 ; Walberg, p. xliii. 2 Gram. p. 82. ^ Harseim, p. 282. * Schumann, p. 23, 5 rorster, Zt. I, 565. 6 Schlosser, p. 21. ^ zt. I, 565. 8 Such. Gram. p. 81 ; Meyer-Liibke, Gram. I, 163, § 163. 9 Stiinming, p. 174. i'^ Boeve, p. 175. ii Schumann, p. 23. 12 Schlosser, p. 21. ^^ Matzke, p. xlvi. i* Such. Gram. p. 81. 60 MANUAL OF OLD FRENCH DIALECTS we note cho.vols (XXXIX, 15, LXVIII, 5) which is of not infrequent occurrence; Gaimar has ous (tllos).^ d. oels, Olds, euls. In the Arundel Fsalt. we have a num- ber of variants, some of which are not found among the above. We note : icoeh,'^ ouls (couls, icouls, common),^ eids,'^ els, eols, ols.^ 17. PRETONIC E. (Here we have no occasion to distin- guish between close and open e.) 1. Fall of Pretonic E. a. Before a vowel. The early fall of pretonic hiatus e in Anglo-Norman is always spoken of ^ as characteristic of that dialect as compared with French of the continent. We have to include here ending accented forms of preterites (like eilmes) or imperfect subjunctives (like eust) which origi- nally had pretonic o in place of the e (oilmes, oilst) ; ^ in the Q. L. R., for example, we find oust (monosyllable), east, and ust. We must take account also of the peculiarly Anglo- Norman contraction illustrated in lele for leale, which has already been referred to on page 49 (pretonic a). In the Cumpoz and Oxf. Psalt. we find the e retained, though in the verb forms referred to the o has weakened to e. The Lois Guillaame keeps this o consistently (oust, poust, etc.).^ The Camb. Psalt. and Q. L. B. show the fall, while the Brandan and Gaimar^ retain the e. After the time of Fantosme the fall becomes frequent ; we may say, therefore, that even in Anglo-Norman the poets do not regu- larly elide the e previous to the thirteenth century.^^ 1 Such. Gram. p. 82 ; Stimming, Boeve, p. 175. 2 zt. XI, 516. 3 Ibid. XI, 520, 524 ; XII, 4, 14, 22. ^ Ibid. XII, 14. & Ibid. XII, 1. 6 Of. for example. Bom. I, 71 ; XXV, 531 ; Zt. I, 5G9; IV, 419; Meyer- Liibke, Gram. I, 319, § 377 ; Stimming, Boeve, pp. xxxiii and 178. ■^ These forms were made the object of especial study by Suchier, Auban, p. 27. Paris summarizes Suchier's results and gives the best general statement for the whole phenomenon, in his Gme. de Berneville, p. xxii. 8 Such. Litblt. XXII, 121. » Cf. Vising, ^tude, p. 82. w Cf. Meyer, Vie St. Thomas, p. xxix. Meyer believes the e is not to be elided in his text. PHONOLOGY 61 b. Before a consonant. The fall of the e here is even more characteristically Anglo-Korman than that of the pre- tonic e in hiatus ; the loss of the latter e becomes general on the Continent too, but that of pretonic e before a consonant does not.^ For convenience of observation, we may divide the examples according as they show the combination con- sonant -\- e -\- consonant, or vowel -}- e + consonant. In the former case the second consonant is usually r, and instances are to be drawn mostly from futures and conditionals {frai, frei, trouvrai, durrai), though examples furnished by other words are not lacking (as vigroiis, pelrin, and the like).^ For vowel -{- e + consonant we note the following : Suchier ^ speaks of seira (< seiera) as one of the Anglo-Norman marks of the Oxf. Psalt. The same fall of the e occurs in Q. L. R. in seven futures (enveirai = enveierai) by the side of twenty-five which keep the e."* Stimming^ gives further examples, like espuntez = espuentez. The examples here are illustrations of the fall of the protonic syllable, and are, of course, to be noted in connection with the similar fall of the post-tonic e after vowels (cf. p. 63). 2. Parasitic Pretonic E. The insertion of an e where it does not belong etymologically takes place under the same conditions as those for which we have just noted its loss ; in fact, this process seems more frequent than the loss of the e. It is not characteristically Anglo-Norman, though very usual there. Stimming treats of this point at length.*^ The svarabhactic e appears often in futures and conditionals, and words in general where one of a group of consonants is r, perderez, ferete (cf. averil, Otinel, 46), though a few ex- amples occur for other consonants, as sabeloun, jovene. The e sometimes acquires such a value as to have other vowels substituted for it, as : fert4 > ferete > ferite ; perdrai >per- 1 Cf. Such. Auban, pp. 33 and 40 ; Paris, Horn. XXV, 632. 2 Cf. Schlosser, Q. L. B. p. 4 ; Hammer, Zt. IX, 85 ; Stimming, Boeve, p. 178. ^ zt. I, 569. * Merwart, p. 10. 6 Boeve, p. 178. « Ihid. pp. 179-181. 62 MANUAL OF OLD FRENCH DIALECTS derai > perdurai (cf. perle > perele > perele, p. 64). Stim- ming considers apart from the cases just considered those in which the parasitic e appears before or after vowels and diphthongs, a.S2Jerdeu (perdu), deeyns (de intus), 7neit (misit), seonge. Such words call for various explanations ; as, anal- ogy, mistakes of scribes, and the like. Doubtless a similar group of examples is to be added here, made up of words in which the pretonic parasitic e precedes the diphthong ie, as peiez, {pieds), veient (vient), etc. Cf. pp. 57 and 58. 3. Pretonic E before a Nasal. Here we need only refer to Stimming ; ^ he distinguishes cases in which the e occurs before a single consonant from those in which it is found before several. In either case e interchanges with a, and Stimming says that such is the case particularly when e precedes a single nasal consonant. It may be questioned whether this statement will hold good, if we take from the list of examples of e before single n the many forms of (menacer >) manacer, where the a dates back to Popular Latin times.^ am — {an — ) occurs often in Alexis, as am.- pairet, L, 2, e, amfant, L, 5, b. In Arundel Psalt. we note anemis.^ It seems very probable that the a merely reflects the general Anglo-Norman fondness for pretonic a already spoken of on p. 49 and referred to again in our next paragraph. 4. Variants of Pretonic E. a. Phonetic (a, i, o, u). We here have to refer, in the first place, to our note on pretonic a (p. 49, § 6, 2, a). Our dialect has a distinct predilection for a in this position, original a being sometimes kept (chavaler), or else supplant- ing etymological e (ascient). Stimming^ gives examples of the other variants, i, o, u, — , which are to be explained usu- ally by assimilation, attraction, and the like. To his instances of u we may add dumurez (Auban),^ Nuvers, espuruns (Ama- das et Ydoine),^ jutas, jutiez (getter) from the Arundel Psalt J 1 Boeve, p. 185. ^ cf. Nyrop, Gram. I, 15, § 12. 8 Zt. XII, 55. * Boeve, p. 177. ^ Zt. II, 343. 6 Ibid. XIII, 86. 7 Ibid. XII, 51, 53. PHONOLOGY 63 and jutta from one of the latest texts, the Apocalypse, 1099. For i we find chivaus, chimin, etc. ; for o, solum, bosoyyie, etc. b. Orthographic (eo, ae, oe). The eo, which we have already recorded (pp. 43 and 53) as a variant of accented e, occurs for e in pretonic position too, though not frequently ; to Stimming's examples ^ should be added the enfeormethe of the Camh. Psalt? We may mention ae also which is of frequent occurrence in the Arundel Psalt. ; meslaescerai, baealtet,^ and noted by Schlosser * for the Q. L. R. raegnad. Schumann ^ gives oe, as distinguished from ae, for the Camh. Psalt. estoerat, portoeras, etc. 18. POST-TONIC E. 1. Fall of Post-tonic E. a. After vowels.^ The loss of the e after vowels has been cited as characteristic of our dialect since the time of the earliest studies on the dialect. In the very first number of Romania, Meyer calls attention to it. This loss swells to a notable extent the total of irregularities in Anglo-Norman, since it brings about the confusion of ee: e in participles or substantives ; it is responsible for feminine possessives like mei, tu (cf. p. 118), and of imperfect endings -ehit (for -eient, cf. p. 124). As to the time of the fall, all agree that it took place earlier in Anglo-Norman than on the continent; we may place it in the course of the twelfth century. The Brandan is the first poetical text with which the phenome- non is usually associated. Stimming^ gives a list of texts and examples. We may add from the Arundel Psalt. im- 1 Boeve, p. 178. 2 Sucli. Qvam. p. 82. 3 Zt. XI, 521 ; XII, 16. * p. 37. ^ p. 27. 6 The general bibliography on this point is the following : Bom. I, 71; Such. Auhmh P- 5; Koschwitz, Zt. II, 482; Vising, iStude, p. 70 ; Lithlt. IV, 311 ; Meyer-Lubke, Oram. I, 261, § 305 ; 290, § 340 {glori, estudi, etc.); Stimming, Boeve, p. 1 (effect on metre); Nyrop, Gram. I, 210. For later Anglo-Norman, cf. Meyer, Bozon, p. Ix. 7 Boeve, p. 182. 64 MANUAL OF OLD FRENCH DIALECTS perfect forms like criou, humiliou;'^ in Amadas et Tdoine^ we note rai : saie, pasme : pasynee. b. After consonants. The loss of the e here, too, is con- sidered an Anglo-Norman pecnliarity. It is by no means as early as that after vowels, and becomes frequent only toward the end of the thirteenth century. Suchier studies this point in detail^ and concludes that the fall of e was gradual, taking place first (and mostly) after r (already in Arimdel Psalt} we note sir), next after / (iiul = nule), finally after m or n (dam, un = dame, une). 2. Parasitic Post-tonic E. Since, as we have just seen, the e fell in so many cases, and with no apparent consistency, we may readily conceive of the state of uncertainty among our scribes as to its proper use and proper omission ; we find numbers of cases where the post-tonic e is added with no etymological right. Stimming speaks of these at length;^ examples are nule (masc), foreste, and the curious instances where the e is inserted between a stem and a flexional s as prest(e)s, coup(e)s, — this being just the contrary of the cases where an e with an etymological background is omitted, as tuz {— TOTAs), veys (= vias), seys (= siAs), and the like. We may add some examples from the Arundel Psalt. where we find the masculine participles tresturnee, eslevee ;^ meismee,^ and jures (= jours). ^ The parasitic e may have other vowels substituted for it: as i (pans > panes >) panis, (Jins > fi7ies >)j^?iis; and u(meilurs > meilicres >) ineilurus (Boeve^). Cf. below (3, orthographic variants). In the Apocalypse we note a remarkable perversion in the example furnished by the rhyme p)erele : cristele (line 1297), where the stages were perle > p4rele > pei'dle (cf . p. 61 and 1 Zt. XII, 19, 26. 2 ji)icl^ XIII, 86. » Auban, pp. 36-39. ^ Zt. XII, 7. Cf. also the rhymes cited by Meyer, Bom. XXV, 255: per e : primer^ Pol : paroles, cors : paroles from Descente St. Paul. 6 Boeve, pp. 182-183. e zt. XI, 619 ; XII, 31. 7 Ibid. XII, 23. 8 jiid^ XII, 2. 9 Stimming, p. 184. PHONOLOGY 65 under the consonant r). In the same text, line 477, we note foiez (=zfois). 3. Orthographic Variants (i, u, a, oe, ae). These are given by Stiinming^ as follows : i, meii, -istis, -int. (This i occurs with great frequency in Arundel Psalt. : chosis, eglisis, taisis, vindrint^) ; u, trovunt; a, ova. (This a for e occurs in nearly every strophe of Alexis L, as nostra, strophe 3 ; tendra, 24 ; lungament, 69 ; anames, 122, etc.) We may add also the oe of the Camh. Psalt. cited by Schumann,^ terroe, pahnoes. In the printed edition (Michel) we find oe in the last few psalms only ; elsewhere ae, as sujlae, XXV, 2 ; fuiaent, XXX, 12, etc. (many cases). In the Arundel Psalt. too, we note ae : terrae.^ I 19. I, TONIC. There is little of importance to note here, and I shall merely make reference to several points treated by Stim- ming. These are the rhyming of i : e ; ^ i:u (cf. p. 79, § 32) ; comparative use of y and i by scribes ; ^ nothing especial arises in the consideration of i before a nasal. ^ As varying orthographies of ^ we find e (esgles = eglise), ei {conqueis),^ ie {fiez; cf. below, § 20), and we may add the u of the Q. L. M. afuhle < affihula.^ 20. I, ATONIC. Here we may add to the two variants given by Stim- ming; ^*^ that is, e (dener) and ei (deiables), three others : ai of Camb. Psalt. LXXIII, 14, daiahles (cf. GUI, 26, daible) ; oi of Arundel Psalt., foiede,^^ and ie of Vie Gregoire, line 1867, dierrai = dirrai. 1 Boeve, pp. 183, 184. 2 zt. XI, 528 ; XII, 12, 14, 47. 3 p. 27. * Zt. XII, 10. 5 Boeve, pp. Iv, 188. e jj)i(i p. ige. ' Boeve, p. 188. » Ibid. p. 187. 9 Such. Gram. § 11, a, 6. 10 Boeve, p. 188. " Zt. XD, 14. 66 MANUAL OF OLD FRENCH DIALECTS 21. TONIC 0. So far as results, noted in rhymes or orthography, are concerned, we do not have to observe a distinction between free and checked tonic q (except in the case of eu which occurs but seldom, cf. below, p. 68, 3). Our first poetical text, Cumpoz (L), shows a distinct preference for keeping original g in the orthography, and even substitutes the o for etymological w < tJ (though the examples given ^ are mostly for u in pretonic position) : jogier, mors, etc. In Brandan, too, we have a similar state of affairs,^ and at a later time, in Angier,^ Chardri, and Adain,^ we note a tendency to keep 0. I record this fact only for the sake of historical com- pleteness ; the really important feature in connection with o is treated in our next paragraph. 1. U AND 0. a. We find u written for o in our earliest texts. In fact, such a use must be very old, though we can- not accept, in its entirety, the reasoning of Liicking,^ at least in so far as it concerns Anglo-Norman. He argues that the writing of u for o must have preceded the time when u (< u) began to have the value of il, since after it signified that value, scribes would not think of using it to designate still another sound (that of the u < o). The use of u for o is likewise to be found on the continent, but it seems to have been favored nowhere as it was in Anglo-Norman.^ No difficulty attaches to the appearance of u for o in prose or in the interior of verse. We find u in the Cumpoz,^ in the Lois Guillaume^ (here o is found in only one word), in Oxf. Psalt.,^ Camb. Psalt.,'' and Q. L. R}' b. V/e have to consider, however, that in some poetical 1 Mall, p. 41. 2 Hammer, p. 87. 8 Meyer, p. 197. * Stimining, Boeve, p. 100. ^ Die AeUesten Franzosischen 3Iund(irten, Berlin, 1877, p. 149. 6 Such. Gram. p. 14. '^ Mall, pp. 41, 40, 47, ^ Matzke, p. xlvii. ^ Harseim, p. 294. ^'^ Schumann, p. 40. ^^ Schiosser, p. 49. PHONOLOGY 67 texts this u, as used for o, rhymes with u aun. The ou, as used before nasals, is doubtless an extension of the ou already spoken of (p. 67, § 21, 2), which was borrowed from Continental French. Koschwitz i^ made a study of the point, and indicates that the development may have been o > ou > u. Our texts, however, do not substantiate this theory; ou appeared at a comparatively late date. The first example I have noticed is that in Chardri,^^ Ms. V (thirteenth to fourteenth century, cf . p. 22) ; a few occur in Boeve " (ount, vount, fount) and Auhan}^ 1 Boeve, p. liii. ^ Walberg, p. xlvi. ' Cloran, p. 48. 4 Zt. VIII, 358. 5 Harseim, p. 295. ^ Schumann, p. 4L 7 Andresen, p. 86. ^ Ulilemann, p. 575. ^ Stimming, p. 19L 10 Koschwitz, Zt. II, 34.3. " Kupferschmidt, p. 417. 12 Ueberlieferiing, p. 32 ss. i^ Koch, p. xxxi. " Stimming, p. 192. i^ Uhlemann, p. 569. 70 MANUAL OF OLD FRENCH DIALECTS 9 The definite history of many of the phenomena arising in a study of o in Anglo-Norman is yet to be written. One cannot read what has been proposed and then, after an observation of the examples for himself, be satisfied that the last word has been said. I offer some fragmentary suggestions on various points, hopiug thereby to invite attention to and discussion of them. A final solution demands a careful and detailed study of o in all the texts ; such a study will surely bear fruit.^ 24. 0, UE, EU, U, AND E. 1. 0. The keeping of o in the undiphthongized (by the side of the diphthongized) form is characteristic of Anglo- Norman. We find o in our older texts, though not with consistent frequency; for example, in Omiipoz'^ and Bes- tiaire;^ in Lois Guillaume^ (pojjle, pot, quor, etc.), to the entire exclusion of ue or oe; in the Psalters o is found, though not in the majority of cases ^ (in Arundel Psalt. we note qiior^), while in the Q. L. R. o prevails in the proportion of four to one.^ We refer to examples of o in later texts, as follows : Angier,^ Chardri,^ Gme. de Berneville,^^ and Amadas}^ (For oi < o + palatal, cf. below, p. 80, § 33, 2.) 2. UE. This is not the place to enter upon, or even to refer to, the extensive bibliography of the interesting gen- eral questions as to the history of the diphthongs ue < o and 1 The points following may be found, amplified in some cases, in an article in Mod. Lang. Notes, XVIII, lOG-111. 2 Mall, pp. 47, 48. » Walberg, p. Ixxxv. * Matzke, p. xlvii. ^ Harseim, p 292 ; Schumann, p. 33. 6 Zt. XI, 624, 525, 52G ; XII, 16. ' riiilin, p. 5. 8 Meyer, p. 19G ; Cloran, p. 48. 9 Koch, p. xxviii (cf. Zt. Ill, 593). lo Paris, p. xxxi. ^1 Zt. XIII, 85. For a general statement, cf. Such. Gram. p. 41. PHONOLOGY 71 ie < E (on p. 56 we referred to the present section for a con- sideration of the latter) : whether the u of ue was pronounced ou or ii ; whether the diphthong was rising or falling ; when the pronunciation o (Mod. Fr. eu) came in, and the like. This latter point is particularly difficult to decide for Anglo- Norman, both on account of the great confusion of orthog- ra^jhies, and because the eu to which we are accustomed in French texts of the continent is extremely rare in Anglo- Norman (cf. just below, no. 3). We have reason to suppose that, for a time at least, ue had a double value in Anglo- Norman, that is, u-e and 6} The earliest text for which I note the pronunciation o claimed for ue is Chardri.^ (There is evidence, however, that the sound 6 existed earlier; cf. below, 6, e.) As to whether ue and ie were rising or falling diphthongs it is difficult to discover essential facts on which to base conclusions. We have already seen (p. 55) that e for ?e is a leading Anglo-Norman characteristic ; this, of course, points to a pronunciation ie. We shall see below (no. 4) that u occurs for ue ; this points to a pronunciation ue. The Oxf. Psal, where we find marks of accentuation, renders ie by i^, ue by both ue and ue. It seems as if the stress must have varied at different periods of the language, or with dif- ferent scribes or in some other (unknown) way; such was obviously the case in our dialect where we have ue, u, e ; ie, i, e; theorists on the original nature or state of these diphthongs must seek their data outside of Anglo-Norman lines.^ An important point to be observed with regard to ue in Anglo-Norman is that it may rhyme with e, and thus give rise to a set of rhymes whose exact parallel is not to be found on the continent; such rhymes are quer: honurer, 1 Koschwitz, Ueberlieferung, pp. 29, 73, 2 Koch, p. xxviii. 3 For general remarks here, cf. Suchier, Zt. I, 291 ; Gram. pp. 40 and 48 ; Nyrop, Litblt. I, 223. 72 MANUAL OF OLD FRENCH DIALECTS chanter, coimter. They are to be foimd in Auban, Donnei, Bozon, and other texts.^ 3. EU. AVe have already referred to the rarity of eu for o (§ 21, 3), and we shall have to record still fewer cases of eu for (ue) q. Stimming ^ gives seut, veiit (in which, however, the u may represent an I which has vocalized), queur, peuple, veuUe, petit, demeure. [demeure should not be included here, as its was o.^] We have to study separately each word and text to determine approximately the phonetic value of the eu. For the Vie Thomas, for example, Meyer ^ assigns different values to the eu, according as it represents general French ue (vent) or g {eurent = orent). 4. U. u, representing o, has always been recognized as a marked Anglo-Norman characteristic ; it is found in Philippe, Oxf. Psalt., Q. L. R., and Brandan, among our earlier texts : buf, put, vult, uvrent, etc.^ Our dialect is apparently the only one in which we find an q represented by an u. How- ever, if we look upon the u as a reduction of the diphthong ue, and not as a variant of undiphthongized o, there is noth- ing striking about the u. In the earlier stages of Anglo- Norman studies the u was treated as such a variant of o,^ and this might seem natural enough in view of the fact of the Anglo-Norman fondness for keeping the o (cf. p. 70, § 24, 1), which is itself a distinct peculiarity. Considering, however, what we have already said as to the breaking up of the diphthongs ue and ie into u, e and i, e, it seems more 1 Cf. Stiirzinger, Orth. Gall p. 46 ; Paris, Rom. XXV, 632 ; Meyer, Bozon, p. lix, 3. 2 Boeve, p. 208. 3 Cf. Zt. II, 509 and Paris, Gme. de Berneville, p. xxxi, foot-note 1. Stimming gave demeur under o on p. 190. On p. 208 he is citing from Stiirzinger, who (wrongly) gives q. 4 Introd. p. xxix. Cf. touz : lens (locum), Apoc. 1. 309. 6 Cf. Grober, Zt. II, 509 ; Meyer-Liiblce, Gram. I, 202, § 217 ; Such. Frangais et Proven(;al, p. 23, (in) Grober's Grundriss, I, 572 ; Gram. p. 41 ; Stim. Boeve, p. 208. 6 Cf. for example, Mall. Cumpoz, p. 50 ; Fichte, Camb. Psalt. p. 63 ; Zt. II, 481. PHONOLOGY 73 probable, to the present writer, that the u in question is out of ue. I know of no suggestion that i (for ie < e) is a vari- ant of E, and yet the history of ie and ue must have been the same in many particulars. It seems to me that to claim, even in Anglo-Norman (where I admit all kinds of irregu- larities), an u for a free tonic o before oral consonants, is going too far. With ue so constant a product in general French and in Anglo-Norman, too, explanations of any phase of the history of 9 that do not take ue into account, where possible, are hazardous. Surely, when we confront vult with vuelt, it is more reasonable to suppose vult a later form of vuelt than to think that, with no apparent reason, o went two different ways, becoming u, in vult, diphthongizing in vuelt. There are cases in which the ue can have played no part, and where we have evident confusion, or interchange, of orthographies by the scribes. These are given by Stim- ming : ^ vult (voluit), nus (noster), etc., and the imperfect endings -ue, -ut, for -oe, -oue. (Cf. below, p. 123, § 64, 1.) 5. E. e for ue is said to be characteristically Anglo- Norman, too. There is little specific discussion on the time difference between the e and the u as used for ue. Meyer- Ltibke ^ treats of ue > e first, saying that it occurs early, then goes on to say, ''One is surprised to find u also for ue,''^ as if e were the more characteristic or usual. Stimming^ says ue > e, especially after the year 1200. If what I have said above (No. 4) as to the derivation of u from ue is cor- rect, it follows that I must consider the original accentuation of the diphthong to have been ue, and therefore any time difference must be in favor of u as older than e. Any such difference in Anglo-Norman is merely relative, and we need not suppose that u was used regularly for a period of years and that afterward e came in ; the difference was doubtless slight, and the use of the one or the other depended on the 1 Boeve, p. 189. Cf. voult in Vie St. Edmond, 1. 454. 2 Gram. 1, 202, § 217. » Boeve, p. Iviii. 74 MANUAL OF OLD FRENCH DIALECTS circumstances which influenced each individual scribe. The texts cited by Stiirzinger and Stimming^ for e are compara- tively late, Adgar, Angier, Auban, etc. {em, fleves, sell, velt, etc.). I note hem (and in tonic position) in Arundel Psalt} 6. Orthographic Variants. a. oe. This is found with great frequency in Norman and Anglo-Norman texts, particularly at the beginning of words, and is supposably a device of the scribes to distin- guish ue ( = ue ai, e > ei, u > id, cf. pp. 50, 53, and 6S) on account of the frequency with which his eye encountered ai, ei, ui which had developed from a, e, u (o, o) before palatals. To my mind oi for q corresponds to ui for (it). For oi ple, etc.* Suchier speaks of it as indicating the sound b (eu)* Cf. my remarks above (c). f. 6. Examples for this have been cited ^ for the Oxf. Psalt. only: pd2'>le, reproce, oil, and oVie. Compare my re- marks above (c). 1 Such. Gram. p. 42. 2 Boeve, p. 208. » Such. Gram. p. 41. * Cf. Vising, Jhrsbrcht. II, 1, 250. 5 Such. Gram. p. 41. PHONOLOGY 77 25. CHECKED p. Here we make note of two points : the presence of on for the Q, and the rhyming of q : o. 1. ou. It goes without saying that checked o remains in most cases. Stimming ^ finds only one example of ou : toust from Boeve and Bozon. I add cours (cordibus) from Arund. Psalt.^ Stimming thinks that the word oustent (obstant, Boeve) may point to an u (g) as a phonetic variant of checked Q in Anglo-Norman. This point demands further study. 2. 0:0. Aside from the cases in which the q occurs be- fore a nasal (cf. p. 78, § 23), there is little evidence in favor of such a rhyme. Mall had supposed that such might be the case for the Oumpoz and Bestiaire, but Walberg ^ affirms that the q rhymes only with itself in the Bestiaire and cor- rects Mall's examples. 26. LOCUM, FOCUM, JOCUM. I do not attempt to say anything new about the various forms of these words, but shall simply refer to places where the Anglo-Norman variants (lieu, liu, leu, etc. ; gieu, giu, ju, geu, etc. ; fe^i, fii, fou, etc.) are especially mentioned or dis- cussed.^ I shall, however, call attention to the form ltd (lids). No mention is made of it in the bibliography cited, except by Stimming, who finds one example of milui in Boeve and supposes it may be another case of "umgekehrte Schrei- bung " (hi > lui, confusion with etymological ui, cf. p. 68). The cases of its occurrence, however, seem sufficiently numerous to justify our giving lui a place beside the other variants of locum in Anglo-Norman; perhaps it is worthy 1 Boeve, p. 189. 2 zt. XI, 515. ^ jj^f-^^ xi, xlv. 4 The best short statement for them in Anglo-Norman (and Norman) is that of Sucli. Gram. pp. 5.3 and 56. Cf. in this connection Meyer- Ltibke, Gram. T, 190, § 196 ; LTilemann, Auban, p. 572 (here we find additional bibliography) ; Stimming, Boeve, p. 204 ; Walberg, Besdaire, p. xlvii ; Busch, Laiit- und Formenlehre, p. 35. 78 MANUAL OF OLD FRENCH DIALECTS of further study, too. I note the following : Arundel Psalt. lui^ (in connection with ltd the form lai'^ < lacum is to be observed) ; Brandan, 1. 86 ; Denis Py ramus, Vie St. Edmond, 1. 729, lid (not in rhyme) ; Madeleine, 7, d, luis, and 10, c, hds:plids; Apocalypse: milui, 218 (p. 208), lid-.celui, 340, 247 (p. 209), luisitidz, 310 (p. 211), luiifui (past, particip. of fuir), 948; one example of fu (preterite of itre) : fui (focum), 843 (p. 232). 27. BEFORE A PALATAL. We shall treat the id arising from this combination in connection with the similar ui < u before a palatal, p. 80, § 33. There, too, will be considered the variant oi. 28. Q BEFORE L MOUILLEE. Here Anglo-Norman offers nothing new. I note the form orguoil in the Arundel Psalt^ (cf. orguel^) ; in Angier (1. 32) we find the reduction of ue to e in veil = vueil. 29. BEFORE A NASAL. I considered o before a nasal in connection with o before a nasal in those cases in which o did not diphthongize (cf. p. 69). It does not diphthongize when checked, of course, and even in free position o alternates in about equal pro- portion with ue. The use of these two in our earlier texts is given by Suchier.^ We find u (Jimne) and e (hem), too, just as before oral consonants.^ Examples of the usual variants for o before consonants occur here also. We note oe in Arundel Psalt. : hoens, soens; ^ eo is found, as in Camb. Psalt. and Q. L. P. : heom, seon ; the Q. L. R. shows ueo also : hueom.^ Stimming finds one or two examples of ou (^moun) and ui (suyn) in Boeve. 1 Zt. XII, 11. 2 iiifi XII, U. 3 /ftiVi. XI, 529. * Ibid. XII, 19. ^ Gram. pp. 73, 74. Add Lois Gme. p. xlvii. 6 Cf. Oxf. Psalt. (Harseim, p. 203), Camb. Psalt. (Schumann, p. 3G), Q. L. li. (Schlosser, p. 43), Angier, Aiiban, Boeve; Stimming, p. 209. 7 Zt. XII, 23, 39. 8 Such. Gram. p. 74 ; Zt. I, 569. PHONOLOGY 79 30. 9 BEFORE N MOUILLEE. Suchier^ treats this point. There are, apparently, only two words of importance to be discussed, cognitum and LONGE + s. In queinte, the uei is kept (u going with the q) ; I note an additional example of queinte in Pyramiis, St. Edmond, 1. 934 ; lueinz does not occur but may be supposed from luein (Camb. Psalt. IX, 21). 31. PRETONIC 0. We have already noted (p. 66, § 21) the tendency of some of our earlier texts to keep o pretonic, and have seen that they may even substitute this o for an original u <^fj. For pretonic o we find the usual variants, u and ou, before both oral and nasal consonants : cf . in the Arundel Pscdt. for exam- ple, oureilles, bounourez.^ The variants i and e are likewise to be noted (connisez, apresmer).^ For the fall of the pre- tonic in hiatus in verb forms {oust, etc.), cf. p. 60, § 17, 1, a. (These remarks apply equally to q and o, since ^ becomes o in pretonic position.) IT 32. 1. U : AND U : I. There are two points to be noted here, the rhyming oi il-.u (p) and that of il : i. We have already considered the former (p. Q>Q) and referred to the phonetic value of the u uei, ui), just as we have original o in free position reflected in the o of our early texts (cf. above, p. 70, § 24, 1) ; in some monuments oi occurs for ui < OJ, but not for ui < uj, and these same texts do not confuse u (o) and u (u) ; this, of course, points to a histori- cal oi (for oj) ; later texts, however, doubtless use oi or ui indiscriminately.^ 3. UE. This orthography (for ui) is particularly note- worthy in Angier ; some examples are pues, puessent, nuet, mues (modios ; ^ this ue can be reduced to u : pusse, pussent. The word proximum numbers pruesme and priisme Oxf. Pscdt., Q. L. R.) among its derivatives; these and other variants of the same word occur in general French, how- ever, and the ue of pruesme may not represent the same phenomenon as that recorded for Angier.* CONSONANTS The only general tendency, if we may call it such, to be noted in connection with Anglo-Norman consonants, is the fondness of the scribes for double consonants ; the doubling is not confined to consonants of any one class, liquids, den- tals, palatals, or labials, but all are treated alike and sub- jected to the process, regardless of etymological justification. Stimming gives examples and texts ; ^ many instances are to be found in Angier : ^ soccors, achettee, obbedience, relligion, 1 Mall, p. 64 ; Such. Gram, p, 35. 2 Cf. Such. Gram. p. 59 ; Vising, ^tnde, p. 86 ; Paris, Vie St. Gilles, p. xxxi. 8 cf^ Meyer, Bom. XII, 196 ; Cloran, Dialogues, p. 50. * Cf. Such. Gram. p. 60. ^ Boeve, p. 240. 6 Meyer, p. 198 ; Cloran, p. 52. G 82 MANUAL OF OLD FRENCH DIALECTS reddevance, douzze, etc. I note in Arundel Psalt., pleinne, orreilles, tuitte;^ in Brandan, frerre, L 85; in Sardenai, 239, chappele; in Apocalypse, 111, dlttez. Vitte has an explana- tion of its own (cf. below, p. 93, § 39). These double consonants have usually been considered to possess no espe- cial phonetic significance, but recent investigations go to show that they may reflect an actual pronunciation of French consonants by the English.^ M AND N 34. We treat these two together because about the only thing noteworthy in connection with them is their inter- change, and this is not at all peculiar to Anglo-Norman, though very frequent there ; such frequency has always attracted the attention of editors, and they have studied the phenomenon from the point of view of the kind of con- sonant following the m or the n ; the kind of vowel preced- ing the m or w ; the bearing of the history of the m, n on the question of the nasalization of the vowels in Anglo- Norman, etc. Consequently some notice is here necessary. 1. Interchange. In Anglo-Norman we must not expect the regular French rule of m -f dental > n\ n -\- labial > m to be carried out consistently ; there exists no apparent regu- larity, n, for example, occurs before labials in many texts, from the very earliest, in Lois Guillaume, XLVIII, we have meyibres, in Oxf. Psalt., enblancet, etc.^ Before /, Anglo- Norman has a fondness for m : ^ emfant, gumfanun, etc. 2. N AFTER R. The history of n in words like jurn, 1 Zt. XI, 526, 516 ; XII, 46. 2 I refer here to the researches of Morsbach, published in Beitrdge zur Roman, unci Eng. Philologie. Festgabe filr W. Foerster. Halle, 1902, pp. 324-330. Cf. liom. XXXI, 618 ; cf. below, p. 80, rr, and p. 99, cc. From the last reference we shall see that the suggestion of a phonetic value for the double consonants was not new in the case of cCy at least. 8 Examples, Stimming, Boeve, p. 215. * Ibid. p. 216. ^^ OF THE UNIVERSITY OF C^l_ip(*Rl PHONOLOGY ^^^^^.-^-^--^ 83 emfern, has attracted the especial attention of editors. It is, naturally, of most importance in poetical texts. In the earliest of these, the Cumpoz, it already shows signs of dis- appearance (enfer-.fer), though by the side of such rhymes, examples occur proving the pronunciation of the n, and it seems to have preserved its value sometimes even in com- binations like enferns, corns. Editors of texts subsequent to the Cumpoz, however, agree that this n is silent in their texts, having no phonetic value, even when written (as is often the case).^ 3. M AND ]Sr Final. A distinction between these two is observed only in the Lois Guillaume and the Oxf. Psalt., among our texts.^ 4. M AND jS" and the Nasalization of Vowels. Some reference to this important subject seems necessary; as in general French, the difficulties are encountered in the early stages of the language. We introduce a note on the ques- tion here because the interchange of m and n is supposed to have a bearing upon the point, and because m and n have been spoken of as having varying values after different vowels. For example, Koschwitz,^ in speaking of mum pecchiet* Oxf. Psalt. XXIV, 12, says the appearance of m for n before a labial, which often occurs in Anglo-Norman (he here cites examples), proves that in this dialect m (n) does not form a nasal combination with the preceding vowel. Again, Mall,^ speaking for the Cumpoz, says : m and n are kept distinct (forms like emposet, 730, are rare) ; the only exception is in the rhyme ums : wis, om : on, and he ex- plains this exception by saying that nasalization was very old in {ou)-\- miiii).^ 1 Cf. for example, Walberg, Best. p. Ivi ; Paris, Gme. de Berneville, p. xxxii ; Stimming, Boeve., pp. x and 216. 2 Such. Lithlt. XXII, 121. » Zt. II, 488. * Mum is one of Meister's corrections ; cf. Meister, pp. 108 and 118. 5 p. 76. 6 Cf. Walberg, Best p. Ivi. 84 MANUAL OF OLD FRENCH DIALECTS These two statements, the one proving the status of the vowel from that of the consonant, the other that of the con- sonant from that of the vowel, mutually complete each other. The idea of Koschwitz seems to be that in a nasal combination the separate identity of m as a labial and of n as a dental would be lost: if mim x)ecchiet had been pro- nounced mun pecchiet, the speaker or hearer would have been aware of no abrupt break from the mun to the pecchiet ; but mum pecchiet proves that the pronunciation was at first m?i-?i pecchiet, and, to avoid having a dental (?i) followed directly by a labial (p) the n was changed to a labial (m). This change of n to m to suit a following consonant indi- cates that m and n had their distinct values, which they would not have had if the preceding u had formed a nasal combination with the n. This is further illustrated by Mall's example, where such a loss of individual values is illustrated by the rhyming of urn : un, in which we have to suppose a nasal value for the ic. Thus the two statements fit into each other. The sepa- rate identity of the consonant in the first case proves non- nasalization ; nasalization in the second case explains the non-individuality of the consonants. After all, however, the two writers meant to assert two things, Koschwitz, non-nasalization. Mall, interchangeableness of m and n. The first proves non-nasalization by interchangeableness; the second, interchangeableness, by nasalization. This is about as far as a study of the question of nasali- zation in the writings of our editors will lead. A general presentation of the point is given by Suchier.^ The con- vincing statement for the early history of the nasalization of the vowels, not only in Anglo-Norman, but in general French, is yet to be formulated. Maybe the solution is to be discovered in Anglo-Norman ; there is a great advantage in beginning the study with that dialect, because we know that nasalization disappeared there toward the end of the 1 Gram. p. 6L PHONOLOGY 85 thirteenth century/ only a century and a half after the date of our first texts ; the circumstances of the disappearance of nasalization so soon after its appearance ought to throw some light on the manner of that appearance; principles thus obtained might apply to general French. 5. Miscellaneous. Stimming ^ considers the following points : — a. Confusion of n and I (cdne). We may add a similar confusion of n and ?•; for example, in Vie Gregoire occurs joevres { = jeunes) lines 958, 1552, and in Vie St. Thomas III, 87, 88 we note a similar jo(ii)vre. b. Confusion in orthography of n and gn : digner = diner. We note an example of this in Sardenai,^ moigne (the regular moine is the rule, however). c. Loss of etymological w .* covendra. d. Insertion of an inorganic n : ensi, boins. In Arundel Psalt. we note seigne (= seige for siege).^ I consider below the Anglo-Norman form verms (p. 114, § 54), and the peculiar ending -dnie, < n -f j (p. 90, § 37, 3). 35. L (BEFORE A CONSONANT). I have little to note here that is remarkable. There is no well-defined usage confined to any particular period. If we should endeavor to divide the examples from our texts under the headings 'I remains,' 'I falls,' 'l> ic^ we would find one and the same text often illustrating all three. Neither will a division of the examples according to the vowel preceding the I justify itself. The state of affairs in the older texts is the following : previous to Brandan, I is kept quite consist- ently : Cumpoz, I remains ; ^ Bestiaire, I remains as a rule, falls in the group -e?s;^ in Lois Guillaume, where one Ms. 1 Cf. Stimming, Boeve, p. 218. 2 Boeve, pp. 216-218. 3 Rom. XIV, 88. ^ Zt. XII, 46. 5 Mall, pp. 65, 77. ^ Walberg, pp. liii, liv. 86 MANUAL OF OLD FRENCH DIALECTS shows I > \i, the variants from all the other Mss. will show the I to be kept. There is only one word in the whole text with I > u for which no variant showing the I kept is given ; ^ Oxf. Psalt., I remains, but a weakening is evident because of the presence of a glide e, chcUemeals. Before flexional s, I is sometimes lost, though examples showing this are in a minority;^ Camb. Psalt., I remains; in rare cases it falls or > ?* ; ^ Q. L. 11., I remains. Schlosser studied in detail'* the history of the I after various vowels. The exceptions are autre, autel, and/;<;; Oxf. Roland, I is kept; glide e appears exceptionally. AVith the Brandayi the weakening of the I becomes the rule ; ^ by the side of heals, oiseals, that text shows hens, oiseus. Gaimar knew the vocalization.^ In Angier the I can remain, fall, or become u.'^ Chardri and Gme. de Berneville both illustrate the vocalization > u.^ 1. L > U. The earliest known examples of ? > u in Anglo-Norman occur in the Domesday Book (date, 1086) in the proper names Bauduin, Tethaut.^ There is little doubt but that I had the value of u in many cases when retained in orthography, and editors of early poems are sometimes constrained to say that while they do not find rhymes attesting the vocali- zation of I, neither do they find rhymes proving that the vocalization might not have taken place.^" It is evident that vocalization of I > u was known very early in Anglo- Norman; it became common toward the end of the thirteenth century. 1 Such. Litblt. XXII, 12L 2 Harseim, p. 320. 8 Schuijiann, p. 43. 4 pp, q, 21, 28, 38, 39, 48, 53. 6 Forster, Zt. I, 5G5. e Vising, Etnde, p. 87. f Cloran, p. 52. 8 Koch, pp. XXX, xxxili, xxxvii ; Paris, p. xxiv. ^ Zi. VIII, 361, § 39 ; for the earliest examples in French proper of. Mom. XVII, 428. 1'^ Cf. for example. Vising, l^tiide, p. 77 {Brandan)) Walberg, Best. p. liii. PHONOLOGY 87 2. L Lost. The simple loss of the I is likewise attested by frequent examples from the time of the earliest texts ; ^ the two phe- nomena, I > It and / lost, seem to have been present side by side, with little or no time difference in the process of their change. We have an indication in the Orthograpliia Gcdlica that I became u when I was preceded by a, e, or o ; this coin- cides with the fact that the earliest examples in rhyme for the loss of I are when the I was preceded by i or u. The rule of the OrtliogrcqMa Gallica doubtless represents the original state of affairs.^ 3. Final L > U (or Lost). Final I preceding a word beginning with a consonant may become u. This is an extension of the vocalization of the I in the body of a word, appears later, and is at first confined to i^roclitic words like del and al ; it is frequently found in teu, queu (tel, quel) too. Stiirzinger gives references.^ We may add Vie St. Thomas, teu pes (I, 56), ceu jour (III, 40). This teu is used with feminines both in Chardri and Vie St. Thomas (teu joie, IV, 85). We note also nu = ne le, Bible Fragment, 854 ; eu — en I, Apocalypse, 378, 623. In the Apocalypse we find the u before a vowel too, solau oscures, solau e, 413, 430. The same analogy (to I + consonant in the interior of words) will explain the occasional loss of final /, particularly in unaccented words, a, de, ne, and the like (for al, del, nel).* 4. Miscellaneous. a. Confusion of I and r. The confusion is frequent from the time of the earliest texts. Cf. Bestiaire, nature : nule, Gaimer, Contraire : bataille. Vie St. Gilles, apostorie : his- torie; Apocalypse, itel : mer (line 1228).'' 1 Reference in Stimming, Boeve, p. 211. 2 See the full note and references of Stiirzinger, Orth. Gall. p. 50, VII, T. 11. 3 Orth. Gall. p. 50. * Cf. Stimming, Boeve, p. 212. 6 Cf. Walberg, Best. p. Iv; Stimming, Boeve, p. 212. 88 MANUAL OF OLD FRENCH DIALECTS b. Z + / > r. This development is noted by Paris ^ not only for original I as in apostolie ; apostorie; cited above, but even for the I < d, as seen in the proper name Gidie, Gilie, Gile, Gire. c. Inorganic I. There are a few examples of such an I both before consonants and final, as voils = voix? In Arun- del Pscdt. I note felidnie, elnemis.^ 36. L AND N MOUILLEES. 1. Loss OF Palatalization. We have to note the loss of the palatalization of the I and n in Anglo-Norman. The fact that the palatalization can disappear not only when V and 7i' precede consonants, but Avhen they are intervocalic or final, and that the I from V can become u just like original / before a consonant — these facts point to an early date for the loss. It became general in the course of the thirteenth century ; we must not suppose, however, that the palatalized sounds were forgotten by our writers, for such was not the case.* 2. Orthographic Variants. It is interesting to note the numbers of variants for V and n' that appear in the Domesday Book.^ For V we find ilgi, ilg, illg, il, ill, Uli, ly, Ug, Hi, II; for n', ingi, ing, inc, ini, ign, in, inn, gn, ngi, ng, ni, im. This list might, apparently, embrace all varieties, but it does not include the ignn of the Q. L. R., nor the iUl of the same (the latter is an evident mistake, however),*' nor the igni and nni of the Camh. PtscdtJ Doubtless, a com- plete list for Anglo-Norman would offer additions to the 1 Gme. de Berneville, p. xxxii. 2 Cf. Stimming, Boeve, pp. 211, 212. « Zt XI, 524, 525. 4 The best general statement and references on the point are tliose of Vising, Etude, pp. 77, 78, and 87. He here speaks of the confusion of palatal and dental sounds in Gaimar and Brandan (cf. for the latter, Zt. IX, 100). For the Cnnipoz and Bestiaire, cf. Walberg, p. liv (no examples of n' : n however, p. Ivi) ; Gme. de Berneville, Paris, p. xxxii ; Boeve, Stimming, pp. 212 and 218. 5 Zt. VIII, 361, 3G2. ^ Schlosser, pp. 54, 61. "^ Schumann, p. 45, PHONOLOGY 89 above. Nevertheless, the favorite orthography for V and n' in our texts seems always to have been simply I and n, or il and in. 37. Several interesting phenomena present themselves in connection with a study of this consonant. 1. n AND RR. The confusion of single and double r is noteworthy. Faulde pays especial attention to it in his article on gemination.^ The grammarian Beza warned peo- ple against the carelessness of the Norman writers in this regard, and the carelessness is strongly reflected on Eng- lish territory. Faulde indicates that the earliest instances of the simplification of rr occur when rr was preceded and followed by e. He cites cases from Bestiaire, Oxf. Psalt., and Oxf. Roland. Meyer-Ltibke ^ speaks of tr > rr after the accent as especially Anglo-Norman, though he adds at once that r is found too, and already in the Cumpoz {"jnere: amere," 1. 2745^) ; Stimming'' gives examples and text ref- erences ; Andresen gives instances from Amadas et Ydoine, and refers to others in the Anglo-Norman Mss. of the Eeim- predigt, and of the Lois of Marie de France.'^ Walberg,« for the Bestiaire, decides to adopt r in all cases except for- eign words and the future of seeir (to avoid confusion with the same tense of estre). We have already referred to the Arundel Psalt., orreilles (p. 82). It seems fruitless to endeavor to make any definite state- ment for the r and rr. To divide our examples according to the place of the accent, or according to the etymological background of the r (as r, rr, tr, etc.), or study the futures (in which so many cases are met), — all leads to but one con- clusion : we are in the presence of another instance of mere confusion. 1 Zt. IV, 547. 2 Oram. I, 446, § 405, 490, § 548. 3 This is an incorrect citation ; the rliyme is pieres : manieres. 4 Boeve, p. 214. ' Zt. XIII, 86. « p. Ixxxix. 90 MANUAL OF OLD FRENCH DIALECTS 2. R BEFORE A Consonant. The loss of the r when preceding a consonant is often to be noted in Anglo-Norman, and proves the weak pronunciation of the r: reto{i')ner, se(r)fs, etc. Stimming gives references and examples.^ We cite, in addition, the following : in Sardenai, 447, occurs the rhyme dame : arme (where arme = anima) ; in Apocalypse, is oscurz : puz ; in the same, 25, we find nurves : pointures. Anglo-Norman thus adds its quota to the many indications we have of an early, general weakening of r -\- consonant throughout Old French.^ 3. K -f I. Here we must call attention to the endings -arie, -erie, -one (and similarly -anie, -enie, -onie). These occur in Anglo-Norman as -drie, -erie, -dnie, etc., as -aire, -eire, -aine, etc., and, finally, as -arte, -erie, -anie, etc. The first set, -drie, etc., is represented in English words like primary, victory ; the second -aire, etc., is the regular French form ; the third, arie, etc., occurs in the latter stages of Anglo-Norman (victorie : vie), — Uhlemann specifies shortly before or after the date of A2iban.^ 4. R AFTER A Consonant. a. e inserted (faiiderai, etc.). This point has already been discussed (cf. p. 61, § 17, 2). b. r lost. Editors have cited but few examples for this ; the instances in Gaimar are well known : estre ifeste, ancestre : geste, entrent : dementent} In Vie Gregoire I note Theotiste : epistre, 1. 1673. We have to do mostly with the group -str (cf. below, no. 6, where an inorganic r is inserted after -st). 5. Metathesis of R. This is of very frequent occurrence in Anglo-Norman, though by no means confined to that dialect. 1 Boeve, pp. Hii and 215. 2 Cf. here the important remarks and references of Neumann, Lithlt. VI, 241, f.-n. 2. 8 Cf. on this point. Vising, ^tude, pp. 71, 82 ; Mall, Cinnpoz, p. 55 ; Uhlemann, Auhan, p. 509 ; Walberg, Best. XLV, LXXXV ; Grass, Adam, 125 ; .Aleyer, Bom. XXV, 250. * Cf. Vising, £tude, p. 87 ; Kupferschmidt, p. 418. PHONOLOGY 91 a. -re > -er. The most frequent examples of metathesis in our texts are met in connection with forms of the verb prendre ; pernons, pernez, etc. ; such are found from the time of the very earliest prose and poetical texts; in addition to these, a few cases like ester < esti-e, quater < quatre, are cited.^ I note, in addition, from Melanges, pover ^povre;^ Vie St. Edmond, 882, overs = ovres; and from Apocalypse, liver, 296, 300, 304, etc., by the side of livere; also quater, 310, 329. It looks as if, in some of these cases, there may have been a development such as livre > livere > liver. b. -er > -re. There may be a question whether certain verbs which apparently show a change of conjugation do not rather illustrate a metathesis of r ; as getter > gettre, lutter > luttre (cf. Morphology, p. 119). 6. I^^ORGANic R. Stimming^ cites some exam^^les from Q. L. R., Gaimar, Bozon, and Ipomedon. In a number of cases the r follows t and st {Olestre, destrin, etc.), though philosophre is found. In Aspremont, I note curorne, 80, and entirely different from the above, poreir = pioeir {pouvoir), 7. Interchange of L and R. Here we have only to refer to page 87, where this point has already been treated. 38. We must note the Anglo-Norman development of p + J, which has been discussed as follows : Mall ■* said that the sound of the c < pj was ts, and quoted examples from Philippe de Thatin, Q. L. E., Fantosme, etc. ; such was like- wise the result of Varnhagen's extended study .^ The deci- sion of the question seems to have depended almost entirely upon the rhyming of the word sace (sapiat) with those words 1 Stimming, Boeve, p. 213. 2 j^om. IV, 376. 3 Boeve, p. 215. * Cumpoz, p. 92. 5 Zt. Ill, 177 ; cf. Bom. VIII, 625 (a summary of V.'s results). 92 MANUAL OF OLD FRENCH DIALECTS in which a similar c had an assured value of ts, such as tace, face, glace, etc. Schlosser^ has a long note on the point, in- cluding a review of the discussion : the Q. L. R. represents the c < pj by c, ch, sch, and sc (last two seldom) ; the ch and sch, he says, indicate the pronunciation c ^ for c < pj was not unknown to the scribe of the Q. L. M. If such is true of that text and of the Camh. Psalt. sace,^ it was doubtless true of later texts ; c is further indicated by English words like approach ;^ consequently we must suppose that Anglo-Nor- man knew both ts and c for the c < pj, the older value being ts. Walberg ^ cites further examples of c having the value of ts, saying that sace : glace is a Norman rhyme. T, D 39. T AND D INTERVOCALIC. Our texts show that Anglo-Norman, so swift in many of its phonetic changes, preserves some of the rare examples that represent the gradual disappearance of t and d, and thus exemplifies the fact that the disappearance was very much later in some portions of French dialect territory than in others.^ The comparatively extensive use of the Anglo- Saxon th in our texts makes it evident that the Normans still pronounced the intervocalic dentals at the time of the conquest, and that this pronunciation was continued in Anglo-Norman for some time, till toward the middle of the twelfth century in any case.^ That the pronunciation th or clh was peculiar to Anglo-Norman is not sure.^ For our earliest texts we note the following : *^ In Cumpoz 1 Q. L. B. p. 63. 2 I ^^se ^ as equivalent to ch in English church. 8 Schumann, p. 50. ■* Stimming, Boeve, p. 235. ^ ^^st. p. Ixvii. « I here refer to Lot's researches, Bom. XXX, 481 ss. ' See here the statement and references of Kluge and Baist, Zt. XX, 330. For useful references, of. Ibid. p. 322. 8 See note and reference by Paris, Bom. XVI, 156. 9 Cf. Meyer-Lubke, Gram. I, 388, § 436. PHONOLOGY 93 and Bestiaire the dentals fall, with isolated exceptions.^ In Alexis (L) th occurs very often. In Lois GuiUaume, where the scribe omits the intervocalic dental constantly, we have one of the rare examples of the use of the Anglo-Saxon den- tal spirant sign, dt (Jlede).^ In Oxf. Psalt. the dental is kept as a rule, though there are a number of examples of its fall. In past participles, forms with d are predominant; in other words the dental perseveres or is lost with no apparent regularity ; or again, certain words, like vie, crier, ocii^e, never have the dental.^ For the Canib. Psalt. Schumann ^ thinks that, in the original, the fall was well-nigh general ; th occurs but once (benetheit, XXVII, 6). The Q. L. R. shows the same state of affairs as the Cumpoz and Bestiaire.^ We may deduce, from this evidence, that within a sin- gle dialect, here the Anglo-Norman, the fall of t, d took place more readily in some words than in others. This point is studied at length by Koschwitz.^ According to him the earliest cases of the fall are shown by Latin d -f k (these two having been assimilated even in Latin times) ; then d ( < t) + r ; fZ + le ; ai, ei, etc. -\- d. A simple statement is given by Paris, too : ^ the d that disappears comes from d intervocalic, or d followed by r or / (vidrent, Eodlant) ; from i intervocalic (vide, muder) or followed by r or I {podrons). It must have been pronounced nearly like English th in this. It shows signs of disappearance at the epoch of the Roland and does fall shortly afterward. The word vite, which is so persistent, even in texts showing in other cases the fall of the t, d,^ and which is written vitte in the Brandan (1. 76), may be considered a learned word in this form (vite), used in referring to the recountal of the events in the life of a saint.^ 1 Mall, pp. 79, 80 ; Walberg, p. Ivii. 2 Matzke, p. xlviii. ^ Harseim, p, 321. * p. 47. 6 Schlosser, p. 67. ^ Ueherlieferung, pp. 58-60. '^ Extraits Roland^, p. 15. 8 Cf. Walberg, Best. p. Ivii. ^ Paris, Bom. XXIX, 590, f.-n. 1. 94 MANUAL OF OLD FRENCH DIALECTS 40. T AND D FINAL. The state of affairs with regard to the final dentals is very- confusing at first sight. For example: In the Oxf. Psalt. the t and d are kept quite regahirl}^ ; in the original they were probably not always pronounced. We have lists of the verbs in this text showing the exact proportion of the fall or the maintenance of final dentals.^ In the Camb. Psalt., certain words always keep the final, others always drop it, others are irregular ; there is a like inconsistency in verb forms, though in the last two parts of the Psalter a tendency to drop the finals is manifest -, t > d seldom ; we note seven instances of tJi;- I note th in Arundel Psalt., oth (audivit).^ In the Q. L. R. the dental remains as d (seldom t) or falls. Here, too, we have details for verbs and various words. After consonants t remains ; there is one example of th.* The pronunciation of these finals, where kept, depended on the elements following them. Both t and d, unsupported by other consonants, had two values previous to their dis- appearance, th (thing) and th (this), the former before pauses and voiceless explosives, for example ; the latter before vowels, and before consonants other than the voiceless stops. The manner and stages of disappearance of these finals have given rise to much discussion. Mall ^ supposed that final t was always pronounced in one manner — voiceless, and that it disappeared gradually, first after e, i {Cumpoz) next after a (Bestiaire), finally after u. Suchier^ devotes some space to a refutal of Mall's position. He insists on the distinction between the fixed (" fest ") and the shift- ing (" lose ") dental (the t in set < septem is a " fest," the d in /eici < FiDEM a ^' lose" dental). The two do not rhyme together regularly, and must have had different pronuncia- 1 Ilarseim, p. 322. 2 Schumann, pp. 47, 48. 8 Zt. XI, 515. 4 Schlosser, p. G8. ^ Cumpoz, p. 81 ss. ^ lieimprediyt, p. xix ss. PHONOLOGY 95 tions, the one (" fest ") voiceless, the other (" lose ") voiced. The final of amed, venud had the same value as the inter- vocalic d in amede, venude (except that the d of amed may have been pronounced as d, that of a7nede as th). When the d of amede began to fall in the course of the eleventh century, so did the d of amed; just as the writing of inter- vocalic d in the manuscripts was optional in the middle of the twelfth century, so was that of final d (but final d could interchange with t, intervocalic d could not). The relative times of the loss of the finals Suchier thinks to have been the following: First of all, original etymo- logical d fell. Perfects in 4t, smdjlt (fiiU) lost t as early as the eleventh century. Verbal forms in -at (perfects in -at, at < HABET, and futures constructed from infinitives + the present tense of habere), lost their t within literary times ; that is, during and after the twelfth century. Suchier specifies the Norman and Anglo-Norman texts keeping, and those losing, the t after a} Vising^ accepts Suchier's division of the two classes of dentals, and dwells upon the sharp line of demarcation between the fixed and the changeable dental. Walberg^ goes over the whole question, formulating a scheme based on the statements of Suchier, and on those of Paris in the Extraits Roland. Walberg considers the examples under three headings. First : finals originally between two vowels, of which the second fell in French, and the final of atonic -at. Here the t fell as early as the Bestiaire, fei : lei ; examples are furnished likewise by the participial endings -atum, -iTUM. For the termination -at, he makes a count of the rhyme -e{t) : e, of e(t) : e(t), also of the elision of the e of e{t). Second: finals supported by other consonants; here the t remained steadfastly, cidt : twit. Third : finals not supported by other consonants ; here the final dental is lost ; examples 1 See here Stengel's remarks, Litblt. II, 329 ; Meyer-Lubke gives a summary of Suchier's results, Gram. I, 500, § 557. 2 :^tude, p. 89. 3 Bestiaire, p. Ivii ss. 96 MANUAL OF OLD FRENCH DIALECTS are drawn from preterites in -it, -at, and at < habet, also futures in -at < habet. Walberg ^ questions two rhymes cited by Mall from the Cumpoz (preterites in -it rhyming with a fixed t) ; in this he is upheld by Paris.^ The above represents the merest outline of the question as it affects Anglo-Norman more especially. There are many interesting points of detail, even in our dialect, not spoken of here, but which are referred to in the citations given. Without a knowledge of these points the student will expe- rience difficulty in assigning examples from any one text to places imder the three divisions as given above. Because so many verb forms have final t, the investigator has to allow for analogies, crossings, and the like. Furthermore it must be borne in mind that even comparatively definite statements, like those of Suchier and Walberg, indicate only the general lines of development. In the study of some one poetical production, for example, it will have to be remembered that the finals, before completely disappearing, were pronounced or not pronounced according to the exigen- cies of the rhymes ; and for each text the editor has to adopt some consistent scheme of orthography with regard to the dentals found in his text (as did Mall, Suchier, and Walberg,^ among others). 41. T AND D MISCELLANEOUS. 1. CoxFusiON OF T AND D IN Groups. Here we re- fer to Stimming ; ^ we find, for example, veindre, Bedleem ; voutrions, pertriz ; forment illustrates the loss of the t. 2. Final T and D. a. t, d sudlever > sullever. The same {t for z) occurs in Brandan, seet, prenget= seez, ]prengez (cf. below, p. 122, 5). b. Final t > d. d for final t is a common orthographic peculiarity of Anglo-Norman ; mound, mord, etc.^ The con- fusion of final d and t is very old, being found in the Domes- day Book ; - for that matter it goes back to Latin texts just preceding the French period.^ c. Final t > k. This change, found on the continent, ap- pears in several instances in our dialect. Stimming ^ men- tions examples from Boeve : bixmc, renc, etc., and refers to others. For the same in first person singular present of verbs, see below, p. 120, § 62, b. The k represents a mere orthographic change in Anglo-Norman — a fact proved by rhyme.^ To the examples mentioned we may add one from Vie Gregoire, 2121, Everivic : eslit. d. Final t lost. This is the most frequent of all the phenomena mentioned, and applies to t in the interior of the word as well as to final t. Examples occur from the time of the earliest texts ; es = est, ai = ait, osa = osast, etc.® e. Inorganic final t. Such a t was added particularly after n, and in this case, was pronounced: tyrant, pay sant, etc.^ It is easy to see analogy here, and likewise in cases where t is added after s, mist (missum),/s^ (feci) ; but after other consonants, as ert (heri), and after vowels, as frait ( =zferai) the t seems due only to indifferent orthography. 1 Boeve, p. 22L 2 zt. VIII, 360. 8 Cf. Bom. I, 327 ; Zt. V, 45. * Boeve, pp. 221, 222. 5 Cf. Sturzinger, Orth. Gall. p. 52. e stimming, Boeve, p. 222. 7 AlTolKri-nn* 7?>oo/ "Tk 1-viii H n lib. vruLi. p. fj^. (jtiij " Walberg, Best. p. Ixiii. 98 MANUAL OF OLD FRENCH DIALECTS C 42. The obscure points connected with the history of palatal c have little light thrown on them from a study of Anglo-Norman. This dialect not only offers nothing original or peculiar, but it reflects nearly every irregularity or pecu- liarity to be seen in French of the continent. As may be expected, all kinds of orthographies abound. " In one and the same manuscript we find the graphic signs of Central French and of South Norman by the side of those of Picard and North Norman, — a fact explained by the presence in England of immigrants from different provinces of France."^ Koch 2 puts the case even more strongly when he says that in his opinion the sounds k and ch (not to speak of the signs used for them) were adopted inconsistently by one and the same individual. The three sounds, k, ch, and ts were all known to Anglo-Norman writers, and each was represented in several ways. Stimming^ gives them as follows : — 1. C PRONOUNCED AS K. Hcrc it was written k and qu, q being used only before u\ ch too was used: eschorchevj auchun. 2. C PRONOUNCED AS TS. c is fouud here, not only before e and i, but before a and o ; co, comencat, etc. ; sc re- flects the change from ts to s in pronunciation : ascer (for which change s alone is commonly written) ; many examples occur in Sardenai; for example, sel, 110, site 21 [we find the opposite, c = s, in the same text, as eucent, 298, auci, 225, etc.] ; ss occurs in issi (ici) several times. In both early and late texts ch occurs here too, and not infrequently ; feche (feis), pecha (pieca), etc. 3. C PRONOUNCED AS CH. Here c is found quite often in the earlier texts, especially when preceding an a (Domes- day Book, Brandan, Oxf. Roland). However, ch early became the popular way of denoting the sound tsh {Brandan, 1 Meyer-Lubke, Gram. T, 357, § 410. ^ Chardri, p. xxxv. 8 Boeve, pp. 231-236. PHONOLOGY 99 Cumpoz, Camh. Pscdt., Q.L.R.). We find cch, too. Faulde,* noting the consistent retention of cc in peccher and seccher in texts of the twelfth century, in instances where continen- tal texts show one c, says the pronunciation must have been k-tsh. The presentation of the question by Stimming is as de- tailed as a work of the nature of the present one calls for. The palatals in each separate Anglo-Norman text furnish abundant subjects for investigation. For such investigations the students may look for models in works like those re- ferred to in the foot-note below.^ G 43. The history of palatal g in Anglo-Norman is essen- tially the same as that of g in Central French. We have to note several varying orthographies. For example, g for dz, goie, mangue, gamhe; ch for dz, bercher, chambe. To the examples given by Stimming ^ and Walberg ^ we may add some from Arundel Psalt., where we find ch for dz'm vencheur, venchance,^ g for dz in goie,^ and a combination of g and ch in estremgclicmz. As may be expected, we find j for g ; as Camh. Pscdt. XXXVI, 21, jahle = gable, and, in the same text, g before e with the same value as before a, as orgeilhis, XXV, 4. V 44. There is little that is characteristic to note with regard to v. 1. Loss OF V IN Groups and Intervocalic. We have already referred (p. 91, § 37, 4, a) to the insertion of an e in 1 Zt. IV, 553. I note an example of j for ch in Vie Gregoire, 1598, jasquns. 2 Varnhagen, Das c im Oxforder Psalter; Zt. Ill, 161-177 (cf. Rom. VIII, 625) ; Buhle, Das c im Lamhspringer Alexius, Oxforder Bo- land und Londoner Brandan ; Greifswakl, 1881 (cf. Lithlt. II, 441). 2 Boeve, p. 237. * Bestiaire, p. xcii. 6 Zt. XI, 519, 534. 6 ji,ia, XII, 4, 26, 52. 100 MANUAL OF OLD FRENCH DIALECTS tlie group vr. The v of the combination is sometimes lost, as is shown in English words like 2^oor, kerchief, and the like ; several examples occur in Boeve} We have a few instances, also, of the fall of the v intervocalic, though most of them occur in different forms of the verb espoenter? 2. Confusion of V and F. This is common in Anglo- Norman and resembles the confusion of t and d (cf. p. 97). Cases occur as early as Gaimar {nafrer, jofne)? I notice in Aspremoyit, 43, life (= Uve), Apocalypse, 144, 512, escriferai, 452, chefuz (both v and /are seen in servfs, 2). W 45. W was a letter much favored in Anglo-Norman orthography, from the date of the earliest texts. In the Cmnb. Psalt. it is used in six different ways, as shown by the words mvrnement, esivanlerai, ewes, ow (oti), towe (tue), swatume} Similar uses in other texts are referred to by Stimming.^ It will be seen that in nearly all the lo takes the place of the half-consonantal u. This usage is con- sidered as peculiar to Anglo-Norman, arising, of course, under English influence. W is used for vu in Boeve : iclt, wnt, etc., and for v simply in Bible Fragment, wus, 3, 112, 526, 689, etc. J 46. The only j we note here is the one of which Stim- ming makes an especial study ; ^ that is, the inorganic j inserted to break hiatus, as baier (bad are), chaier (cadere), etc. It is written either as y or i. It is peculiar to Anglo- Norman, in that it occurs only before the accented syllable, whereas the parasitic i found in eastern and northeastern France appears after all vowels in all positions. For R -f J, N + J, see under the consonant r, p. 90, § 37, 3. 1 summing, p. 219. ^ p^i^^ Boeve, p. 220. 3 Und. p. 220. * Schumann, p. 38. 6 Boeve, p. 220. ^ ibid. pp. 237, 238. PHONOLOGY 101 47. S BEFORE A CONSONANT We now enter npon one of the most important chapters in the history of Anglo-Norman consonants, and of Old French consonants as well, because the Anglo-Norman developments indicate to us many of the developments for general Old French. We may say, in a summary fashion, that phe- nomena observed in any portion of French territory seem to respond to general tendencies ; only variations, of more or less importance, are peculiar to the separate dialects. Of the dialects, however, no single one is as important, probably, for the general history of the subject as is Anglo- Norman. The main points of the question appear at once when we consider, for example, the English words blame, meddle, forest, which preserve to this day Anglo-Norman tendencies of centuries ago. In the first word (blame) we note complete disappearance of the s (of blasmer), in the second (meddle) we have d substituted for the s (of mesler), m the third (forest) we see the perseverance of the Latin s. ^ The first extended study of the point was that of Kbritz. His work was reviewed by Paris,^ and, as usual, the master's review was, in effect, a wonderfully clear presentation of the whole subject, and subsequent grammarians^ have had little to do beyond accepting the divisions and adopting the conclusions of Paris. The etyma showing s before a con- sonant must be considered in two distinct series : first, s + J, F, v; -f- B, D, g; L, M, N, r; second, s + p, t, c. in the first series s is voiced, in the second it is voiceless. Of especial interest in Anglo-Norman is the history of the s before I and n, and before t. For s + Z we find si, dl, II, and disappearance of s: isle, idle, ille, He. For s + n we 1 Ueher das s vor consonant im Franzosischen, Strassburg, 1885. 2 Bom. XV, 614-623. 8 Cf. Meyer-Llibke, Gram. I, 476, § 529 ; Nyrop, Gram. I, 3ol. 102 MANUAL OF OLD FRENCH DIALECTS find S7i, (In, gn (sgn, mi), and disappearance : disner, didner, digner, (disgner, areinnad), diner. For s + t, we find st, lit {gilt, slit), and disappearance: ostel, osalit (eght, oslitel), otel. It is not necessary to give full examples. Koritz^ cites them from eight early texts (though some of his examples are to be controlled by later editions or critiques of these texts). In addition to the references to be found below, Stimming gives many examples from various texts.^ 48. S BEFORE L AND N. 1. General Remarks on Substitute Letters. The first point to be noted with regard to every combination which we shall have to record as a substitute for s + a con- sonant (dl, dn, gn, etc.) is that in the case of each one there was, for a time, discussion whether the substituted letter had a phonetic value, or represented only the replacing of one silent letter by another. The early investigators took the latter position, as a rule, as we shall see below. Koritz,^ for example, considers the substitute signs as silent substi- tutes, and dates the disappearance of the s + consonant from the time when such substitutes began to be used. This posi- tion was one which might easily have suggested itself to early students and editors (of Anglo-Norman texts) when they confronted the apparent inconsistencies in the Mss. in the transcribing of Anglo-Norman representatives of Latin words having s -f consonant. The earliest texts, both in prose and poetry, contain examples of nearly, if not quite, all possible variants. The Q. L. R., for example, has si, U, dl, sn, dn, gn, nn. There is now no question but that these letters before the I and n signified an actual pronunciation, however unem- phatic and transitory that pronunciation may have been, and that they preserve some of the stages through which s passed before its early disappearance, when in contact with I and ?i. 1 pp, 10-18. 2 Boeve, p. 224 (dl); p. 224 (dn); p. 216 (gn); pp. xl, liii, 225 (st). « p. 32. PHONOLOGY 103 2. DL. Mall spoke emphatically of the d of adne and medler as indicating the silencing of the s. More minute study, however, soon brought out the fact that, since this d for s was present only before the dental liquids, I and n (not before m, for example), tliere must be a phonetic reason for such limitation, and the d came to be looked upon as a step in the complete assimilation of the s to the I (or 7i). The s was a voiced fricative, pronounced, doubtless, with but little energy, and it was but little separated from the voiced stop d, whose value (in adne or medler) may not even have been that of a complete stop, but more like Eng- lish th in this} The explanation just mentioned seems sufficient from a phonetic point of view; nevertheless, others have been attempted. For instance, mesler > mesdler > medler has been suggested.^ [The mesdler here spoken of is probably to be compared with disgner and oshtel, all three being cross- ings in orthography ; that is, oshtel = ostel + ohtel; disgner = disner -{- digner ; mesdler = mesler + med/er.] Again, the following has been proposed, and by no less an authority than Foerster : ^ sl> ll> dl; sn > 7in > dn. His idea encountered but little favor.* 3. LL. The usual treatment of II has been that of a good illustration of consonant assimilation in French ; as such, it matters little whether we adopt the stage si > II directly or si > dl > II. The II occurs in the Domesday Book {Gille- bert < Gislibert) and our earliest texts.^ We have the suggestion, new so far as Anglo-Norman is concerned, that the II may represent an I mouillee, and much 1 Cf. Mall, Cumpoz, pp. 88 and 90 ; Behrens, Franz. Stud. V, 183 ; Schlosser, Q. L. R. p. 73 ; Paris, Bom. XV, 618, 620. 2 Merwart, Q. L. B. p. 8, f.-n. 1. 3 Zt. XXII, 265, f.-n. 1. 4Cf. Zt. XXII, 513; Bom. XXVIII, 145; Jhrshrcht. V, H. 2, p. 291. 6 Zt. IV, 556 ; VIII, 362 ; Bom. XV, 618 ; Schlosser, Q. L. B. p. 72. J 104 MANUAL OF OLD FRENCH DIALECTS is brought forward in favor of this view.^ This question merits further study, and several additional circumstances will have to be considered. For example, there is this much against the old idea of assimilation ; namely, that the final stage of assimilation of two consonants is usually repre- sented by a single remaining consonant; on the other haud, Anglo-iSTorman has a fondness for double consonants, notably ^v> II and rr (cf. above, p. 81). Since the supposed develop- ' ment of /' < si must necessarily be compared with that of n' < sn, the orthographic signs for the two should be com- f pared : we find U and nn ; the latter is sometimes equivalent to n' (cf. p. 88). As II is supposed to stand for V < si, in how far does nn occur for n' <. sn? It is found in areinnad in Q. L. R., which is so rich in variants for s -f- consonant.^ Does nn occur for sn after vowels other than i ; that is, is it used to mark assimilation too, as II supposedly is ? Investi- gation along these, among other, lines may throw new light on the point. 4. DN. The d here is the same as that described under s + / (p. 103, 2), and all references given there apply equally well here. No English words have preserved the stage dn^ corresponding to dl {meddler). 5. GN. Here again we must note that early scholars thought of gn as a mere orthographic variant of sn, with no phonetic value, and they discussed whetlier sn was changed to gn directly, or whether sn was first written as n, and this n then confused with gn.^ Suchier ^ called attention to the fact that in Q. L. R. s is 1 By Walberg, Best. pp. Ixiv, Ixv. Paris makes no objection to the suggestion in his review of Walberg, Bom. XXIX, 590. [On Hne 14 of p. 590, read LXV instead of XLV.] 2 Schlosser, p. 73. 8 Cf. Koschwitz, UeberUef. p. 27, f.-n. ; Schumann, Camh. Psalt. p. 49 ; Neumann, Zt. VIII, 383, f.-u. "^ Zt. I, 429 ; cf. Schlosser's rectification of Suchier's statement, q. L. H. p. 73, f.-n. 243. PHONOLOGY 105 replaced by d or g, according to the consonant following the s; speaking of gn, he says he thinks it equivalent to g-n (that is, both pronounced), but not n'. In his study of the Q. L. R. Schlosser^ said there was no doubt but that the gn < sn was there equivalent to n'. This idea is developed by Paris ^ in his review of Koritz, in which Paris says that g is found for s only before n and after i ; that is, under conditions where ign usually repre- sents n'. If ign for isn occurred in Anglo-Norman alone we might consider it a mere graphic variant ; but it is found in regions quite diverse. It is probable that after i, the voiced s before n developed a sound which in combination with n approached n' (cf. below, p. 107, str > istr). 49. S BEFORE T. At first sight, matters here seem involved. Before the voiceless stops, s surely remained in early Anglo-Norman ; modern English words, like forest, beast, seem to point to its consistent retention ; yet, in later Anglo-Norman, not only did the s certainly fall, but Anglo-Norman orthography illustrates the stages of the fall. We shall treat of these points below. 1. Early Eetention of S. The early monuments do not show the silencing of the s. Such are the C^impoz,^ Bestiaire,'^ Oxf. Psalt.,^ and Camh. Psalt.^ It thus becomes evident that the s had not fallen in France at the time of the Norman conquest, and that the consideration of s before t must be different from that of s before I and n. Paris places Anglo-Norman and Wallonian together as illustrating the keeping of the s before p, t, and c. 2. Fall of S. a. Date. We have a famous example, for a long time supposed to be the earliest to show the fall of the s -f i ; this 1 p. 73. 2 Horn. XV, 619. 3 Mall, p. 90. * Walberg, p. Ixvii. 5 Harseim, p. 323. ^ Schumann, p. 48. 106 MANUAL OF OLD FRENCH DIALECTS is the one furnished by the rhyme est : met, 1. 882 of Phi- lippe's Bestiaire. Walberg ^ proposed to shatter the force of the old example by offering the reading s'en est in place of me{s)t, which reading he adopts in his edition/ commenting again upon it. The new reading has not been accorded uni- versal acceptance;^ the objection to it, however, is based on the sense of the passage, and there seems to be no question but that the old reading is so doubtful that it carries no weight as exemplifying the fall of s -|- 1} The s before t had certainly begun to fall in the early part of the thirteenth century, but we have direct evidence that all traces of the s had not been lost before the latter part of the same century. This testimony is offered by the Ortlio- grapUa Gallica,^ where we find the rule that when s is joined to t, it has the sound of h, and that est, plest should be pro- nounced eglit, pleght. This statement reads almost like a commentary on texts near the OrthograpMa in time, as the Chevalier, Dame et Clerc, where we find 7niht (118), conulit (ISO), Jiht (508), etc., and mushter (8), oshtel (27). b. Stages. The rule of the Ortliographia Gallica just cited, while proving the perseverance of the .9, shows, too, one of the stages through which the s passed before disappearance, — the stage h. Neumann ^ developed this idea, based among other things on the rhymes of German poets who seem to have tried to transcribe phonetically words borrowed from the French. Koritz ^ did not like the idea, but Neumann adhered to it,^ and it is generally accepted now. The 5 cer- tainly did not disappear suddenly, the h was obviously not a mere sign of the lengthening of the preceding vowel, but represents one of the many series of changes s may have 1 Bom. XXVII, 146. ^ Best. p. Ixvii. 3 Cf. Rom. XXIX, 591 ; Jhrsbrcht V, H. 2, p. 291 (Vising). 4 Cf. Vising, I.e., and Bom. XV, G21. ^ pp. 8 and 49. 6 Zur Lmit- unci Flexionslehre des Altfranzdsischen, Heilbronn, 1878, pp. 106-109. 7 Svor Cons. p. 34. ^ LithU. VI, 243. PHONOLOGY 107 undergone before disappearance. Neumann himself does not claim long life for the h stage.^ 3. English Words with ST. li s -\-t did undoubtedly disappear in Anglo-Norman, how are we to account for English words like feast, forest, where s is kept ? For this we have two suggestions : that of Behrens,^ that the quies- cence of the s was gradual and not completed till the four- teenth century ; and that of Paris/ that the English borrowed such words from French before s was silenced, or else that the pronunciation of the s was weakened (" ebranlee ") at one time, and afterwards strengthened again (" raffermie "). 4. STE,>ISTE,. It is worth noting in connection with the keeping of s -f ^ in the early texts and in connection with the supposed palatalizing influence of the s on a following I or n (cf . above, p. 104) that Horning in his suggestive note on s -f consonant > ?/ * explains cheveistre < capistrum in the Gamh. Psalt. as follows : in order to facilitate the enun- ciation of the group, the s was pronounced softly and thus developed a y out of the group. 5. Inorganic S + T. It goes without saying that we find examples of s inserted before a t where it has no ety- mological right, and from the time of our earliest texts {mercist, nuist, toste, etc.).^ Cf. below, p. 121, 3, b where this s appears in certain verb forms. GENERAL SUMMARY (48-49) As has already been brought out, the date of the fall of the s before liquids, and that before t are quite independent of each other. In the case of the liquids, Anglo-Norman preserves in its orthography some of the phonetic phenomena 1 Cf. here Paris, Bom. XV, G21 ; Behrens, Franz. Stud. V, 183, and the confirmation of the theory presented (incidentally) by F. Wulff in Mem. Fhilol. present, a G. Paris par elev. sued., Stockholm, 1889, p. 256. 2 I c. 3 Bom. XV, 622, f.-n. * Zt. XXIIII, 414. s References and examples, Slimming, Boeve, 227, 228. y 108 MANUAL OF OLD FRENCH DIALECTS attendant upon the fall before I and n ; consequently we may not say more than that the quiescence of the s before I and n was taking place in French at the time of the Norman Con- quest. In the case of the other liquids, and the other con- sonants of the first series formulated by Paris (cf. p. 101 -), the fall had taken place before the Conquest. Before t the fall is much later, instances are found for the first part of the thirteenth century ; a trace of the s (I refer to h) is present in the latter part of that century, and the general fall was probably not complete till the fourteenth century. The strong hold that st once had in Anglo-Norman is reflected in English derivatives like feast and tempest. 50. S INTERVOCALIC. Ano'lo-Norman writers were inconsistent in their use of s and ss, employing them indiscriminately for the voiced and voiceless s. A favorite method of indicating voice- less s was by the use of sc ; for the same purposes we find z and c, though both of these occur for voiced s also.^ 51. S FINAL. It is very evident that final s and z may have the same value in Anglo-Norman (that of s) from the earliest time.^ A remarkable exception is presented by Guillaume de Berneville, who observes an absolute distinction between final s and z, while all traces of t and d final after accented vowels are lost. For example, -as, -es, -eis, -us never rhyme with -az, -ez, -eiz, -uz.^ Other similar cases are cited by Stimming,'' who supposes that the old value of the z (that is, ts) may have been preserved in some instances, as is evident from English proper names containing Fitz. 1 Examples and references, Stimming, Boeve, 224, 228 ; cf. Zt. XIII, 8G. 2 Some general references here are, Stimming, Boeve, pp. 226, 230 ; Walberg, Best. pp. Ixiv, xcii ; Meyer-Lubke, Gram. I, 505, § 561 ; II, 239, § 173 ; Vising, ^tudej p. 95. 8 Paris edition, p, xxxii. * Boeve, p. 230. PHONOLOGY 109 1. Miscellaneous. Several orthographic confusions may- be noted here. In the Arundel Psalt., where s follows a consonant, we have both s and z written : oilsz, jursz, que- ransz,^ etc. ; many more examples might be cited ; we find also seiesz (verb),^ and -stz in mercistz, enlistzJ'' In the Camb. Psalt. we find an example of x for s in reix, V, 1. The d for s in the same text is a variation of the d for z already referred to above (p. 96) : Jlechisums nod genuilz (XCIV, 6). We note one example of t for s in Vie Gregoire, vert (1. 2017). 1 Zt. XI, 517, 519, 521. 2 jj^ia. XII, 9. 8 Ibid. XII, 9, 10. MORPHOLOGY DEFINITE ARTICLE 52. 1. General. In the Psalters and the Lois GuiUaume we find the regular Old French forms. Li is not axDOCopated in the Psalters {Oxf. and Camb.) ; it is in Q. L. R. and Lois Guillaume, though only in the construction Vum (or its vari- ants, rem, etc.).^ 2. Accusative for Nominative. The first examples of le and les as nominatives that I have noted in our texts are in Chardri, where they occur by the side of the regular li, which is used with certain restrictions, le and les being the more frequent.^ In Auban, too, the latter are found, though not as frequently as liJ On the use of the nominative article with an accusative noun, see below, p. 113, § 54, 1. 3. Confusion of Masculine and Feminine Articles. The use of the masculine le for feminine la is old in Anglo- Norman, dating from the Bestiaire^ {le gambe, le allegorie) and Q. L. P., and becoming more and more frequent toward the end of the Anglo-Norman period. The use of the le seems to have arisen in those cases in which the vowel (of la) could be elided. However, detailed study of this point will doubtless reveal irregularities for whose origin no such explanation exists. 4. Lu. Two interesting forms of the article are hi and lui. The older of the two is doubtless lu, variant of the accusative le, though it is difficult, in some cases, to distin- 1 Schlosser, p. 43 ; Matzke, p. 41. ^ Koch, p. xxxviii. 8 Uhlemann, p. 621. ^ Walberg, p. Ixxiv. 110 MORPHOLOGY 111 guish when la is an article and when a variant of the pro- nominal form lui. Lu is found in the Arundel Psalt. : en lu seignur,^ in Quatre Livres des Rois,^ in Fantosme (1. 290), in Brandan,^ and it is made the subject of a special note by Meyer in connection with Angier.'' It is found in Vie St. Thomas, too (II, 34, il e lu mi), and often in the Pderinage. This lu has been noted especially as the equivalent of del, that is, the genitive. It occurs in Auhan (62, as nosces lu her, 991, le cors lu martir),^ often in the P^lerinage (882, la femme lu rei, 852, la Jille lu rei), and a number of times in the Vie St. Thomas (III, 19, 118 ; IV, 100). There is, prob- ably, nothing distinctive about such nsage, it corresponding to general Old French, as nosces le her, etc. 5. Lui. Even if lu be the original form, it was never used with the frequency which we note for lui. Lui can be graphic variant, in Anglo-Norman, of either lu or li (cf. p. 80, § 33, 1). Consequently we lind lui as nominative and accusa- tive singular, and nominative plural. Examples are given by Stimming,^ to which are to be added some from the PUe- rinage (788, la gent lui rei), from Angier^ {lui vilein), and from the Apocalypse (65, p. 202, lui dreins) . It will be seen that most of the texts cited are of a comparatively late date, Gaimar and Angier being the exceptions. Stimming cites instances of lui as feminine singular (from two texts). 6. Miscellaneous. In addition to lu, mentioned above, we note the following forms from the Arundel Psalt. which are not cited from any other text, as far as I know : leu {en leu cuer) ; lau {lau tue glorie) ; leis {leis 2yeissuns).^ In each of these cases the extra vowel in the form of the article seems to anticipate the corresponding vowel in the word following the article. 1 Zt. XII, 21. 2 Ed. Michel, p. 434. 3 Brekke, p. 48. * Bom. XII, 199. Ji Uhlemann, p. 577. ^ Boeve, p. xii. 7Cloran, p. 57. » zt. XII, 31 ; XI, 529, 620. 112 MANUAL OF OLD FRENCH DIALECTS ADJECTIVES 53. The only point we have to note here — beyond re- ferring to forms like vifs (whose history is the same as that of nouns like colps, p. 113, 2) and beyond recording that in Anglo-Norman we must not expect a masculine or feminine noun to be preceded or followed consistently by the cor- responding masculine or feminine form of adjective or par- ticiple — is the early appearance of f eminines of the graxdis, FOLLis type with an -e. The Cumpoz, Oxf. Psalt., and Bran- dan show fole, forte, sueve, ardante, etc., by the side of the older fo7i, ardant, suef, etc.^ The new feminines abound in Angier : quele, tale, ■passante, etc.^ For the form tuit used in the singular as well as plural, of. above, p. 68, § 21, 4. In Vie St. Edmond, 2930, we have one of the rare examples of beleisur.^ NOUNS 54. We shall treat here of the substitution of accusative for nominative ; of the peculiar forms ne/s, versus, etc. ; and refer to some miscellaneous points. 1. Accusative for Nominative. The origin of the reduction of the declension from two cases (nominative and accusative) to one (accusative) is usually associated with Anglo-Norman; the justice of the tradition is derived, not from the non-appearance of the phenomenon at an early date elsewhere than in Anglo-Norman, but from the consist- ency that marks the reduction in Anglo-Norman. To ilhis- trate our meaning, we find that in Norman, for example, traces of the change are to be detected already before the year 1180 in Roman du Mont Saint Michel, and the change is marked in Guillaume le Clerc (first part of the fourteenth century) ; but, on the other hand, it is never met with in AVace (1125-1174)."* In Anglo-Norman, on the contrary, 1 Cf. Mall, p. 100 ; Hammer, p. 114 ; Brekke, pp. 40, 44 ; Meyer- Lubke, Gram. II, 80, § GO. 2 ^^jw. XII, 199. 8 Cf . Meyer-Lubke, Gram. II, 93, § 66. « Paris, Vie St. Gilles, p. xxi. MORPHOLOGY 113 examples are found in an iminterrupted succession; for example, the Ctimpoz,^ Lois Guillaume,^ Oxf. Psalt.,^ Camb. Psalt.,* Brandari,^ Angler,^ Chardri/ Vie St. Thomas,^ Boeve,^ A7nadas,^^ Chevalier, Dame et Clerc,^^ Auhan,^" Donner,^^ etc. The nature and manner of this change are studied by Koschwitz^^ for the Oxf. Psalt. Mall, too (I.e.), studies irregularities such as nominative article or adjective with accusative noun. We must not forget that the opposite process of the one referred to occurs; that is, the use of nominative for accusative.^^ For the time order of this phenomenon in Anglo-Norman as compared with other dialects we have several statements.^^ Brekke^^ attempted statistics to show that the author of the Brandan was more careful than Philippe de Thaiin in his use of the cases ; Brekke resents having his author classed with Philippe in this connection. A similar line of inves- tigation, that is, trying to draw deductions as to purity of language from the comparative frequency with which two given authors avail themselves of a given phenomenon, has never been followed to any extent. 2. Nefs, Sacs, Colps, etc. These, and similar forms, are very frequent in Anglo-Norman; for example, Oxf. PsalQ^ Camb. PsalQ^ Q. L. B.,"^ Brandan,^^ Chardri,^'-^ Boeve,^ and Auban.'^^ They do not occur in the Cumpoz,^ 1 Mall, pp. 97-100. 2 Matzke, p. 1. 3 Zt. I, 569 ; II, 485. * Ibid. I, 569. 5 Brekke, p. 81. ^ Meyer, p. 198 ; Cloran, p. 54. 7 Zt. Ill, 594. ' « Meyer, p. xxviii. 9 Stimining, p. xiv. ^^ Zt. XIII, 86. 11 Bom. I, 72. 12 Uhlemann, p. 613. 13 Bom. XXV, 532. i* Zt. II, 485. i^ Ibid. IT, 488. 16 Meyer-Llibke, Gram. II, 39, § 25 ; Such. Fran^ais et Provencal, p. 163 ; (in) Grober's Qrundriss, I, 638 ; Warnke, Zt. IV, 234. 17 p. 81. 18 Zt. I, 569. 1^ Schumann, p. 46, c. 20 Schlosser, pp. m, 80. 21 Hammer, p. 108 ; Brekke, pp. 25-30. 22 Koch, p. xxxiv. 23 stimming, pp. xiv, 181, 220. 24 Uhlemann, p. 598. 25 Mall, p. 105. I 114 MANUAL OF OLD FKENCH DIALECTS with the exception of verms, nor need we expect to find them to the exclusion of nes, sas, etc., in the other texts. This set of words must be studied in connection with the question of the value of final s, which, apparently, was not stable. The early adoption of the accusative for the nomi- native probably impressed the form of the accusative on the minds of our scribes, and the nominative, under consider- ation, doubtless represented to them the traditional -s (of the nominative) added to this accusative. The scribes iusert nefs, sacs, coJps, etc., even in rhymes (as Chardri, Boeve, Auban), but thereby again demonstrate uncertainty in the use of their own language, because evidence is against the pronunciation of the p, b, v, c -\- s in rhyme. 3. Miscellaneous. I append here some references on the following points : confusion of genders ; ^ neuter nomina- tives, jugement, conseil, etc. ;^ formation of compound nouns ;^ masculine pere type with s ; ^ feminine ^ewr type without s;^ distinction between ans and anz,fils o^ndijilz.^ We note, too, further instances of the uncertainty of our scribes, in that they sometimes omit an etymological post-tonic e before flex- ional s (as colurs = colubras, pers = petras), and again they insert an inorganic e (as coupes — colps, manteles = mantels, etc.).'^ PERSONAL PRONOUNS 55. FIRST PERSON. 1. Je and Joe. Our older texts offer nothing striking here. Already in Chardri je, as tonic form, rhymes in e : pri je : marche, clemant je : cunge, etc.® The form joe is made the subject of a note by Paris in his edition of the Donnei.^ He says that he had usually regarded joe as a variant of jo 1 Meyer-Liibke, Gram. II, 450, § 302. 2 j^jvZ. II, 13, § 8. 8 Ibid. II, 630, § 547. * Ibid. II, 33, § 22. 6 Ibid. II, 32, § 21 (bis). ^ Litblt. VI, 116. 7 Stimming, Boeve, pp. 182, 183. ^ Stlirzinger, Orth. Gall. p. 45. a Bom. XXV, 532, f.-n. 6. MORPHOLOGY 115 or jue, but that in the Donnei we have to suppose a dissyl- labic pronunciation, joe. He refers to joe as an Anglo- Norman form. The same occurs, for example, in the Bible Fragment, line 793,^ and coe is often used in the same text, as 116, 137, 194, etc. (cf. tue, below, § 56, 1). 2. Mi. This form is cited by Vising,^ who says it is equiv- alent to mei {moi) in Fantosme and Adgar. It occurs also in Picard, Norman, and Lorraine, and is doubtless the mi referred to by Meyer-Liibke ^ as derived from mihi. It is late and rare in Anglo-Norman, more frequent elsewhere. 56. SECOND PERSON. 1. Tu AND Vous. An interesting syntactical phenome- non arises here ; that is, the confusion of tu and vous, the two being employed without any apparent distinction in meaning in one and the same sentence. Suchier speaks of this point.'* The form tue occurs in Aru7idel Psalt.;^ if not a corrupt reading, it is worthy of attention in connection with jue, and recalls the early Italian tue.^ 2. Ous. In Donnei, 816, 973, we find jous ( = je vous) and quideus (= cuidiez vous). Similar forms occur in Chardri and Ipomedon. The ous for vous is not uncommon in Old French.^ 57. THIRD PERSON. Here I shall speak of the use of the tonic for the atonic forms, and of the general confusion of forms, singular and plural, masculine and feminine. 1 Bom. XVI, 206. 2 luul V, 68. ^ Qram. II, 102, § 75. * Auban, p. 8. Cf. additional references in Zt. V, 164. 6 zl XI, 521. 6 Cf. Meyer-Lubke, Ital. Gram. § 148. (In Italian translation, § 88.) _ ^ , 7 Cf. Bom. XXV, 533. Tobler's original article on the point is m Zt VIII 496 ; additions to this were made by Behrens, Zt. XIII, 408. Cf. also, Meyer-Lubke, Gram. I, 379, §429; II, 110, §78, and 408, §325. 116 MANUAL OF OLD FRENCH DIALECTS 1. Tonic Instead of Atonic Forms. This phenome- non, quite common throughout Old French, is very marked in Anglo- jSTorman from the time of the earliest texts ; as Cumpoz, Oxf. Psalt., Arundel Psalt., Alexis, etc. : jetent sei, set pastierent, jo toi pri, etc. (It goes without saying that the confusion is not confined to the third person.) Koch, in his edition of Chardri, made distinctions (and deter- mined his readings thereby) according to the position, or rather significance, of the given pronoun in the sentence ; but in this he was criticized by Mussafia.^ In Angier^ we find that the use of atonic forms is accounted a rarity. 2. Confusion of Forms in the Singular. a. Masculine, lu = lui, lu = lo. We have already referred (p. 110, 4) to lu; both lu and li (which occurs here as through- out Old French) are regular phonetic reductions of lui in Anglo-iSTorman (cf. p. 80, § 33, 1). Lu occurs frequently in Chardri and Angier : quant il lu vit, a lu apparut, etc. Any syntactical principle indicating a preference for lui or for lu or for li is not apparent ; Cloran ^ suggests that in the Dia- logues Gregoire lu after prepositions is rare. We must note, too, that lu is used for accusative lo in the same texts in which hi is found for lui. Koschwitz^ suggested a distinction between Anglo-Norman lu and gen- eral French le, saying that lu goes directly back to lo (that is, it is a variant of lo), while le is derived from lo through the stage Id. For is = il, see below (3, a) under plural forms. b. Feminine. I give here some references on the follow- ing points : use of lui, le, and la as equivalents of feminine dative li-, use of le for la, accusative ; ^ for le = ele, see just below, under lei = ele ; el = ele is found in the twelfth cen- tury in Norman and Anglo-Norman.^ For the correspond- 1 Cf. Koch, p. xxxix ; Zt. Ill, 596. 2 cf. Meyer, p. 199 ; Cloran, p. 56. 8 p. 57. ■* UeherUefcmnti^ p. 25. ^ Stimming, Boeve, p. xxii. 6 Cf. Meyer-Lubke, Gram. II, 109, § 78 ; Such. Franqais et Pro- vencal., p. 136 ; Walberg, Best., p. Ixxiv. MORPHOLOGY 117 ing plural els — eles, see below; lei for la is cited by Cloran^ for the Dialogues Gregoire; de lei soccoure, etc. ; a variant of this lei in the same text is le : en le entrerent, etc. 3. Confusion of Forms in the Plural. a. is = il. is occurs for both singular and plural — for il and ils. So far as I know it is peculiar to Angier : qu^is eit, qu'is orent, etc. It occurs in both his Vie and his Dialogues Gregoire.^ b. es = eles. This is given for Angier by Meyer as the feminine corresponding to the is just mentioned; the only instance cited is, deu donst qu'es puessent; Cloran does not cite it for the Dialogues. c. il = eles. This usage is referred to by Suchier ^ as occurring often in Anglo-Norman. d. els = eles. This corresponds to el = elle mentioned above. e. les = lis, dative plural. Eeferences for this peculi- arity are given by Stimming/ who attributes the use (and that of lu7' as accusative) to English influence; the same occurs in Wallonian and Picard, however.^ UNIVERoi OF DEMONSTRATIVE PRONOUNS 58. Here we note the following: eel and cest occur early for the nominative in Norman and Anglo-Norman ; ^ the neuter eel, particularly in the locution puet eel estre is confined, for the most part, to these two dialects ; ^ the neuter ceo occurs often as an adjective in Boeve (ceo trait ur), but this is a peculiarity of texts of comparatively late date ; ^ ist, mascu- line, and iste, feminine, seldom found in general Old French, occur frequently in Dialogues Gregoire.^ 1 p. 58. 2 Meyer, p. 199 ; Cloran, p. 55. 3 Zt. IV, 419 ; cf. Meyer-Lubke, Gram. II, 110, § 78. * Boeve, p. xxii. 5 Cf. Meyer-Lubke, Gram. II, 115, § 83, and Vising, Zt.fr. Spr. u. Lit. XXII, H. 2, p. 25. ^ Meyer-Lubke, Gram. II, 132, § 96. 7 Ibid. II, 135, § 98 ; Walberg, Best. p. Ixxxv. 8 Stimming, Boeve, p. xxv ; cf.Vising, ref. above. ^ Cloran, p. 57. 118 MANUAL OF OLD FRENCH DIALECTS POSSESSIVE PRONOUNS 59. There is little that is characteristic to be noted here ; variations are coufiued for the most part to the forms of the feminine. 1. rp]MiMNE. Frequently we find mei = ineie, tu = tue, which illustrate the loss of post-tonic e (cf. p. 63, § 18, 1, a) ; these forms are frequent in the Psalters, especially the Camb. Psalt. (as XXX, 16, en la tu main; LVI, 7, mei aneme), also in Arundel Psalt., moi oreille, mei ureisun} We have already spoken of the feminine article le for la (p. 110, 3) ; we find the possessives sa and ma replaced by se and me in our earliest texts, as Camb. Psalt. LXI, 11, se owe; Arundel Psalt., se felonie,^ and similar cases in the Q. L. II? 2. Miscellaneous Forms. We note in the Camb. Psalt. mis, tis (XXVI, 4, 9, 16, etc.) constantly where the Oxf. Psalt. has mes, tes. The former shows also meins (XII, 4), mieins (XVII, 34), and nod for nos (cf. above, p. 109, § 51, 1). In the Arundel Psalt. we note mens, meiens, muen, ten.* In Vie Gregoire we note sis (64, 655) and sous (1730, 1741). RELATIVE PRONOUNS 60. Our texts offer nothing extraordinary here; M early makes a place for itself beside qui or chi ; it is found in Cumpoz, Lois Guillaume, the Psalters, and Q. L. E. with gradually increasing frequency. In the accusative ke for que appears as early as the Camb. Psalt. and Q. L. B Chardri uses ke, as well as ki, for the nominative; Koch changed the ke to ki in many cases (in which procedure Mussafia did not agree with him).^ In Boeve, the usual form for both cases is ke. 1 Cf. Zt. I, 569 ; XII, 17, 48 ; XI, 515. '^ Zt. XI, 519. 3 Schlosser, p. 8. " Zt. XI, 516, 533 ; XII, 5, 17. 5 Matzke, p. li ; Such. Lithlt. XXII, 120, f.-n. 2. 6 Koch, p. xxxix ; Zt. Ill, 595. 5 MORPHOLOGY 119 VERBS 61. INTERCHANGE OF CONJUGATIONS. This process is a most frequent one in Anglo-Norman. It is difficult to detect any underlying principle, because there is hardly any change that does not work two ways. In Anglo-Norman, as in general French, the first weak con- jugation {-er < -are) seems to exercise the greatest power of attraction, though, as we shall see, it sometimes suffers losses to other conjugations. 1. -EiR > -ER. We have here a change that is character- istic of our dialect, as compared with French of the conti- nent; it seems to have originated in Anglo-Norman, and about the middle of the twelfth century.^ The change of conjugation affected the infinitive most of all ; four infini- tives were altered the earliest of all, setting the example to the rest; these were aveir, poeir, saveir, voleir > aver, poer, saver, voler.^ We notice voer (for veeir, videre) already in the Arundel Psalt.,^ and the same occurs as voier (itravaillier) in Vie Gr ego ire, 351. 2. -RE, -iR > -ER. This change is later than the one first mentioned, and illustrates the strength which the -er conju- gation had acquired in Anglo-Norman, by virtue of the older change (-eir > -er). We find terier, repenter, attender, and the like, in the course of the thirteenth century.* 3. -ER>-RE. We note gettre, leetre ( -iR. We find donir, demorir, mangir (Otinel, 26), and the like.^ 5. -lER > -IR. The change in conjugation here affects more especially the infinitive, the third plural preterite, and the past participle: lessir,froissirent,Jichi} 6. -oiR AND -DRE. Anglo-Norman, too, furnishes exam- ples of these varying forms. Already in the Q. L. R. we find manoir and maindre.^ 62. PRESENT INDICATIVE. 1. First Person Singular. a. -e. Our texts offer nothing remarkable here, and from the point of view of time no fixed line of demarcation may be drawn between those texts showing -e (as supporting vowel or analogical) and those not having it. It is present in the Brandan,^ while Chardri, later and careless in his grammar, has no -e ^ {pens, merveil, eim, etc.), nor has Aubau'^ nor Guillaume de Berneville.^ Boeve shows -e in the first person present of the first weak conjugation, but usually omits it for the third.^ b. -c. Quite a number of cases of this ending are to be noted in Anglo-Norman, first of all in the Cumpoz,^ Arundel Psalt. (venclie), ^^ and Brandan}^ In Angier it is frequent ; Cloran^^ cites aourc (adoro), comanc, conseilc, arc (ardeo), tiejic, perc (perdo), regierc, sere, etc., also rench and renconch (record) ; in the Vie Gregoire I note renc (1843), conmanc (1950), pramec (1100). In Boeve we have renc}^ In still later texts I note Otinel {tienc, 12), Melanges^* (senk, pleink), Bible Frag' 1 Cf. Meyer-Lubke, Gram. II, 164, § 121 ; Stimming, Boeve, xxix. 2 Cf. Such. Auban, p. 47 ; Zt. II, 343 ; Bom. XXVI, 88, f.-n. 8 Meyer-Lubke, Gram. II, 172, § 127. 4 Meyer-Lubke, Gram. II, 189, § 136. 6 Zt. Ill, 596. 6 Uhlemann, p. 622. 7 Paris, p. xviii. ^ Stimming, p. xxvi. » Mall, p. 110. ^^ Zt. XII, 36. 11 Brekke, p. 55. ^^ p. 59. 18 Stimming, p. xxix. i^ Rom. IV, 376, 377. MORPHOLOGY 121 ment (venc, 825). The extent of this phenomenon and the character of some of the examples indicate that we may not have to do here merely with an orthographic interchange of final t and c (cf. above, p. 97, c). 2. Second Person Singular. a. -z. Meyer-Liibke ^ calls attention to a phenomenon that is encountered particularly in Anglo-Norman texts; in verbs whose stems end in labials, as deveir, mover, or heivre, the s of the second person is replaced by z, as Cumpoz, deiz; Bran- dan, moz, etc. b. -is. The second person in -is is sometimes found, con- sequent upon the substitution in orthography of post-tonic i for e (cf. above, p. Q>5, 3). 3. Third Person Singular. a. Here we have to consider the question of the t (of -et). Merwart^ gives statistics as to what verbs and tenses do or do not show the t in Q. L. B., and makes reference to the discussions of Mall, Paris, Koschwitz, etc. Hammer^ studies the question for the Brandan, and says that his text and the Cumpoz betray the same state of affairs : the t remains as a rule, may fall for the sake of the rhyme, but elision never occurs in consequence.^ For Guillaume de Berneville, Paris finds a dozen examples of the retention of the t (indicative of first conjugation, subjunctive of others).* In a foot-note he quotes from Suchier,^ to the effect that the elision of the -e did not take place in the verse until some time after the t had ceased to be pronounced. b. St = t. The forms dist, fest, vest (vadit, as Vie St. Edmond of Pyramus, line 1362, vest, lest), dust (debuit), cunust, morust, are to be explained by a rule recorded in the Orthographia Gallica,^ which says that in the present and pre- terite an s must be inserted between the vowels e, i, o, u, and the t.'^ The phenomenon is encountered already in Gaimar. 1 Gram. II, 239, § 173, 2 q_ l^ 7^. p. 3. 8 p. 103. * Cf. Mall, p. 109 ; Brekke, p. 56. ^ p. xix. * Beimpredigt, p. xxxiii. '^ p. 8. ^ Cf. Litblt. VI, 116. 122 MANUAL OF OLD FRENCH DIALECTS 4. First Person Plural. Here we refer to the state- ment already made (cf. above, p. 83) as to the early identity of m and n in the endings -urns, -wis (-oms, -ons), which rhyme together already in the Cumpoz. 5. Second Person Plural. -et {-ecT) = -ez. This peculiar form is found in the Camb. Psalt} (devet, pernet, entendet, corned, seied) ; in the Arundel PscdtJ (adoret, aprimet, pabUet); Brandan^ (seet, prenget)) Chev. Dame Clerc, freiet, 553, and F^lerinage (Jiuyiiset, 721). The confusion of t (d) and z has already been considered (cf. above, p. 96) ; it is entirely orthographic in Anglo- Norman, and in that respect peculiar to our dialect. 6. Third Person Plural. We have to note here the ending -ihit (cf. above, post-tonic e, p. 65, 3), and an oxytone accentuation (to be noted later for the imperfect indicative and subjunctive). Examples for this latter I have observed only in the corrupt Apoccdypse ; here in two instances we see Uasphement : habitant (110), portent : seint {^xi^civ^i, 800). Often we find the present ending in -unt, as a rule to rhyme with sunt or unt ; for example, sigyiijiunt : sunt (73), liabi- tunt : sunt (960), etc. (frequently). 63. PRESENT SUBJUNCTIVE. 1. Forms in -ge (augez = aillez, garge = garde, perge — perde, prengez = preniez, etc.). Similar forms are very popu- lar in Anglo-Norman. A few references are : Camb. Psalt} {aherged), Arundel Psalt.'^ (aicge), Angier,^ Chardri,^ Boeve,^ and Apocalypse {auges, 236). 2. First and Second Person Singular. The early appearance of the (analogical or the supporting) -e here is characteristic of Anglo-Norman as compared with conti- 1 Zt. I, 570. 2 p)ia, XII, 15, 23, 48. 3 Cf. Brckke, p. 56 ; LtblL VI, 371. 4 Meyer-Liibke, Gram. II, 19G, § 138. ^ Ibid. 227, § 1G3. 6 Zt. XII, 35. ' Meyer, p. 200 ; Cloran, p. 59. 8 Koch, p. xl. ® Stimming, p. xxxi. MORPHOLOGY 123 nental dialects ; it is esi^ecially prevalent in the first and second persons.^ It is found in the Psalters and the Q.L. R. (cf., on the contrary, the third person deliurt < delivrer in the latter text). The -e is the exception in the Lois Guil- laume,^ and in some later texts, like Guillaume de Berneville and Aicban, it does not appear.^ The examples from Boeve illustrate the distinction between the first and second per- sons as compared with the third; forms without -e are confined to the third person."* 3. Third Persox Singular. We have just considered the late establishment of -e in this person. The history of the final -t is the same as that of t in the corresponding form of the indicative. Some circular combinations result from the absence of the -e after palatal stems ; for example, Oxf. Psalt. cerst {chercher), esculurst (esculurgier), Camb. Psalt. juszt (juger); cerche, juge, etc., occur, too, however.^ 64. IMPERFECT INDICATIVE. The points of interest here are confined, for the most part, to the first weak conjugation. 1. Oldest Forms. These were oive or one, found in the Psalters and the Q. L. R. ; in the third singular the post-tonic -e is dropped, out being the regular form.^ By the side of owe, oue appear oe, ot, oent in the Camb. Psalt., '^ though they become general only in later texts. These are to be consid- ered as further developments out of oive, oue, and not vice versa.^ The -ue, -ut where found ^ are doubtless variants of the -oe, -ot, -one, -out, due to the scribes (cf. above, p. 73). 2. -EiE, -oiE. The introduction of these in place of -oue, -oe, did not begin in western French territory as early as in 1 Cf. Meyer-Lubke, Gram. II, 207, § 146 ; LithU. V, 70. 2 Matzke, p. li. ^ Uhlemann, p. 622. 4 Stimming, p. xxvii. ^ Meyer-Liibke, Gram. II, 221, § 157. 6 Such. Gram. p. 31. ' Ficlite, pp. 24, 25. 8 Cf. Mall, Ciimpoz, pp. 66, 67 ; Lucking, Aeltst. Fr. Mund. p. 210. ® Cf. Stimming, Boeve, p. 189. 124 MANUAL OF OLD FRENCH DIALECTS other parts,^ and consequently it is not a charaoteristic of our first texts. In Q. L. R. -eit occurs {rjuardeit, passeit) but only in five cases as compared with 105 instances in which -out is retained.^ The rujowe {rujir) of the Oxf. Psalt., and the enqiierrout of Chardri illustrate the prevalence of -oue. A favorite method of studying the beginnings of the pro- cess of substitution has been that of noting the rhyming of -oue with -oie (-eie). Difficulties are encountered in several directions : some texts will rhyme the imperfects of a given conjugation only with imperfects of the same conjugation, as Guillaume de Berneville,"'^ for example. Again, on account of the possibility of choosing from variant readings, what one editor claims as a rhyme of -out: -eit may be discarded by another student of the same text. For example, Kupfer- schmidt^ noted the rhyme despendeit : soldout for Gaimar; Vising^ rejected the same. (In another place, however, Vising cites four cases of the rhyme out: eit in Gaimar's Havelok.^ The first detailed study of the point is that of Suchier,^ who considers the usage of five (minor) texts. Evidently the substitution was not universal among poets in the thir- teenth century, though the tendency toward general adop- tion of -eie is to be noted from the beginning of that century and gradually becomes accentuated. Angier uses -ot and -eit in the singular, but -oient always in the plural.^ 3. -EiNT. This ending (for -eient), an example of the loss of post-tonic e (cf. above, p. 63), is very usual in Anglo- Norman : esgardeint, soleint (Chardri, Jos. 1158, Pet. Plet, 474), poreint {PMerinage, 511). Examples abound in manu- scripts executed during and after the thirteenth century.^ In the Apocalypse we find -eint rhyming with ceiiit, forment, 1 Meyer-Liibke, Gram. II, 323, § 258. 2 Merwart, p. 13. « Paris, p. xxxiii. * p. 417. ^ ^ttide, p. 103. ^ Ihid. p. 14. "^ Aiiban, p. 5. 8 Cf. Meyer, Bozon, p. Ixv. » Cf. Meyer, Rom. XXV, 255. MORPHOLOGY 125 and the like. Again, when the e is not dropped we find ensement : diseient, 252, forme7it : cUsoient, 275 ; or the rhyming of preterite and imperfect ; as, crierent : disoient, 289, crierent: teneint, etc. We have already noted a similar oxytone accen- tuation for the -eyit of the present tense (cf. above, p. 122, 6). 65. IMPERFECT SUBJUNCTIVE. Here, again, we have to call attention to the accent on the last syllable (chantassSnt). The phenomenon is by no means confined to Anglo-Norman, but is very frequent there. ^ Under the influence of the accent the -ent may become -ant (Soussmit, Cumpoz ; veissant, Q. L. B.), and then -ont (Dia- logues Gregoire, rendissont, gardesont, etc.^). 66. FUTURE. This tense, in its various forms, illustrates several phe- nomena of Old French in general, as well as of Anglo- Norman. Some of these may be referred to, as follows: tr, dr > rr, crerrat (Camb. Psalt.),^ cf. above, p. 89 ; con- fusion of TV and r, dirrai (Q. L. R}), cf. above, p. 89 ; fall of atonic e after a vowel, envei-rai, cf . above, p. 63 ; fall of same after a consonant, jurrez,^ cf . above, p. 64 ; insertion of e between consonant and ?', fauderai, cf . above, p. 90, 4, a ; pretonic i > e, vesterai, manterai)^ confusion of -ai and -ei, servirei (Alexis, L, str. 99), Camb. Psalt. serei, clianterei, XC, 15 ; GUI, 33 ; in this Psalter we notice a continuation of the -ex in the second person in enseigneres, CXVIII, 171. 67. PRETERITE. Here we note a few isolated points, which we classify under the different types of the preterite. 1. -DEDi. The Oxf. Psalt. offers some of the rare exam- ples which we have for the preservation of the old -ie: 1 Cf. Meyer-Liibke, Gram. II, 388, § 307 ; Stimming, Boeve, p. Ivii. 2 Cloran, p. 60. ^ Meyer-Liibke, Gram. II, 396, § 314. 4 Merwart, p. 9. ^ Meyer-Lubke, Gram. II, 393, § 314. 6 Ibid. 395, § 314 ; Schlosser, Q. L. B. p. 57. 126 MANUAL OF OLD FRENCH DIALECTS cunfundies espandies, espandierent} We note another case in Vie St Edmond, 2524, espandie : lie. 2. -s Preterite. Here are three noteworthy points. First, the influence of the present on the preterite in forms like crienst, repunst, mainst (in the Psalters)} Second, the confusion of mest (mansit) and mist (misit) which is pecul- iar to Norman and Anglo-Norman texts.^ Third, the exten- sion of weak forms of this type (as desis) to regular weak verbs ; for example, garesis (for garis), saisesis (for saisis), etc., in the Psalters and Q. L. P.* In the Arundel Psalt. we note degiierpisis^ (fesis occurs just before this), in Vie St. Gilles, guaresis (3606), in Sardenai (L and 0), traisist (98). The Anglo-Norman here furnishes examples of a phenomenon not rare in general Old French. 3. -ui Preterite. Here we call attention to the remark of Paris ^ in comparing the forms out and ot, plout and p/o^; out and plout represent the normal products, ot and pZo^ the abridged (cf. above, imperfect indicative, p. 123, 1). These perfects, out, plout are to be distinguished from those in -ut (aperceut, but, etc.) which rhyme only with themselves. 4. -ierent > -iRENT. This change, in common with a similar one of -ie to -i in infinitives and past participles, is referred to in our next section (67). 5. -EUMES > -UMES. On the fall of the atonic -e here (cf. above, p. 60, § 17, 1, a). 6. -RENT. For the oxy tone here, cf . the remarks above on the imperfect, p. 125, 3. 68. PAST PARTICIPLE. We find quite a number of examples of the rhyming of -iee (usually first reduced to -ie) and -lej that is, to two 1 Meyer-Lubke, Gram. II, 348, § 272. 2 Ihid. 374, § 289. 8 Ibid. ; Schlosser, Q. L. B. p. 8. « Meyer-Lubke, Gram. II, 375, § 289. ^ Zt. XI, 521. « Ihid. IV, 520. "^ Ome. de Berneville, p. xxxiv. MORPHOLOGY 127 rhyming in -/ {baissie \fie). These are referred to by Stiin- ming.^ 69. MISCELLANEOUS, IRREGULAR OR PECULIAR FORMS. We here ax^pend, in the alphabetical order of their infini- tives, some striking forms of verbs which have been referred to in various places. Avoir. Avant occurs for the present participle in the Psalters,^ as Camb. Psalt. XXXVII, 14, avanz. Comiaitre. Conissiez, coneissiez occur in the Camb. Psalt. and Q. L. E.^ I note a similar example of the pretonic i in Vie St. Edmond, of D. Pyramus, alisum (1660), menisum (1665). Forms with que- are seen in Vie St. Gregoire, as qiienoistras (1723), qenu (1724). Creire. We find crei as third singular preterite in Vie St. Gilles, 1. 3611.'' Dire. We find disum in Cumpoz and Bestiaire.^ In Vie St. Gregoire we note dierrei (1867). Este7'. We have a present, estait, analogical to vait in 'Cumpoz and Brandan,^ and a preterite estout, analogical to out in Camb. Psalt., Q. L. R., and Brayidan? Etre. We encounter several interesting forms here. For the infinitive we find in Vie St. Gregoire istre (1562). In the present tense, we note in Alexis, L (str. 44, e) soi; su for sui is of course common in Anglo-Norman, consequent upon the reduction of -ui to -u (cf. above, p. 80). For the first person plural we find semes in Vie St. Gregoire (2271). The rare esmes is said to be found only in Norman and Anglo-Norman texts.^ Paris cites it for the Vie St. 1 Boeve, p. 202. ^ Meyer-Liibke, Gram. II, 247, § 183. 3 Schlosser, Q. L. B. p. 62. 4 This refrence is given (incorrectly) as line 3615 in Bom. XI, 596. 5 Meyer-Lubke, Gram. II, 2.32, § 169. 6 Ibid. II, 290, §224. "^ Schlosser, Q. L. B. p. 19. 8 Meyer-Liibke, Gram. II, 281, §211. 128 MANUAL OF OLD FRENCH DIALECTS Gilles,^ and says it is almost unknown after the twelfth century. He refers to Burguy,^ who cites several examples of esmes (eimes, ermes). In Vie St. Gilles, line 504, we find the spelling aimes. I note also in the Bible Fragment (11. 125 and 634) two cases of eimes.^ For the present subjunc- tive I note susums in Vie St. Edmoncl, line 948. For the imperfect indicative we find an odd set of forms in the Vie St. Gregoire: ereit, eroit, eroient; they are numerous, found, however, by the side of the regular ere, ert. Meyer ^ says he has never seen these forms elsewhere. They are evident creations on -ebam imperfects (one never finds erot). We note esteum in Chardri, Jos. 1712. For the future we see ierc in Camb. Psalt. CXLV, 1. Faire. For f aimes we refer to the note by Paris on esmes, just quoted. He refers to the examples previously given by himself,^ and adds one from Oxf. Psalt. We note an example in the Vie St. Gregoire, 983, where the faimes has the sense of a subjunctive; a few lines above, 979, occurs fagons. For faisum, cf. the note on disiim, above. The forms of the future, frai, etc., are very frequent in Anglo-Norman cf. (above, p. 61, § 17, 1, b), but not confined to our dialect.® Getter. For gettre, cf. above, p. 119, 3 ; for jutta, etc., cf. above, p. 62, 4? a. Plaire. For the present subjunctive, plaise occurs already in the Psalters. Pouvoir. In Aspremont, line 63, we find poreir for poeir ; ^ this recalls the imperfect poreint of the PUerinage, line 511 (cf . above p. 91, 6). For the present participle we find poant in Oxf. Psalt., but both 2)oant and i^uissant in Camb. Psalt.^ 1 Qme. de Berneville, p. xviii. 2 Grammaire Langue (V Oil, F, 269-270. 3 Bom. XVI, 180 and 201. * Ibid. XII, 201. 5 Accent Latin, P- 71. « Meyer-Liibke, Gram. II, 396, § 314. 7 Ihid. 234, § 169. » Bom. XII, 448. 9 Meyer-Liibke, Gram. II, 248, § 183. MORPHOLOGY 129 Prendre. Metathesis is here frequent: pernons, pernez, etc. (cf. above, p. 91, 5, a). These forms occur in Norman too.^ Savoir. I note soi for sais in Melanges, V, 8.^ Examples of siez, siet (sapis, sapit) are given by Stlirzinger.^ I add three of siet from Vie St. Thomas (I, 94 ; II, 2, 16). No satisfactory explanation for these has been given.'' Trouver. I note troffe as present subjunctive in Sar- denai (0).^ 1 Meyer-Lubke, Gram. II, 257, § 189 ; Litblt. V, 69. 2 Bom. IV, 377. 3 Orthographia Gallica, p. 39. 4 Meyer-Ltibke, Gram. II, 306, § 243. s j^^m. XIV, 91. TEXT SELECTIONS It is not my purpose to make text selections a feature of my completed Manual ; but on account of the peculiar inter- est of the Anglo-Norman dialect, both for students of Eng- lish and of French, it has seemed desirable to add some representative readings ; these will facilitate the use of this portion of the Manual for practical purposes. In many cases I have selected extracts from editions now difficult of access; I have tried to choose extracts which illustrate phenomena described on the pages preceding the present appendix. I add, in the form of foot-notes, variant readings from the manuscripts, suggestions of reviewers, and the like, so as to furnish the student with somewhat of a critical apparatus for his reading. In the case of poetry, I retain the original numbering of the lines as found in the editions cited. PHILIPPE DE THAUN: CUMPOZ^ Prologus Philipes de Thalin At fait une raisun Pur pruveires guarnir De la lei maintenir. 5 A sun uncle I'enveiet, Que amender la deiet, Se rien i at mesdit En fait u en escrit, 1, rhilippe CLS; taim S.—b, le enveit S. — Q, le LS ; deit S.— 7, Se il de rien ad m. S. — 8, v en f. v ^S' ; ne en fait ne C. 1 Edition Mall, p. 1. 130 TEXT SELECTIONS 131 A Hunfrei de Thaiin, 10 Le chapelain Yun E seneschal le rei : Iqo vus di par mei. Salutatio ad Patrem Or oez sun sermun Cum le met a raisun. 15 Icil Deus ki tut fist E ki tuz jurz veir dist, II guart I'anme de tei. Que il n'i ait desrei, Qu'ele ne facet rien 20 XJ tuz jurz n'i ait bien, Ne li seit purluigniee La joie apareilliee. Maistre, un livre voil faire E mult m'est a cuntraire 25 Que tant me sui targiet, Que ne I'ai cumenciet; Kar mult est necessaire Cele ovre que voil faire ; E mult plusur clerc sunt 30 Ki grant busuin en unt, Ki pur mei preierunt E m'anme beneistrunt. E sainz Augustins dit La u fait sun escrit, 9, unfrei C, unfrai L, unfraid V; taun S. — 10, ydun LV, yhim C, yun S. — 11, Le chapeleiii S, lu CL. — 12, dit LS. — 14, Cumme S. — 15, E cil O ; deu S. — 17, le alme S. — 18, ait ia d. L. — 19, face LS, fait C. — 20, ni S, nen L, ne C. — 21, purluine S. — 22, apareille S. — 23, Mais un L. — 25, targe Z, targiez S. — 26, cummeciez *9, acu- mencet C. — 27, ker L; necessarie CS. — 28, Cel CL, eel liure S. — 29, plusurs clers CS.— SO, Ke ^. — 31, prierunt X. — 33, saint CS ; augustin S, austius L ; le dit LS. — 34, u il f . CL. 132 MANUAL OF OLD FRENCH DIALECTS 35 U numet le librarie Ki mult est necessarie As priiveires giiarnir De la lei maintenir : Iqo fut li saltiers 40 E li antefiniers, Baptisteries, graels, Hymniers e li messels, Tropiers e leQunier E canes pur plaidier 45 A eels ki le mal funt, Envers Deu se forfunt, E cumpoz pur cunter E pur bien esguarder Les termes e les cles 50 E les festes anuels. Par qo devum guarder Nostre lei, celebrer Des Pasches, des Noels Les granz festes anuels. 55 U par cest les tendrunt, U viaz i faldrunt. D'iqo me plaist guarnir Cels ki unt a tenir Nostre crestientet 60 Sulunc la Trinitet. ^o dit sainz Augustins, Ki fut mult bons divins, 35, libraire *S'. — 36, Ke >9. — 38, A la OZ. — 41, Baptisterie CS; grahels C, e graels LS. — 42, Li hyraners C. — 43, Tropeir S ; le- cuners C, leconiers, L. — 44, pur parlers C, porparliers L. — 46, Et vers deus S. — 47, ciimpot C, compote Z, compot S. — 49, clefs L. — 50, annuels S. — 51, Par cest L ; deuiins C. — 52, e (et S) eel. LS. — 53, de iioels C. — 54, Les] Des CLS; annuels S. — 55, le tendr. L ; U par cez latendrunt C ; par ceo les tendr. S. — 57, a guarn. L. — 58, Ces O, Ceals L, Icels S ; lunt a t. ^. — 50, cristientet (xpist. L) CL. — 61, dist C; saint S ; austins L. — 62, bou S; devins CLS. TEXT SELECTIONS 133 A\asimkes pot estre Que il unkes seit prestre, 65 S'il ne set cest librarie Dunt faz cest essemplarie. Eeprehexsio Allegorice per Proverbia Que ferat pasturel Ki nen at nul drapel ? Cum guarderat berbiz 70 Ki nen at nul pastiz ? E Deus, cum cumbatrat Ki ses armes nen at ? N'avum fei ne creance, Ki doiist estre lance 75 Cuntre eels enemis Ki sur nus sunt espris. II pernent la citet, Le mur unt enfundret, Fait i unt grant baee, 80 Vunt i od grant huee. ^0 est pur le seignur Ki se siet en la tur, Ki ne se pot defendre N'od els bataille prendre. 85 Cument pot hom leer Que bien curget par mer Nef, ki seit desquassee E desuz enfundree ? 64, il lacking S. — 65, siet L ; libraire S. — 66, dum faiz C; essam- plaire ^S*. — 67, frat S. — 60, cument S. — 70, qui nen nad C ; ki nad *S'. — 74, Kb S; deust Z, dust 6'. — 75, c. noz ^. — 77, prennent C ; cite CLS.—1S, effundre C, esfundre L, enfundre >S'. — 79, Fait vnt S, Hardi unt C ; baudee 6'. — 80, Vunt lur huee ^. — 81, par CL ; le lacking *S'. — 82, Ki siet sus en L, qui set en C. — 84, eals L ; batailes ;S'. — 85, pot lorn L. —86, Ki L, qui C ; curt C, curge L. — 87, Ke S ; esquasse S. — 88, desus C; enfundre S, esfundree L. 134 MANUAL OF OLD FRENCH DIALECTS Dire qo pot li prestre : 90 Senz cumpot pot bien estre, Bien set us de mustier E ses festes nuncier. E jo li respundrai Par raisun e dirrai : 95 Horn set par us chanter Cum esturnels parler. Ne larrai nel vus die, Nen est pas juglerie, ' Ne n'est Grius ne Latins, 100 N'Hebreus ne Angevins, Ainz est raisun mustree De la nostre cuntree. Bien poent retenir ^o dunt jos voil guarnir, 105 Se il volent entendre E bone guarde prendre. Mais cuit qu'al quant dirrunt, Ki puint de sens n'avrunt, Qu'en vain me travaillai, 110 Quant cest livre ordenai ; E jurrunt, pot eel estre, Le vertuus eel estre Que unc ne soi rimer Ne raisun ordener. 89, Dirre coe pot 1. p. C ; Dirra (Dirrat S) cupet li p. LS. — 90, pot lacking S. — 01, Bien ws set demustrer *S'. — 92, Ses festes et denuncier S. — 93, lui L. — 94, et mustrai S. — 95, Horn CL, Le S. — 96, Cum S ; e C ; et L ; asturnel CS, estornel L. — 97, lerrai L; nel S, ne OZ. — 98, Ne est C, Nest ^\— 99, griu CLS.— 100, Ne liebreu ne ang. L^ Ne peitevins ne ang. S^ ne nen est ang. C. — 103, pot lem ^S*. — 104, Co dunt vos v. Z, Coe dum ges v. C, Ceo dunt io v. S. — 105, atendre S. — 100, Et voilent g. p. S. — 107, M. quid quaiquant L, Mais que que alq. C, Mais li quel ke d. S. — 108, Ke 8, ni 6^.-111 E] lacking in all. —112 Le L, Les CS', uertur ^S'. — 113, rimeier L, TEXT SELECTIONS 135 115 Mei ne chalt que fols die, Jo ne m'en repent mie ; Asez sunt malpaiiiers Pur mult petiz luiers E humes pur blasmer, 120 Neient pur amender. Unc pur fols nel truvai Ne ne m'i travaillai ; N'unc ne fut asnes net Ki bien loast citet. OXFORD PSALTER 1 PSALMUS I 1. Beneurez li huem chi ne alat el conseil des feluns, & en la veie des peccheurs ne stout, & en la cliaere de pestilence ne sist 2. Mais en la lei de nostre Seignur la voluntet de lui, & en la sue lei purpenserat par jiirn e par nuit. 3. Et lert ensement cume le fust qued est plantet de juste les deciirs des ewes, chi dunrat sun frut en sun tens 4. Et sa fiiille ne decurrat, & tutes les coses que il unques ferat seriint fait prospres. 5. Nient eissi li felun, nient eissi ; mais ensement cume la puldre que li venz getet de la face de terre. 115, Mei ne chat S, Mais ne chalt C, Mais qui chalt L ; fol S. — 116, Je (7, men S, me CL. — 117, Asunt S ; mas parlers C. — 118, E purl,; mult lacking C ; petit S; luers CLS. — 120, Naient U, Nient S, E nient L. —121, Vnkes p. fol /S'. — 122, mi CL, men S. — 123, Ne ne fut S ; asne CLS; ne LS. — 124, cite LJS. Variants (given by Michel) Cod. Cott. 1. Beonure. harun. cunseil. et. pecheurs. et. 2. Seignor la volunted, e. Deest de lui. 3. E. ki. froit. son. 4. E. e. ferad. 1 Edition Michel, pp. 1 and 212. The accent marks in the first psalm I take from the copy in Bartsch, Chrestomathie, 7th ed., col. 53. 136 MANUAL OF OLD FRENCH DL\LECTS 6. Empurice ne resurdent li felun en juise, ne li peclieur el conseil des dreitiu'iers. 7. Kar nostre Sire cunuist la veie des jiistes, e le eire des feluns perirat. PSALMUS CXXXVI 1. Sur les flums de Babilone, ilnec seimes e plorames, dementres que nus recordiums de Syon. 2. Es salz els milliu de li, suspendimes noz organes. 3. Kar iluec demaiiderent nus, chi chaitis menerent nus, paroles de canz ; 4. E chi menerent nus : Loenge cantez a nus, des canz de Syon. 5. Cument canterum-nus le cant del Segnor en estrange terre ? 6. Si je oblierai tei, Jerusalem, a obliance seit dunee la meie destre. 7. Aerde la meie langue as meies jodes, si mei ne remem- berra de tei ; 8. Si je ne proposerai Jerusalem el cumencement de la meie ledece. 9. Eemembrere seies, Sire, des filz Edom, el jur de Jeru- salem ; 10. Chi dient: Voidez, voidez, desque al fundament en li. Variants Cod. Cott.—Q. Empurigo ne surdent. peclieor. 7. I'eire. Cod. Cott. 1. Babylouie. 3. caitis. chanz. 4. E chi ki. 7. re- membera. 8. al. leece. 10. fundement. Psalt. Corb. 1. Babylonie, iloec. recordiuns. 2. de millui de lui. 3. ki caitifs m. voz p. 4. E ki en veie menerent nus : L. c. ^ n., de c. de S. 5. C. c.-n. del Seignur. Desunt le cant. G. Se jo. e ^ o. don6e. 7. Aerdet. k mes joes, se m. ne remembeiat. 8. So jo ne purposerai. comencement. leece. 10. Ki d. : Voidient, v., d. el. TEXT SELECTIONS 137 11. Fille de Babilone caitive ; beneurez clii redunrad a tei la tue giieredunauce, laquele tu regueredunas a iius, 12. Beneurez chi tendra, e esgenera les tues^ enfanz a la pierre. CAMBRIDGE PSALTERS PSALMUS I 1. Beoneiiret li heom ki ne alat el cunseil de felims, e eu la veie des pecheurs ne stout, e en la chaere des escharuisetirs ne sist. 2. Mais en la lei del Seignur la volentet de lui, e en la lei de lui penserat par jur e par nuit. 3. E iert ensement cume fust tresplantet dejuste les ruisals des ewes, lequel sun fruit durrat en sun tens. 4. E la foille de lui ne decurrat, e tuit ceo que il ferat serat feit prospre. 5. Nien issi felun, mais ensement cume puldre, lequel degetet li venz. 6. Pur ceo ne resurdrunt li felun el ju'ise, ne li peclieiir en la asemblee des justes. 7. Kar cuneut li Sires la veie des justes, e I'eire des feluns perirat. Variants Cod. Cott. 11. Babylonie. redunrat. reguerredunas. 12. esge- rera. tueiis. Psalt. Corh. 11. Babilonie. ki reguerdonerat. guerredonance. guerdunas. 12. ki tendrat, e esgenerat 1. suens emfanz. Variants (given by Michel) 1. conseil. de. 2. Meis. Seinur la voluntet. 4. serrat fait. 5. Nient. meis. 6. assemblde. 1 Meister collation : tuens. 2 Edition Michel, pp. 1 and 244. 138 MANUAL OF OLD FRENCH DIALECTS PSALMUS CXXXVI 1. Sur les fluez de Babiloine, iluec sesimes e ploiarames, cum nus recordissiims de Sion. 2. Sur les salz eu miliu de li, suspendimes noz estrumenz. 3. Ker iluec nuns demandowent ki cliaitifs nuns niener- ent, paroles de chauQun ; e cil ki nuns tormentowent : Lie, chautez a nuns des chanz de Sion. 4. Cumment chanterums la cantike Damne-Deu, en aliiene terre ? 5. Se jeo serai oubliez de tei, Jerussalem, en oubliance seit ma destre. 6. La meie langue alierged a mun guitrun, se jeo ne me recorderai de tei, se jeo devant ne metrai Jerussalem en r comencement de ma leece. 7. Remembre te, Sire, des filz Edom el jurn de Jerussalem, des disanz: Esfowed, esfowed juske al fundement de li. 8. La fille Babiloine deguastee ; bonelired ki guerdunerad a tei la twe feiede, ke tu guerdunas a nuns. 9. Bunewred ki tendrad, e aburterad ses petiz enfanz a la pierre. QUATRE LIVBES DES ROIS^ Ll QUARZ LiVRES DES ReIS .XXV. Al nuefme an lu rei Sedechie, el disme meis, el disme jur del meis, vint Nabugodonosor li reis de Babilonie a tute se ost a Jerusalem, si 1' asejad e ses engins i levad. Mais cil dedenz tindrent la cited jesque al unzime an lu rei Sedechie, le nofiue jur del meis. Lores i fud la famine tant grande que tenir ne porent la cited,2 1 Edition Le Roux, p. 434. 2 What Ollericb calls " Moderne Hand " here substitutes cit'ed. TEXT SELECTIONS 139 Si s'enfuirent^ nuitantre cez ki dedenz erent, e cil de Chaldee f urent ^ a siege, e li reis Sedechias s'enf uid ^ par la cliampaine del desert, E li oz de Chaldee le sout, si V pursewid, e prist e retint es plaine * de Jericho, e tuz ses hummes s'enf uirent,* Si r guerpirent en champ, e cil menerent lu rei Sedechie devant le rei de Babilonie a Antioche ; ^ e li reis de Babi- lonie Fist devant li ^ meime ses fiz ocire, e ses oilz crever, e de chaene ® le fist lier e en Babilonie mener. El quint meis e el setme jur del meis, co fud li dise-nofme an^ del regne lu rei de Babilonie que Nabuzardan li cunesta- ble ^^ de la chevalerie de Babilonie vint a Jerusalem ; E tuchad le f u e arst lu temple nostre Seignur, e le palais lu rei, e tute la cited ; " E fist les murs de tutes parz agraventer, E les remasilles del pople ki furent la remes, e ki fuid s'en furent al rei de Babilonie, e les altres qu'il truvad tuz menad en chaitivier en Babilonie ; Mais del poverin de la terre i laissad partie que il s'entre- meissent de la guaignerie. Lores prist les riches columpnes de araim ki al temple furent od tutes les basses ^^ e lu vaissele '^ ki fud el temple, de argent e de araim ; e tant i out que I'um " ne sout lu peis. E Nabuzardan prist les pruveires e les tresoriers del tem- ple, e un des cunestables e des privez lu rei^ e altres une masse, si's enveiad en Antioche al rei. E la les fist ocire li reis, e jetad cez de Juda hors de lur pais. 1 Ollerich's collation : s'en fuirent. ^ oilericli : i furent. 3 Ollerich : s'en fuid. 4 oUerich : plaines. 5 Ollerich : s'en fuirent. ^ Modern hand : antiochie. 7 Modern hand : lui. 8 Ollerich : chaenes. 9 Ollerich : ans. i*^ Ollerich : cunestables. 11 What Ollerich calls " Zweite Hand" here substitutes cit'ed. 12 Zweite hand : basses. i^ Ollerich : vaissel. i* Ollerich : lii. 140 MANUAL OF OLD FRENCH DIALECTS Si fist Godolie le fiz Aica le fiz Sapha maistre sur tant de frapin cume en la terre remest. Cume li paisant surent que li reis Nabugodonosor out fait Godolie maistre de la terre, Ismael le fiz Natanie, e Johan- nan li fiz Caree, e Saraia li fiz Thenamech, e Jeclionias li fiz Machati, e lur cumpaignuns vindrent a lui en Masphath. E Godolias lur fist serement que mal ne lur freit, si lur dist : Mar averez pour de servir a cez de Caldee ; en ceste terre remanez, e le rei de Babilonie servez, e bien vu3 esterrad. . . . BE ANDAN ^ Donna aaliz la ""erne Par q' ualdrat lei divine Par q' creistrat lei de terre E remandrat tante guerre 6 Por les armes henri lurei E par le cunseil q' ert entei Saluet tei mil emil f eix^ Li apostoiles danz benediz. Que comandas 66 ad enp's 10 Secund sun sens entre Mis En let^ Mis & en romanz Esi cu fud li teons cumanz De saint Brendan le bon abeth Mais tul defent ne seit gabeth. 16 Q'^nt dit q'' set e fait q"* peot Jtel seruant blasmer ne steot Mais si q' peot ene uoile Variants (from Oxford Ms. given by Suchier) 1, Donna aaliz] De ma halt. — 2, ualdra. — 4, remei/idm. — 5, Par. le. _6, conseil. de tei. —7, Salue. — 8, li apostolies donz beneiz. — 9, commandas. — 10, Si cum. en \etre. — 11, remans. — 12, issi. fut (usually found), tuens. — 13, brandan. abet. — 14, tu le defende. gabet. — 15, poet, — 16, seriant. stuet. — 17, cil. puet. uoiliet. 1 Edition Suchier, Bomanische Studies, I, 667. TEXT SELECTIONS 141 Dreiz est q^ cil mult se doile Jcist semz deu fud nee? de reis 20 De naisance fud des ireis Pur 66 q® fud de regal Im Pur oc en tent a noble fm Ben sout q^ lesc'pture dit Ki de cest mund fuit de delit 25 Oc^ deu de eel tant en aurat Que pP demander ne saurat Pur oc guerpit cist reials eirs Les fals honurs p^ iceals ueirs Dras des Moine pur estre vil. 30 En cest secle cti en eisil Prist elordre eles liabiz Pu's fud abes par force esliz Par art de liu mult iuincZrent Qui ale ordre bein se tmdrent 35 Tres Mil suz lui par diuers leus Munies aueit b'^ndan li pius. De lui pnanz tuz ensample Par sa vertud q^ ert ample Li abes brendan p'st enp«pens 40 Cu home q' ert de Mult g*nt sens De g^nz cunseilz e de rustes Cti cil q' ert forint iustes De deu p^er ne fereit fin Pur sei ep? trestut sun lin. 45 E p^ les morz ep^ les uifs Variants (from Oxford Ms. given by Suchier) 18, mult se doile] puis sen duillet. — 19, fu neiz. — 20, naissa?ice fu de. — 21, ceo (usually so written), fu. reial. — 22, Pur oc] de mielz. — 23 [B]ien. la scj'iptwe. — 24, fuiet le. — 25, ciel. aura. — 26, demander plus, saura. — 27, cist (lacking). — 28, le. onurs. ices. — 29, m[u]inie. — 30, icest siecle. — 31, abiz. — 32, fu abez. — 34, ki (usual) alordre blew. — 36, brandaws. — 37, tons. — 38, anple. — 39, brandans. — 41, cowseilz.— 43, feseit. — 45, [mor]s. 142 MANUAL OF OLD FRENCH DIALECTS Quer astrestuz ert amis Mais de une en li p*st talent Dimt deu p^er p'nt pP suiient Que lui mustrat eel parais 50 V adam fud p'mes asis Jcel q' est nost^ heritet Dun nus fumes deseritet Bien creit q' leoc ad g*nt glorie Si cu nus dit ueire storie 55 Mais nepHant uoldret uetheir V il deureit par dreit setheir Mais par peccet ada forest P^ quei z sei nus fors mist Den en p'et tenablement 60 Cel lui mustret ueablement Amz q'l murget noldreit uetlieir Quel sed li bon deurunt aueir Quel lu le' mal aueir deurunt Quel merite il receurunt 65 Enfern pried uetlieir oueoc E quels pein^s auruiit ileoc Jcil felun q' ];)ar orguil Jci prennent par eols escuil. De gurrer deu e la lei 70 Ne ent^ eols nen unt aiiiur ne fai Jed dunt lui p's est desir Voldrat Brandans par*^^" sentir OcZ sei phnes cmiseilz en p^nt Qua un deu serf 9fesse rent 75 Barinz out nun cil ermite Murs aut bons esamt uitte Li fedeilz deu en bois estout Tres cenz moiiies od lui out Variants (from Oxford Ms. given by Suchier) 56, [s]eier. — 60, [u]isablemcnt. — 62, deuront. — 66, iluec. — 68, e[scu]el. — 70, fei. — 76, sai?ite uie. — 78, aueit. 80 TEXT SELECTIONS 143 De lei p^ndrat conseil elos De lui uoldrat aueir ados Cil li mustrat par plusurs diz Beals ensaples ebons espiz Quil il uit en mer z enterre Q'^nt son filiol alat querre 85 Co iml m'noc q^ fud frerre Del lui V cist abes ere. GAIM AE 1 Quinte iur apres reis Harold vint; Contre Norreis bataille tint. 5225 Co fu Harald fiz Godewine, Ki des Norheis fit discipline. Co fut al Punt de la Bataille : Norreis trouat, pernant almaille. Li reis Harold done les sewi, 5230 Ireement se combati. Laltre Harald el champ oscist, E de Tosti ensement fist. Sur les Daneis out la victorie, La gent del Suth sembla grant gloria. 5235 Mais bom ne sout conter demis Cels kel champ furent oscis. Totes les nefs, e Iur herneis, Ad feit saisir Harald li reis. Variants (from Oxford Ms. given by Suchier) 81, S[il].— 82, respiz. Variants 5228, Quart II; Haralt A Harald J/L. - 5225, le fiz DL.-- 5227, le (for al).-5228, aumaille ^.-5221), siwit D siwi L, suit H; done omitted, H.-5230, Irousement ^- " ^^l^^' ^'\^°^' i" ' Norois II -5234, de Sud L; gloire 2>.-523o, demi if.-52o6, Ce. A Qui H; pais LII -5237, Iur (for les) H. - 5238, Ad omitted M. 1 Edition Hardy and Martin, I, 221. 144 MANUAL OF OLD FRENCH DIALECTS Le fiz eel rei i fust troue : 5240 Cil fust a Harald amene. Merci cria, treu pramist ; Haralde homage de li prist. E de trestuz les remananz, Prist bons ostages, e vaillanz ; 5245 Od vint nefs les les sat aler : Done airent tant ke sunt en mer. Cine iors apres sunt ariuez Franceis, od bien vnze mil nefs, A Hastinges, desur la mer : 5250 Hoc firent chastel fermer. Li reis Harald quant il oi, Leuesque Aldret ad done saisi Del grant auer e del herneis Kil out eonquis sur les Norreis. 5255 Merleswain done i lessat ; Pur ost mander en Suth alad. Cine iurs i mist al asembler, Mais ne pout gueres avner, Pur la grant gent ki ert oscise, 5260 Quant des Norreis fist Deus justise. Tresken Suthsexe Harald alat, Tel gent cum pout od li menat. Ses dous f reres gent asemblerent ; A la bataille od lui alerent, 5265 Li uns fust Gerd, laltre Leswine, Variants 5239, celui DH (for eel rei). — 5239 and 5240 omitted L — 5240, Cel A Si, ^. — 5241, triu i7.— 5243, tuz if.— 5244, bons omitted H. — 5246, font H (for airent). — 5248, bien omitted H \ IX H (for unze) ; miliers L (for mil nefs). — 5251, co old Z), co oi L, ceo oit H (for il oi). — 5252, Aired DLH\ i H (for ad).— 5253, de i). — 5254, des H (for sur les). — 5255, idunc DH. — 525G, el jy. — 5259, quot D (for ki ert).— 5201, Sudreie L. — 5262, Tant D, Tels L, Tieus H (for Tel gent) ; cum sei D ; od sei Z. — 5263, Les; genz i). — 5265, Gered i, Gerard If; Lefwine D.— TEXT SELECTION^ 145 Contre la gent de vltre marine. Quant les escheles sunt rengees, E del ferir aparillees, Mult i out genz dambesdous parz ; 5270 De hardement semblent leoparz. Vn des Franceis done se hasta, Deuant les altres cheualcha. Taillefer ert oil apelez, loglere estait, hardi asez. 6275 Armes aueit e bon cheual : Si ert hardiz e noble vassal. Deuant les altres oil se mist ; Deuant Engleis merueilles fist. Sa lance prist par le tuet, 5280 Com si CO fust vn bastunet : ' Encontre mont halt le geta, E par le fer receue la. Trais fez issi geta sa lance ; La quarte feiz, mult pres sauance, 5285 Entre les Engleis la lanca, Par mi le cors vn en naffra. Puis treist sespee, arere vint, Geta sespee, kil tint, Encontre mont puis la receit. 5290 Lun dit al altre, ki co veit, Ke CO estait enchantement Variants 5266, lost H (for la gent). — 5267, eschieles D ; furent LH; rengies D. — 5269, dambes DH, damp L. — 5270, leuparz BL. — 5273, Tail- lifer D, Talifer //. — 5274, luglere I)LH\ ert BL, omitted H\ ardiz D, hardiz i. — 5276, bon L (for noble). —5277, cil omitted H.— 5279, diet B, cued i. — 5280, Si cum B, Si come //; hastuned Z. — 5281, lengetta jET. — 5282, sa (for la).— 5284, mult omitted^; par X, puis H (for pres).— 5285, la omitted i. — 5286, naurat Z>, nauera H.—5281, 5288, lespee BH.—62S8, Et geta //. — 5289, halt BL, haut H (for puis) ; le (for la) BH. L 146 MANUAL OF OLD FKENCH DIALECTS Ke cil fesait, deuant la gent. Quant treis faiz out gete lespee, Le cheual, od gule baiee, 5295 Vers les Engieis vint a esleise, Si i ad alquanz ki quident estre mange, Pur le cheual ki issi baiout, Le iugleor apris li out. Del espee fiert vn Engieis ; 5300 Le poing li fait voler maneis. Altre en fiert tant cum il pout : ' Mai guerdon le ior en out. Car les Engieis, de totes parz, Li lancent gauelocs e darz. 5305 Lui oscistrent, e son destrer : Mar demanda le colp primer. Apres i co Franceis requerent, E les Engieis en contre fierent. La out asez leue grant cri : 5310 Desci kal vespre ne failli, Ne le ferir, ne le lancer : ]\Iult i aut mort maint cheualer ; Nes sai nomer, nos mentir, Lesquels alerent mielz ferir. Variants 5292, Cil se fiert H (for Ke cil fesait). — 5294, ad la DLH (for od). — 5295, a omitted DLH ] eslessie Z>, eslesse H. — 5296, Alquant Z>, Alquanz Z, Auquanz H (for Si . . . ki). — 5297, si DL.— 5298, enpres venout H. — 5299, espie B. — 5300, piiin i>, poin i, poign //; des mains L (for maneis), demanois H. — 5301, Vn altre H. — 5304, launcerent. — 5305, Si 1 H. — 5306, cop DL, coup H. — 5307, requierent DL, les requierent //. — 5309, Assez i out H. — 5313, Ne D\ ne ruis DH, rois L (for nos). — 5314, Li Englois alerent bien ferir H. TEXT SELECTIONS 147 ADGAE, Vns moines ert Eueshamneis, Ki mut ama Deu e ses leis. La mere Deu reserui bien E honiira sur tute rien. 5 Ne laissa rien en nule guise, Qu'apartenist a son seruise. Puis quant cil bers deueit murir, A sei uit mut diables uenir. Cuiluert erent e mult engres ; 10 Deuant sei les uit ester pres. Mais quant ewe beneite esteit Getee, cum I'en faire deit, S'en fuirent tuit li felun De tutes parz de la meisun. 15 E puis tut dreit, cum dut murir, La Dame uit a sei uenir ; La cliere mere al Salueur Dist li : " Ami, ne aies pour ! Ensemble od mei t'alme merrai ; 20 En ioie, en repos la metrai ! '^ Cil la prist dune a saluer E icest respuns a chanter : " Gaude, Maria ; virgo chere ! " E rendi I'alme en ioie entiere. 25 Cum li respuns esteit pardit, El ciel tramist son esperit. Bien deiuent Crestien trestuit Seruir la Dame e ior e nuit, Asaier,^ cumfaitement 30 Peussent faire sun duz talent ; Ne mie sulement li grant, 1 Edition Neuhaus, p. 151. 2 Foerster (corrections added to the edition, p. 251) suggests : E asaier as a probability. 148 MANUAL OF OLD FRENCH DIALECTS Mais fol e sage e nunsauant. A tuz rent bien sulimc merite ; Plusurs defent de mort subite. 35 E se il en muerent de tele mort, As almes rent ele grant cumfort, Cum ele fist a I'alme de un moine, Ki ert d'un mostier de Burgoine. — Den nus duinst des pechez pardun 40 Par fei e par comfessiun. FANTOSME i Li sires d'Engleterre ad en sun cuer pesance Quant sun fiz le guerreie, qu'il nurri ad d'enfance, 80 E veit que cil de Flandres I'unt mis en errance. Pramis li unt la terre des Engleis a fiance ; Mielz volsist mort que vie qu'il eiist la puissance, Tant cum il pout d'espee ferir u de lance. Establist sun barnage par fiere cuntenance, 85 Vait encuntre Lowis, le riclie rei de France, Cuntre le cunte Plielippun, dunt vus oiez parlance, E dan Maheu sun frere, chevalier de vaillance. Mult aida Deu le pere le jor, quant il I'avance, E mustra de sa guerre bele signifiance ; 90 Que le sucurs de sun fiz, u plus fud s'esperance, Fud le jor agravente senz nule demurance. ^o fud Maheu le puigneur, sur qui vint la lance ; N'aurad mes li reis Henriz pur lui nule dutance. Li cuens de Buluine ad receu mortel plaie, 95 De si qu'as espuruns a or li sane vermeilz li raie. Ne purrad james guarir, asez ad qu'il asaie. Tant est sis freres plus dolent, e plus suvent s'esmaie, Variants 78, quor pensance L. — 80, 81 omitted in L. — 83, peust, de sa lance L. — 87, En i. — 90, ot sa fiance L. — 92, Ceo fu i. — 96, ad assez L. 1 Edition Hewlett, p. 208. TEXT SELECTIONS 149 E jure sun serrement, " La pretiuse plaie," James vers le rei Henri n'aurad nule appaie. 100 Ore chevalche Lowis, si fait le jofne rei, E Phelippe de Flandres est mis en grant desrei. Li cuens Tiebaut de France demeine grant podnei. Ja saverad li reis Henri asez u mover sei. Franceis li muevent guerre, Flameng et Cupei, 105 Li cuens de Leircestre, si i sunt ses fiz tut trei. Icil de Tankarvile ne I'aime pas de f ei ; Cent chevaliers a armes ameine en sun cunrei, Ki tuit lui sul manacent de mettre en tel desrei Ne li larrunt de terre le pris d'un palefrei. 110 Seignurs, en la meie fei, merveille est mult grant. Pur quel li suen demeine le vunt si demenant, Le plus honurable e le plus cunquerant Que fust en nule terre puis le tens Moysant, Fors sulement li reis Charle, ki poeste fud grant 115 Par les dudze cumpaignuns, Olivier, e Rodlant. Si ne fud mes oi en fable ne en geste Un sul rei de sa valur ne de sa grant poeste. Purquant lui vunt tuz maneqant, il en jure sa teste Ne larrad pur riveier ne pur chacier sa beste. 120 Or chevalche li cuens Phelipe ovoc sa grant cumpaigne, E guaste Normendie par bois e par champaigne. N'en oissiez le rei Henri qu'il une feiz s'en plaigne, Ne querre nul achaisun que la guerre remaigne. Mult ad li juesne reis espleitie, qui si bien se baigne 125 Encore en ad les mandemenz des baruns de Bretaine. 99, n'avera nul apaie X. — 100, reis i^. — 101, de Flandres X, omitted D. — 102, Tebaud — pothnei L. — 105, Leecestre, si sunt L. — 106, Tankervile. — 107, en meine a sun cunrei. — 110, mei, en est L. — 111, mana9ant L. — 112, e phis cumquerant L. — 115, Oliver e Rolant X. — 116, Ce ne fud (by erasure) L. —118, li vunt managant si en jure L. — 120, od sa grant cumpainie L. — 121, campaignie L. — 122, que une feiz se plaine L. — 123, remaignie L. — 125, Vncore L. 150 MANUAL OF OLD FRENCH DIALECTS ANGIER (Vie de Saint Grecjoirey Icist Eleuthere par nom Un inort jadis resuscita, 324 E por itaiit In amena A cele houre, ovec sei, Gregoire Priveement eu I'oratoire; Si I'en requist, por De amor, 328 Q'orast por lui Deu, q'icel jor Veaus noil trespasser lu donast Q'od les enfanz lu jelinast ; Mais ne demoura fors brefment 332 Pues q'is orent ens emblement Amdui, tant com lor plot, ore, Quant Gregoires li ami De Tant fort e legier sei sentit 336 Qe s'il vousist sanz nul respit Jeliner jesq'a I'endemain, Tant sei sentit vigrous e sein, Bien lu pot faire sanz nul gref 340 Del mal del ventre ne del chief, Si q'il soi merveilla de sei Comment qo pot estre e par qei Q'il ne mangot com il sent faire. 344 Ne voeil d'iqo plus avant traire, Car il meisme asez reconte A quel cele aventure amonte Enz el tierz livre renome 348 Del Dialoge translate, La ou par sa raison escleire Les vertuz del dit Eleutlieire. ^0 poet asez cliasqun voier, 352 Per q'il voilge tant travaillier Q'iloec en dreit lu deinge querre. 1 Edition P. Meyer, Bom. XII, 156. TEXT SELECTIONS 151 Mais iceste avant dite guerre La quele encontre sei enprist, 356 Ja seit grant ennui lu feist, One por QO ne fut plus oisdif, Ainceis ert tant plus ententif Nut e jor d'orer e de lire, 360 Ou d'estudier, ou d'escrire. Si q'onqes ne cessa nule hore. Qei feroie plus de demore De reconter com pues vesqit, 364 Com nostre Sire lu rendit Por son servise sa merite ? — Gregoire en cele iglise ainz dite Lone tens tot aresie maneit 368 Ou de cru letim lu pesseit La seinte Silvia sa mere, Quant un jor en guise d'un frere Lui trovot un angle escrivant 372 Qui bien resemblot i)ar semblant Tins oem qui fust de mer jete, De peril de mort escbape, La nef de qui fust perillee, 376 Par tempeste de mer brisee. Cist itel lu appareisseit La ou par costume escriveit. Si lu requist por De amor 380 Q'elist merci de sa dolor, Pitousement, od voiz plorable. Gregoire qi fut merciable, Douz, francs, pitous e deboneire 384 Tantost sanz demorer, en eire. Sis deners trest de s'aumosnere, Si lu donot od f ranche cbiere ; E cil ilors graces rendant 388 Partit de lui liez e joant. Mais ne demora fors briefment, 152 MANUAL OF OLD FREXCII DIALECTS Quant eis le vos tot freschement Un autre jor a lui venir. 392 Si se perneit a dementir E se pleiust qe poi ot reQu Encontre iqo q'ot molt perdu ; E Gregoire erraument regiers 396 Lu donot sis de ses deniers Doucement e de quer verai, E li perillie sanz delai Merciz rendant s'en vait joious. 400 , Mais el tierz jor este le vous Tot de novel par devant lui Pleingnant e plorant a ennui, Com s'il etist tot oblie 404 Quanq'il lu ot ainces done, E dist : " Por amor De, beau sire, " Qui en son regne lu vos mire " E por la salu de vostre emme 408 " Qe Deu la defende de blemme "E de damage e de peccliie, " Aies merei del perillie, ** De ma mesaise, e de ma perte 412 " Qui tant par est dure e aperte "Aiez, por siente charite, " Compassion e piete ; " Si me fai consolation 416 " De quel qe seit de vostre don, ** Dom mis doels seit asouagiez." Gregoire atant s'est esbrusciez, Com oem de charite espris, 420 Douz, merciable e francs e pis, E dist a son chamberleuc lors : " Va tost e sis deners onqors " Lui aporte, si tu les as. 424 — Sire," fist s'il, "jo nes ai pas, " Si Deu me saut, n'or ne argent TEXT SELECTIONS 153 "Taut dom negun confortement " Lii pelisse a Oste liore faire." 428 Lors fut Gregoire en grant arvaire Desquant soi vit en tel destreit, Car d'une part pitous ereit E d'autre triste e angoissous 432 Del povre qui vit soufreitous, E de sa destresce demeine. Neporquant a la fin aceine Regiers son chamberlenc a soi, 436 Si lu dist : " Va tost, par ta foi, " Si cerche par trestot laienz " Huges, almaires, vestimenz, " Si riens par aventure i truisses 440 " Dom lu reconf orter pelisses, "Q'il ne s'en aut triste e plorant." ' Atant respondit li servant : ^' Sire, sachez qe des piece a 444 " Enquis ai trestot quanq'i a, " Mais veirement dener ne maille " N'i troefs ne nule rien qui vaille ^' N'en robe, n'en veisselement, 448 " Estre I'escuele d' argent " La quele a vostre maladie " De legun sovent replenie " Vos traniet vostre bone mere. 452 — Done," dist Gregoire, od bele here, " Va donqes tost, e si I'aporte "Al povre qui se desconforte, " Q'is eit, seveaus, de tant solaz." 456 ^o q'il ot dit fut fait viaz Einsi com il I'ot commande, E cil qui ert povre quide La requt, s'accuillit sa voie 460 Graces rendant od molt grant joie, Mais ]0 quit quant repeirira 154 IVLINUAL OF OLD FRENCH DIALECTS De chef, rens maes ne li qiierra, Ainceis lu voudra, sanz mentir, 464 Quanq'ainz lu a donne merir. A quel vos tendroie lone tens ? Mais tant par eroit en toz sens Gregoire espirez de vertuz, 468 De miracles seinz reqenuz Enpres la visitation De I'angle dom faz mention, Qe tut cil qui od lui vivoient 472 , Tot autritant lu redoutoient Com s'il fust per a seint Andre Qui de s'iglise ert avoue. CHAEDEI 1 La Vie des Set Dormans La vertu deu ki tuz jurs dure E tuz jurs est certeine e pure Ne deit pas trop estre celee. Car quant il fet chaut u gelee, 5 Nues voler, escleir u vent, De ceo n'unt merveille la gent, Ne de la terre ne de la mer, Pur ceo k'il sunt acustumer De veer cele variance, 10 Cum deu le fet par sa pussance j E ne puroec mut esbaifs I serrium, se ententifs Pussum estre del penser. Variants (given by Koch, p. 189) 1, LO ke ; L tutjurz, tutjurs ; ibid. 2. —5, Nuwes. —6, Z De ceo num m'ille, etc., De ceo um merveyle. — 7, tere. — 8, 0, keil. i Edition Koch, pp. 76 and 164. TEXT SELECTIONS 155 15 20 25 30 35 40 E deu nus vousist itant tenser. Ne purrum pas a chef venir, Se deu nel vousist nieintenir/ Ki purreit or sanz encumbrer Les esteiles del eel numbrer, Ne la hautesce del firmament Ki tant est cler e tant resplent, E la laur de tut le munde E de la mer ki est parfunde : ^ Mut purreit ben esmerviller Ki weres en vousist parler. Mes nus en pen sum mut petit, Car aillurs avum le nostre alit Enracine par grant folie En mauveste e en tricherie. Car d'autre penser n'avum cure Eors de cele malaventure K'en cest secle veum user. Trop i delitum, seinnurs, muser, Si n'avrum fors hunte e dulur Pur teu penser a chef de tur. Cil ki de quoer vout deus amer E retrere vout del amer De cest munt ki tant travaille, Mut se delitera sanz faille Des uvrainnes Jhesu Crist K'uncore fet e tuz jurs list. Leal serra ki par teu penser OF THE UNIVERSifY OF Variants 14, L E ren (or veu?) n'uousist, etc. — 17, LO ore. —22, LO ke.— 23, L p?irreit len.— 26, Zaffit. — 28, mauveiste.— 30, LO For. — 31, LO ke en. — 33, LO for. —34, tel au chef. — 35, ke deu. — 37, ke. — 39, overaynes. — 40, ke ; LO unkore ; L tutiurz, tuzjurs. — 41, LO lei. 1 Mussafia {Zt. Ill, 604) suggests a period at end of this line. 2 Mussafia suggests an interrogation point here. 156 MANUAL OF OLD FRENCH DIALECTS Lerra sa grant folie ester. Pur teus curages tenir E le ben k'en poet avenir, 46 Une aventure viis cunterai, Dimt ja ren ne iiientirai, D'un miracle ke fist Jhesu, Ki pitus est e tuz jurs fu. Ki deus eime de bon curage 60 Or i tende, si fra ke sage. Ne voil pas en fables d'Ovide, Seinnurs, mettre mun estuide, Ne ja, sacliez, ne parlerum Ne de Tristram ne de Galerun ; 55 Ne de Eenart ne de Hersente Ne voil pas mettre m'entente, Mes voil de deu e sa vertu, Ki est pussant e tuz jurs fu, E de ses seinz, les set Dormanz, 60 Ki tant furent resplendisanz Devant la face Jhesu Crist. Car si cum il est escrit Vus en dirrai la verite De cbef en chef cum ad este. Petit Plet Tuz ne sunt pas amis verais 1630 Ki vus losengent, de deus en treis. Ki beit e mangue a ta table Variants 44, ke peot.— 46, L ni. — 47, L Dun. —48, LO tutjurs.— 49, LO aime. — 50, Ore i antende ; Zfrad. — 51, L touid' ; de Ovyde. — 52, seynurs. — 53, LO Renard. — 50, maentente.— 58, L tutiurz, tuziurs. — 59, O des ses, etc., les VII, D.— 60, LO ke. 1629, OV verray(i)s. — 1630, OV ke ; LO dous ; O u treys.— 1631, Ke ; maniuwe. TEXT SELECTIONS 157 Ne serra ami cuvenable, Se essae nel as avant — Mar le crerras ne tant ne quant ! 1635 Car ceus ki plus vus promettereint A chef de tur vus decevreiiit. Teus juent e bel vus rient, Au paraler de vus mesdient. Cil vus eiment e pres e loin, 1640 Mes il vus faudrunt au grant busoin Ceo est Pamiste de main en main : Tant as — tant vans, e tant vus eim. De teus en avrez vus asez, Deske les eez ben espruvez. GUILLAUME DE BERNEVILLE La Vie de Saint Gile^ Gires est en la veie mis, Gerpist sa terre e ses amis ; II nen ad n'or n'argent od sai, Cheval ne mul ne palefrei; 646 II n'en porte ne veir ne gris, Meis povres dras de petit pris ; Meis Deus ki est riches d'aveir Lui truverat sun estuveir. II ad tute la nuit erre ; 650 N'est merveille s'il est lass^: N'ert pas a us d'aler a p^ ; 1633, LOV Si assae; OV avawnt. — 1635, ki plus promettrey- ent, V ke uns pmett'unt. — 1636, V cliif ; L deccuerint, O -eyent, V decevemnt. — 1637, Teus iuwent, V Teus venint. — 1638, L Al, V A paraler. — 1639, L aiinent. — 1640, V bosoign, LOV al. — 1641, C'est for Ms. ceo est; L le ; LV mein en mein. — 1624, LO ai(y)m — 1644, Deske vus eez, V Deskes. Manuscript Readings 641, mise. — 649, tut. —650, si est. 1 Edition, Paris et Bos, pp. 20, 38, and 46. 158 MANUAL OF OLD FRENCH DLILECTS Ne pur kant mult s'est efforce, Kar la gent dute de sa terre Ke nel sivent e facent guerre. 655 E il si funt, plusurs parties Vunt lur curlius e lur espies, Querent a munt, querent a val, Ambure a pe e a cheval ; Mei est avis k'en vein le funt: 660 Ja de lur olz meis nel verrunt. Entre le Eodne e Munpellers 1230 Ert le pais large e pleners De granz deserz e de boscages; Assez i out bestes sauvages, Urs 6 liuns e cers e deims, Senglers, lehes e forz farrins, 1235 Olifans e bestes cornues, Vivres e tygres e tortues, Sagittaires e locerveres E serpenz de mutes maneres. Gires n'en prent nule pour, 1240 Einz se fie en sun bon seignur. El bois entre ki mut fud grant E veit le Eodne costeant : Or en penst Deus par sa merci, Car pur s'amur ad tut guerpi ; 1245 Se il n'en prent de lui conrei, Ne mangera, car il n'at quel : Ne porte od sei ne pain ne vin Dunt il se digne a eel matin, Ne tant que vaille un hanetun 1250 Entre vitaille e guarisun. Manuscript Readings 654, Ki. — 656, Unt. — 1230, &. — 1231, grant. — 123G, Urs e t. 1243, peust. — 1247, nine,— TEXT SELECTIONS 159 Seigneur, oez \\n bel miracle : Hoc u ert en s'abitacle 1505 E en sa loge u il uroiit E nostre seignur depreiout, Si vit une bisse sauvage Tut dreit errante a I'hermitage. La bisse fud durement bele 1510 E vint tut dreit a la venele Par la sente k'ele trovad : Entre les branches se musQat, Ne dutet pas, meis dreit enz veit. Gros out le piz e plein de leit : 1515 As pez Gire se veit gesir, Presente sei de lui servir. Gires ad la bisse veue Ki a ses pez est estendue : Mult se feit lez, kar ben suschad 1520 Ke Dampne deus lui enveiad. Tant cum iloc el desert fud, Del leit de la bisse ad vescud. Or escutez cum el le sert : Le jor veit peistre enz el desert; 1525 Quant vent a I'ure de disgner, Ne I'estot pas pur lui aler : Ele set ben le terme e Pure, Si sacliez bien plus ne demure K'el n'en venge dreit a la fosse ; 1530 Ele fud bele e grasse e grosse : N'i out si bele en la contree, Ne ne serad ja meis trovee. Gires li feit a une part Manuscript Readings 1505, E wanting. — 1507, E si. —1509, f. mut d. — 1518, estendu. — 1519, lez wanting. — 1521, iloc wanting. — 1523, ele. — 1529, Kele. — 1530, gras, — 1531, Nout, centre. — 1533, le f. 160 MANUAL OF OLD FRENCH DIALECTS Une logette en sun essart 1535 U gist la nuit pur la fraidure ; L'endemain veit a sa pasture. De tel conrei cum jo vus di S'est li sers Deu vescu meint di : Quant il ad pris tant cum li haite, 1540 Nen ad messaise ne suffreite. VIE DE SAINT THOMAS 1 FEUILLET II Li apostoilles ad grant angoisse, Suz ciel ne siet ke faire puisse. E li messager [an] cuntraire, 4 Ore funt un, ore funt el traire.^ Mut avroit en quor pesance^ S'il curuQast le roi de France E des Franceis la commune. 8 Ceste est de dous parties la une, Mes ceste requiert dreiture, Lei e leaute e mesure. D'autre part li rois Henris, 12 De terre, de aveir poestifs, Ove ses evesques e ses clers Est tut encuntre e en travers. Si li rois ne se entremette 16 Lowis, ne siet quant pes i mette, Kar mult est de descreciun, De mesure e de grant resun. Manuscript Readings 1638, uesqui. 1 Edition P. Meyer, p. 11. 2 " Corr. 07' dans I'un des deux cas." 8 " eii ou en ?" TEXT SELECTIONS 161 BOEVE DE HAUMTONEi XL VI Li emfes vint devaunt le emperur a vis fer, 290 harcliement commenqa a parler : " Entendez vers moi, beau duz sire cher, ky Yus dona conge cele dame acoler ? Ele est ma mere, ne yus enq?(er celer, e ka?mt a moi ne volez conge demaunder, 295 jeo v?