EXAMINATION OF THE LANGUAGE OF THE SCRIBE

The inferences concerning the language of the scribe are to be drawn from the forms unprotected by the rhyme and mètre and especially from those cases where they differ from the forms ia rhyme. In our poem sufficient uniformity occurs in the speech characteristics to show clearly the dialect of the last rédacteur.

(a) Phonology. â free gives e ; Bartsch-Mussafia Law completely observed ; -âge (-aticum) 216 (so also in rhyme), and -ac/ie, 195; -ier (-arium) ; vassaus, 722 (i. e. vassal, not vassel); -iee never reduced to ie ; eve (aqua), 347, 370.

É free gives oi ; ei atonie gives oi ; cortoisie, 933, damoisele 270, etc. ; but neiant 318, 590, 1165, beside noiant 723, 827 ; fres 644 (in rhyme /rots 474) ; -el final remains 197, 280, etc. ; bia.u 230 is built on biaus ; -el before a consonant becomes -iau 85, 155, 255,^ 261, etc.; pronoun els gives aus 169, 708, 817, 881, 899; solaus (soleil) 60 ; -uel -{- cons. gives iau ; iauz 643.

o free gives ue; hom (not heum) 190, 620, etc. ; fors (not fuers} 1173, etc. (also in rhyme); pou 268, 1166, and poi 973 (also in rhyme) ; bloi 641 (also in rhyme) ; leu 730, 848 ; feu 73, 78, 80, 107, 117, 230 ; jeu 654, 919, 956, 1041, 1047 ; jeuent 804.

sol 586, 628, 929, 983; seus 88; solement 33; seulement 777; eure 88 ; preu 25, 987, 1005 ; 308, 382, 995; deus (duos) 730.

^ Cf. infra.

ty-

52 Language of Scribe.

o checked, (1) remains 13, 21, 34, etc., etc., or (2) gives ou 1180. Pretonic o, (1) remaÎDS 25, 39, etc., etc., or (2) gives ou 222, 436, 648, 740; 436, 809, 812. The writing jalous : vos 289 shows that this was the transition time between o and ou. UBi gives (1)

0 94; (2) ou 816, 889. aut gives (1) o 186, 339, 214, 802; (2) ou 77, 313, 802.

é -\- l mouillé gives ell, mervelle 1103; pretonic e -{- 1 mouillé gives ell, mervellier 166, 494, 614, 982; vellier 115, 120, 681; aparellier 37 ; mellor 748, 763, 992. 6 -\- 1 mouillé gives -oill, -oily Cardoil 30; voU 284, 504, 507, 592, 792, 1092; vaille 284, 387. Pretonic o -\- l mouillé gives oill-, moillier 991 ; recoillier 1172.

é before n or n mouillé gives a^, talng 1037, /am^e 1186; but preigiient 381. Pretonic e before n mouillé gives ai, ensaignier 9, 205 ; saignor 8, 277, 447, 772, 830.

en often written an ; gênerai in œiui, anuier 279, 637, 669, 707, €tc. ; asanble 262 ; ensanble 475, sanblant 223, 342, 380, 465, 559 ; talant 606 ; manace 576, 653 ; but entendre 2 ; aventure 3, etc. ; -ment (mentem, mente) not -mant 57, 91, 92, etc: ; feme, not famé 990, 1097, 1101, 1109, 1172 (;.same 1169); anor (honorem) 287, 411, 588, 764, 844, 1177, 1179; en (= on) 18, 236, 241, 304, 347, 418, 419, 420, 463, 1169, 1185. The nasal before labials is written n 203, 380, 476, etc.

ai is written ai and e. The confusion is complète ; cf. trere 14, tere 16, fer e lSd,fet 183, meson 121, faire 6S7, fait 223, plaise 145 ; afaitement 156, raison 19, etc. Even final ai is written e; e (habeo) 614, conté (prétérit) 613, doteré 412, etc., ai is also written ei; cf. leissier 199, beside laissier 1084, 1104, etc., and lessier 201, 1080, etc., pleisir 587, heside pkdsir 580.

al -\- eons. gives au passim; but v lait 283, 373, 964, 966, 986, 1172, hesiCie viaut 293, 563 ; and valt 1109, 1167, beside vaut 1166.

01 + cons. gives (1) ou, doucement 51, 429, 652 ; couchier 514, 690 ; voust 1145 ; (2) o, cop 616, 1134 ; docement 171, 397 ; cochier 516, 813; decoper 39; vost 403. ul -\- cons. gives u, hus 180, 190, 213.

gua reduced to ga, garde 525, etc. c before â gives ch. ndr (from nr) 14, 761 , 1077, etc. (also in rhyme) ; but nr (from 7ir, ndr) 378, 943, 1073, 1149 ; Idr (from Ir) 12, 757, 761, 976, 1079, 1171.

n mouillé written gn after tonic vowel 2, 381, 623, 624; written gn before tonic vowel 8, 9, 100, 111, 205, 277, 283, 447, 510, 772, 830, 873, 955, 1056, 1189; written ngn, estaingniez 505, plaingnoient 167.

s for 2, fruis 443, girois 499, aprimies 596, bras 519 (cf. b7'az in rhyme 510), solaus 60, Ace. Plu. Poss. Pron. vos 724, 740.

desduit 280, but (^e(i^t^^ 437, 788, 811 ; déduire 281, 371 ; efforça 360, 363; meffet 1137, but desfenda 328.

Language of Scribe. 5â

SI gives se 183, etc., etc., before vowels and consonants; sic gives si 2, 61, etc., etc., before vowels and consonants.

Penre for prendre 1149. ids (Jmis) 513 (but us in rhyrae). REM gives both rien 631, 632 ; and î^iems 650 (and 1038 in rhyme; 7-ien oblique in rliyme fréquent 185, 310, 340, etc.) ; nom. riens 1185. Artu 20, 29 (author^s ïorm Artur 743). donique gives donc 967, 1016 ; dom 1083.^ de unde gives don 256, 323, 544, 719, and dont 486, 715, 797, 1106.

(b) Case flexion. Nom, Sing. Masc, subject and in direct address, the flexional s is regularly preserved ; the only exceptions are as foUows : ostel 249 ; H oste 360 ; l^oste 455 ; mon 'saignor 830 ; -âge — sage 135-136 (cf. -âges — sages 571-572) : cest sanc 713 ; jeu parti 956 ; tlie nouns in 31-32, Gaiwain, Keu, seneschal, Vvain probably seeraed to the scribe to be oblique ; sol 585 perhaps attracted by the plural idea. siî'e, the historié form, is used in the combination mes sire; elsewhere, 223, 1078, it has assumed s. Nom. Sing. Masc, adject. after êtrey with s passim and 163-164, 345-346, 775-776, 685-686, 699-700; withouts, délivre 702, preu 987, 1005, 221-222, 249-250.

Ace. Sing. Masc. in a few cases has s; nus 180 (separated from its verb), 213 {but nul 254, 280, etc.); miaudre 761 (cf. mellor 748, 993, and in rhyme 763) ; biaus 547 ; hère the scribe was evidently influenced by the nominative pères imraediately preceding. biau 230 is an oblique form built on the Nom. biaus (cf. bel, passim).

Nom. Sing. Neut. : the regular neuter form miaus is used 588. In the two other passages where the neuter occurs, esprovez 762, d^^oiz 967, it falls in with the Masc. and takes s,

ne Vuns ne l'autres 680, nu a nu 585, sol a sol 628 do not absolutely comply with the rules of gender and case, but this is explained by their being set phrases. Parallel cases occur in Chrétien de Troves, who observes the rules of declension ; cf. Eric et Enide, 11. 4689, 5248.

Nom. Plu. Masc. There is no confusion by the scribe with the oblique form; even in 1196, where the author must hâve written ses amis, since it rhymes with pa^ls, the scribe has restored si ami in spite of the rhyme.

Ace. Plu. Masc. naturally has s regularly. tôt jorz 34 is probably a mère clérical slip; cf. toz jorz 67, 68, 135, 202, 569, 635. moût 179 may be taken as Sing. meaning '^ many a one."

Part, after Ace. Plu. Masc. -j- avoir has s, saluez 157.

A comparison of the observation of the rules of declension in rhyme ^ and in the body of the poem shows at once that for the

^ Foerster, l. c, p. Lxxiir, says dom is the prevailing form in manuacript A of Chrétien's works. Ms. A is a west-Champagne ms. 'Cf. supra, 'Case Flexion of Aulhor.'

(54 Language of Scribe.

author the nominative and oblique forms are almost completely confused, while the scribe, in an overwhelming majority of cases, bas maintaiued the distinction. In gênerai, where the oblique form used by the author is protected by rhyme, the scribe lets a directly modifying adjective or auy form closely connected in position and construction maintain its agreement with the nouu. But even in spite of this close union, he sometimes changes the first form to couform to decleusion rules; e. g. 194 Mes tant iestes biaus chevalier: chier, oblique; 622 Gauvains remest pensis et morne: tome ; 833-834 si parent et ses amis Quidoient bien qu^il fust ods, When we recognize the fact that the scribe observed the rules of declension better than the author, it explains some seeming violations of thèse rules by the scribe as cited above : thus the scribe left preu 987, 1005, and bons 869 because they were closely linked to other adjectives to which he could not restore the s because they were in rhyme ; in 360, 455, 702 to hâve restored the Nom. forms would hâve given thèse lines a syllable too many. We are safe in assuming that the scribe uniformly observed case flexions and made it his ruie to restore them where they had been neglected by the author, failing to do so only in the very few cases where we find violations outside of rhyme syllables.

(c) Article. Nom. Sing. Maso, li passim, V 388, 471, 721, 746, 874. Aœ. Sing. Masc. lo 48, 83, 117, 120, 156, 210, 229, 353, 418, 454, 474, 478, 552, 562, 641, 643, 644, 651, 657, 808, 820, 921, 992, 1042, 1065, 1101, 1122, 1137; 963; lou 50, 78, 97, 107, 109, 198, 209, 220, 226, 303, 353, 363, 444, 535, 562, 566, 602, 641, 646, 654, 750, 782, 816, 891, 904, 1061, 1104, 1126.

(d) Personal pronoun. je passim; jo 507, 531 ; / 327, etc. ; g^ 143, 942, 950. voSj nos passim.

3 Fers. Conjunc, Obi. Masc. Sing. (1) lo; referring to a person 167, 187, 238, 315, 387, 460, 472, 657, 875, 1139, 1197; lo me 759; referring to an animal 234; abstract 114, 421, 723, 727, 746 ; lo me 723 ; lo vos 947 : (2) lou ; referring to a person 67, 97, 322, 340, 381, 397, 414, 468, 549, 551, 562, 900, 996, 1118, 1134, 1139, 1143 ; abstract 247, 417, 578 : (3) le; le vos 336, 488, 947, 1045 ; following its verb, por prendre le 564, only case where the pronoun follows the verb.

ïhus we see that the conjunctive pronoun is lo or lou, weakened to le only after the verb or when it précèdes vos. In the latter position, one case occurs of lo.

3 Pers. Disjunc. Masc. lui passim, li 99, 1148, 1152. 3 Fers. Nom. Sing. Fem. ele and el, both passim.

(e) Possessive pronoun. Nom. Plu. vo 1035 ; Ace. Plu. vos 724, 740; Sing. vostre passim. soe (not suoe) 251, etc.

• Language of Scribe, 55

(f ) Relative Pronoun. Note cui oblique 143, beside later qui 827.

(g) Verbs. Endiiig of 4 Plu. (1) -ons 181, 406, 932, 977; (2) -071 175-176, 188, 960. 4 Plu. Près. Subj., soions 932. 4 Plu. Imperl". Indic. does not occur. 3 Sing. Imperf. Indic. 1. -oit (not -ot as in rhyme) 636, 967-968, etc. être : 3 Sing, Imperf. Indic. uudiphthongized, ert 9, 25, 263, etc., etc. ; the Fut. is diphthongized 195, 289, 626, 942, etc.; 5 Plu. Près, is diphthongized 194, 571, etc. ; but once uudiphthongized 877. aller ; 3 Sing. vait 782, vet 909, 911, va 153, 389, 467, 874, 1019, 1146, and iu rhyme. eussey seûssey etc. 201, 401, etc. ; but poïst 320, beside peussent 755.

Ending of Second Person Plural. (Près. Indic.) -ez 87-88, 144, 146, 193, 364, 438, 533, 558, 593, 594, 673, 736, 752, 845, 924, 928, 937, 945, 953, 997, 1088, 1089, 1095, 1113, 1159. After a palatal, -iez 1112. iestes^ estes, faites 194, 344, 501, 708, etc., etc. (Imperative) -ez 134, 146, 245, 448, 500, 618, 722, 744, 938, 939, 998, 1189. Affcer a palatal, -iez 244, 617, 1072. sachiez (Schwan, § 420) 242, 452, 542, 567, 614, 698, 703, 866, 988, 1003, 1087, 1093, 1097, 1098. soiez 448, 744, 938. (Future Indic.) -ez 149, 933, 1073; -oiz 114 340, 509, 637, 769, 976, 1091; -o/s 499. (Présent Subjunctive) -62 337 ; -oiz 611. After a palatal, -iez (505), 955, 1156, 1032, 343. (Imperf. Subjunct.) -iez 163, 570, 706. (Imperf. Indic. and Conditional) -iez (2 syllables) 574, 609, 935, 1040; 341, 342. (Prétérit Indic.) Sing. fus 160; Plu. -stes 440, 724, 726, 1041, 1042.

(h) CoNTRACTED FORMS. For se lo, sil 469. For si lo, sel 106, 599, 1088; and sou 105. For ne lo, (1) nel 168, 188; (2) nou 971 ; (3) no 239, 412, 458, 955 ; 192, 284, 293, 1159. Note no = ne once 760 ; elsewhere ne, For je lo,jou 609. Note^o =je 607, 531 ; elsewhere je. For qui loy quil (constructive form) 185. For ne les, nés 15, 1071, 1112. For je les,jes 1037.

The full forms occur in the following cases : si lou dites 245, si lou dient 416, si lou feri 1118, ne lou prendra 549, je lo désir y 420, qui lo contredit 727. No cases occur of the full forms for se lo, ne les, je les, a les, en lo. Chrétien de Troyes uses sel (si lo), nel, jel, nés, jes, and also the uncontracted forms, cf. Eric et Enide, 1. 2518, Je le, etc.

For en lo, (1) el 158, 210, 872; (2) o 990, 1055, 1135, 788, 830, 1021 ; 0 pas petit 152. o lit revenue 515 ; (3) ou 666. For a lo, (1) au, passim; (2) al 204. For a les, as 804, 805, 1133. Notice 11. 216, 356 where a, introducing an infinitive phrase, coalesces with a following article ; au herberjage deviser ; as mes aconter.

(i) Hiatus and elision. Hiatus occurs 413, 515, 666. It occurs also in Chrétian de Troyes, cf. Yvain, 11. 1666, 1891, 2384, 6138, etc.

66 Language of Scribe, •

Elision is facultative in: (1) si (sic); elided 479, 773; not elided 2, 41, 347, 521, 586, 606, 711, 762, 870: (2) se (si); elided 195, 200, 283, 286, 296, 317, 322, 401, 405, 467, 553, 559, 825, 964, 966, 1052, 1112; not elided 183, 214, 311, 380, 431, 563, 612, 636, 675: (3) que; elided 58, 224, etc.; not elided 49, 54, etc. : (4) je; elided 327, 942, etc. ; not elided 147, 864, etc. : (5) ne (non); elided 130, 181, etc. ; ne (nec) ; not elided 1144, etc.: (6) ce; elided 150, 1074, etc., not elided 590, 1165, etc.: (7) H Art. Nom. Siug. Masc. ; elided, l'en 18, 304, 418, 419, 420, 463, rosfes 388, 471, 721; l\ms ne Vautres 680; Vostel 249 was probably written by the author for lo-\-ostel; not elided, li uns 108, 159, 174; li ostes 263, 372, 426, 435.

(j) Tutoiement and vousement. Tutoiement : shepherd to Gauvain 160 (in rhyme); Gauvain to shepherd 182 (in rhyme). Vousement; shepherd to Gauvain 163, 176, 191-195 (assured by rhyme) ; Gauvain to shepherd, none. Elsewhere (Gauvain ta castellan and vice versa, Gauvain to maiden and vice versa, G. to knight and vice versa) vousement. *

The above exami nation of the scribal forms shows clearly that the manuscript belongs to the Champagne dialect. This is at once made évident by noting the principal points of resemblance and of opposition to the Champagne as characterized bv Prof. Foerster* and M. de Wailly.^

Resemblances. a gives e ; Bartsch-Mussafîa Law observed ; -âge beside -ache (cf. Éric et Enide^ 1. 1006, damache : sache) ; -ier ; -iee never reduced ; E gives oi ; -el -\- cons. gives iau; -uel -{- cons, gives -iau ; o gives ue ; seus ; sole and seule (Champ, sole) ; eure, preUy deus (duos) ; o checked and atonie o generally remain ; atonie e -\- l mouillé gives ell ;^ atonie o -^ l mouillé gives oill ; e -\- n mouillé gives aign ; atonie e -\- n mouillé gives aign ; al -jcons. regularly gives au (but isolated valt, vialt, not found in Chrétien) ; ol -\- cons. gives o or ou (Chrétien, o or ol) ; id -f cons» gives u; gua- gives ga- ; Idr, ndr^ isolated nr (Chrétian, Idr, ndr) ; isolated writing ngn for n mouillé ; "* se for si ; si for sic ; sel for si le; poisse and peûsse (Champ., poisse) ; donique gives donc and dom.^

^ In his introduction to his édition of CUgès (1884).

^"Mémoire sur la Langue de Joinville," and "Mémoire Bur la Langue de Keims."

^ This is the form used by Foerster in the large and the sn)all édition of Erie et Enide, which he considers the work of Chrétien containing the most Champagne traits. Cf. the édition of 1896, p. XLi.

* Mémoire sur la Langue de Joinville, p. 277, and s. v. -aingn, -angn, eingn, -oingn, pp. 435-457.

^dom is the prevailing form in the west-Champagne manuscript of Chrétien de Troyes ; cf. Foerster, /. c, p. lxxiii, § 25.

Language of Scribe. 57

Tlie point noted above that the scribe observeci case flexion much better tlian th:^ author accords with the examination niade by the Marquis de Wailly of the chartes of Jean, Sire of Joinville in Chanipaojne. Tlie chartes he examined are ail dated and are embraced in the period betwcen 1239 and 1315. Five of the chartes are of ihe foiirteeuth century. Ont of 837 occurrences of the nominative singuhir masculine, 835 hâve the flexional s} This furnishes conclusive évidence that at least in some parts of Chamnairne the rules of declension vvere broken down much later than in the Ile-de-France.

noSy vos are the forms employed for the personal pronoun.

3 Sing. Imperf. Ind. 1, -oi/, not -ot; vaii 'dudva; the individual words vassaus ; eve ; fres, not frois ; aus (= eux) ; hom not huem ; JorSy notfaers ; soe, not suoe-, leu, feUyjeu', anor.

A certain number of variations occur frojn the language of Chrétien de Troyes. Thèse, however, may be brought under three heads :

1. Forms that while not occurring in Chrétien^s works, or occurring there as the exception, are Champagne forms. Such forms are most liable to corne up in Eric et Enide, since it shows, as his earliest preserved work, the strongest marks of his home dialect. In his later works he gives évidence of Ile-de-France influence.^

s for z occurs in isolated cases in our poem. The sarae is true in Eric et Enide.^

Isolated /ir-forms : this is characteristic of Champagne.*

As regards /o, loti for the article and pronoun, though Prof. Foerster "^ considers such forms occurring in mauuscripts of Chrétien as due to Eastern influence and substitutes le, yet thèse Eastern forms occur as far west as Ile-de-France^ and might therefore easily be the prevailing forms with a Champagne scribe.

The shorttned form ?io, Poss. Pron. Nom. Plu., since it is assured by the rhyme, belongs to the author ratïier than to the scribe.

The forra -0i7, -oiUe from o -\-l mouillé is not the characteristic Champagne form, but belongs to Picardy and Lorraine.^ It comes up, however, in one of the Joinville Chartes^ beside -uel and -euil, bat this is a charte showinîî stron^: Eastern influence.

^ Mémoire sur la Langue de Joinville, p. 229. 'Cf. Foerster, Eri.c et Enide (1896), p. lxii. 3 Cf. 11. 2250, 3711, 8872, 4924, 5975.

* Foerster, C/iV/ès ( 1884), p. lxxiv, l 29. ^l. 0., p. i.xvri, ^ 21,

« Metzke, A. N. S. L, lxv, p. 80.

' Cf. Matzke, dia/ectische JiJùjenthuemlichkeiten der mouillierten X, p. 86.

* VVailiy, Mémoire sur la Langue de Joinville, p. 350, 29.

58 Language of Scribe,

li occurring sporadically for lui, disjunctive masc, is an Eastern characteristic ^ that we must assume to hâve affected Champagne, as is natural enough.

2. Forms that seem to hâve belonged to the author and to hâve been left by the scribe, el fréquent beside ele — a common form in Ile-de-France.^ The el then may be an Ile-de-France form of the author that the scribe has left because to change it would spoil the mètre.

Sporadic -ez, beside customar}^ -oiz in the Future and the Présent Subjunctive also probably represents the original writing that has escaped scribal change.

3. Forms pointing toward the Ile-de-France and thus indicating that the scribe belonged to West-Champagne or else marking the spread of Ile-de-France forms in the dialect. While e before n is frequeutly written a, this is not uniformly the case as in Chrétien de Troyes ; so also the indefinite pronoun, while weakened from Ile-de-France on, is written en, not an, and the suffix is written ment not niant.

For pretonic ei we find the Francien oi.

The writing of touic e -\- l mouillé, in the one place where it occurs, as -ell is probably due to the scribe's accepting the French -eilt but writing it phonetically, unless it be a parallel form to atonie -ell, or by Picard influence.^

The first person Plu. of the Subj., in the one place it appears, has the Ile-de-France form and not the characteristic -iens.

The gênerai homogeneity of the forms throughout the poem indicates that the language of any scribes who may hâve retouched it was not strikingly removed from that of the author. It also indicates that between the Francien author and the Champagne scribe or scribes there intervened no scribe belonging to another section. There are also signs that the scribe was very conservative in making changes. Thus though he restored in gênerai the neglected case flexion, he did not do so when it involved any radical change, and seems to hâve shrunk from it even when it involved merely a complète change of form, as miaudre to mellor, mon saignor to mes sire.

^ Cf. Apfelstedt, Introduction to Lothringischer Psalter, p. Lxvil ; Goerlich, Frx. Slud., Vil, p, 123. Note also li in rhyme, Roman de la Rose, vol. IV, 11. 17490, 18162, 20648.

'Cf. Rustebeuf, no. 1, 1. 90; 40, 89; 40, 131 ; 40, 138, etc.

3 Cf. Matzke, L c, pp. 69-72.

Episodes, 59