“S “ “Ahh
> im) /// nea < Z
TO
TOMO
MI
-eweee
, *
MOM
oe
| = ~—s i= — - _ — —_ i— — — —_ ——-:= ~ tall =. -
SLOREROTROLERODORAONODEDDD "UAE TA,
UE
eg wrrrs
yah ii OMG GU av 0 eNO UA
ye
ITM Pe
«
U -- ’ _
TITTLE,
“pesCOUODEDDAGERERIENUEEUL REE TERS ED DEAGOP DUO EE
the
TELL LLL PLL
Avg “ iis
ar) rs eres / < wey, . HE paper UTTNTE
VEeeeeedahs tee hewenkeaeriebededed vas
‘=
= = = = .— ‘= — = (= ‘= ‘= + ‘= = = ‘= = —— __= ——) = dl = ss _ = .a 4 ‘_— — — oo - _— ‘= Sew = ~ - = = = ew oe — ‘> = ‘oe = = -
t
FROM
Pern ey tT te
ibis
see
al
Digitized by Google
Original from
UNIVERSITY OF MICHIGAN
Digitized by Google
Original from
UNIVERSITY OF MICHIGAN
LE VENGEMENT ALIXANDRE
Accepted by the Department of Modern Languages,
September 1924.
A CLASSIFICATION OF THE MANUSCRIPTS
OF GUI DE CAMBRAIS
VENGEMENT ALIXANDRE
A DISSERTATION PRESENTED TO THE
Facutty or Princeton University IN CANDIDACY FOR THE DEGREE oF Doctor oF PHILOSOPHY
BY
BATEMAN EDWARDS
_ PRINCETON
1926
co we A 37S” ie a ay
Original from
Digitized by Google | UNIVERSITY OF MICHIGAN
PREFACE
For Gui de Cambrai’s continuation of the Roman d’A- lerandre | have adopted the title Le Vengement Alixandre, thus differing from previous writers on the subject, who call the poem either La Vengeance Alexandre or La Vengeance d’Alexandre. In the first case we have modern orthography coupled with medieval construction, and in the second de with an objective genitive after vengeance is not in accord with modern usage. The latter objection applies equally to Gréber’s cumbersome title,! La Vengeance de la mort
d’Alexandre. It is Gui de Cambrai himself who furnishes the solution. In the third stanza the author, after giving his name, indicates his theme :
El non al vallant conte a cui Clermonsapent | Et por Symon sen frere, saciés seiirement,
_ Sont cist ver ici fait qui ci sont en present. Guis de Cambrai les fist en lor tesmognement Qui por ceste oevre ara gueredon bel et gent ; D’Alixandre viut dire et de son vengement.
This seems a definite statement of. title, and has led to my choice of the form Vengement for the poem of Gui de Cam- brai, as over against Venjance, the term employed in the explicit of Jean le Névelon’s work.
In citing passages from the Vengement Alixandre, | follow the numbering of lines which I shall adopt in my edition of the poem. For the Venjance Alixandre, I use the numbering of Professor Schultz-Gora in his edition. In speaking of the Parma manuscript, * 1 make my references by stanza and line,
4. Grundriss, Vol. II, part 4, p. 584. 2. See infra, Chapter 11.
VI PREFACE
but wherever possible I supplement these by giving also the corresponding line in the Vengement or in the Venjance. One of the preliminaries to an edition of a text is to seek such light regarding the readings of the original version as may be obtained by establishing the interrelation of the manu- scripts. When, however, can we be reasonably sure that we have some guide to the author’s own wording ? This, for the Vengement, is a question to which I seek an answer in the — present study. As one essential step we shall have to weigh
certain evidence which might seem to be in conflict with any and all schemes of manuscript classification. To this evidence, which may be called negative, I devote the first section of my third chapter, where I try to simplify the problem by a con- sideration of some of the mental processes of scribes engaged in copying a text.
Instead of the term laisse, which does not lend itself to adop- tion in English, or of tirade, which as an English word has unsuitable connotations, | employ ‘stanza’, which does not entirely correspond, but which can be conventionally extended to include /aisse without undue violence to its meaning.
Chapter [. Chapter II.
Chapter III. Chapter IV.
TABLE OF CONTENTS
Delimitation of the Problem....................... 4 The Amalgamated Version of the Vengement Alixan-
dre and the Venjance Alixandre................ 4 Manuscript Relations in the Vengement Alizandre... 413
Remarks on the Establishment of the Text......... 47
Google ;
ivy Google
CHAPTER I
DELIMITATION OF THE PROBLEM !
The history of Alexander the Great, in its legendary form, was widespread in medieval Europe, in both Latin and vernacular versions.’
For the French the Roman d’ Alexandre is the most signifi- cant monument. In its termination, however, this work, judged by the standards of the time, was incomplete. When the story ends, Alexander has been poisoned by Divinuspater and Antipater, and has been given fitting burial by his peers. Concerning the murderers we know only that after the scene of the poisoning they fled par deviers le marine.
- This incompleteness in disposing of the loose threads of the
story was noticeable to such an extent that we have two accounts of how the peers went in search of the traitors, overcame them and put them to death:* one composed by Gui de Cambrai,® the other by Jean le Névelon.§
Of the latter version I shall have occasion to speak only by way of comparison with the former. My purpose in this study is to clear the ground for an edition of the Vengement Alixandre of Gui de Cambrai. With this in view I shall study
4, For explanation of special abbreviations employed, see Bibli- ography.
2. See AlGr for a discussion of the sources and diffusion of the legend in the French territory; also Rom., 14 (4882), pp. 243-332.
3. Mich., p. 529, 38.
4. Up to the present time no Latin source has been discovered for ° either of these continuations. \
3. For discussion of the identity of this Gui de Cambrai with that of the author of Balaham et Josaphas, see Appel, pp. xLIv-xLV11; ; Armstrong, EM, 19.
6. For the form of the name ‘to be adopted, see Schultz-Gora, pp. 4-5, and Walberg, Rom., 32 (1903), p. 155; also FF, 1906, p. 22, note 4.
1
2 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
which has until now been imperfectly known; next I shall
attempt a classification of the manuscripts of this version;
finally I shall add some preliminary remarks on the establish-
ment of the text in view of an edition. | The Vengement is contained in eight manuscripts,! all
of which attach it to the end of the RA. In referring to these
manuscripts, the list of which follows, I shall use the letters
applied to them by Meyer.
D — Paris, BNF 13094, f° 272 r°-299 vo. —
F — Parma, Bib. Palatina 1206, f° 173 r°-190 v°.
G — Paris, BNF 25517, fe 284 r°-343 r°.
H — Paris, BNF 786, f° 84 v°, col. 1,-94 v°, col. 2.
I — Paris, BNF 375, f° 241 r°, col. 2,-216 r°, col. 4.
J — Paris, BNF 24366, pp. 221, col. 1,-242, col. 4.
K — Paris, BNF 792, fo 144 r°, col. 1,-151 r°, col. 4.
L — Paris, BNF 789, f° 95 v°, col. 1,-103 r°, col. 2.
The date of the poem is anterior to 1191, as hasbeen estab- .
lished by Meyer.’ Meyer has likewise described seven of the manuscripts of
the Venjance Alizandre of Jean le Névelon® as has also —
Schultz-Gora.* Since they had not examined closely manu- script F, as had neither Sachrow nor Walberg, they omit men- tion of it in this connection. Study of the manuscript, however, reveals it to be a fusion of this version with that of Gui de Cambrai. I reserve for the following chapter an examination of this fusion. Meyer, in opposition to the views of the other above- mentioned workers, held that the poem of Jean le Névelon was a century later than that of Gui de Cambrai,® his reason
. Described by Paul Meyer, Rom., 44 (1882), pp. 247-322.
. AlGr, IJ, pp, 255-57.
. Rom., 14 (4882), ibid.
. Pp. 7-8. -
. See AlGr, II, pp. 263-64; also Rom., 32 (1903), p. 150, note 2.
OF 1B GS DO
#
in some detail the literary procedure and importance of one +” manuscript which contains portions of the Vengement, and
DELIMITATION OF THE PROBLEM 3
being apparently that the Venjance occurs in none of the earlier manuscripts. This argument does not hold, for we shall see! that the composition of the Venjance antedates manuscript J, which Meyer himself ascribed: to the middle or the second half of the thirteenth century. J contains at the end (p. 242) a reference to the translation of the relics of St. Quentin in the year 1228. Since the relics were again moved in 1257? it seems certain that this manuscript was written before the second translation took place. Since both the Vengement andthe Venjance were written before manu- script J, it is at least possible that the poems were contem- poraneous. I find nothing in the language of the Venjance to argue against a date as early as that of its rival.
Although the general subject of both poems is the same, neither author shows any indication of having known the other’s version. As I have said, some such addition as theirs to the RA was for the medieval genius almost a necessity, and the argument itself made it impossible to avoid certain points of contact, for instance the series of single combats which are a mark not only of these poems but of the RA and also ofall poems in which military exploits play a prom- inent part. Likewise all names and incidents common to the two versions are found in the RA, which thus could con- stitute an adequate intermediary.
1. Infra, p. 34. . 2. See Acta Sanctorum, Vol. 13, pp. 725ff.
CHAPTER II
THE AMALGAMATED VERSION OF THE VENGEMENT ALIXANDRE AND THE VENJANCE ALIXANDRE
The manuscript of the Roman d’ Alexandre preserved in the Biblioteca Palatina of Parma (Manuscript 1206), beautifully written in an Italian hand of the fourteenth century,! has twice received summary notice: in 1839 by P. Lacroix,? and in 1882 by P. Meyer in his description of the manuscripts ot the RA.? Lacroix in his report attributes the continuation which follows the RA to Jean le Névelon whom he calls Jean le Nivelois, while Meyer credits it to Gui de Cambrai. There is a half truth in each of these statements, for closer inspection shows that we are dealing here with an amalgamation of the two versions.“ My object in this chapter is to examine the pro- . cess of fusion and to judge how far the redactor has been successful in his task.. |
In its present condition the Parma manuscript consists of 190 folios with thirty-seven lines to the page, disposed in nineteen signatures of five sheets each. The beginning and the
4. The Italian origin of the manuscript is confirmed by the orthogra- phy: for example vinti (vingt) 20, 4 [Vengement 364}, and 30, 10 [ Ven- jance 651); cf. also the writing a for mute e: Alizandra 4, 17 [Vengement 16], 6, 10[ Vengement 141], 40,9, 43, 1; parlament 6, 9 [Vengement 140}; fussant 45, 12 [Vengement 195]; puissané 16, 7 [ Vengement 214].
2. Jacob, P.L., Dissertations sur quelques points curieux de l'histoire de France et de l'histoire littéraire, vi1, Paris, 1839, pp. 71-72; also in Migne, Dictionnaire des manuscrits, Vol. II, 1853, p. 969.
3. Rom., 14 (4882), pp. 258-60.
_ 4 Fora discussion of the relations of this manuscript of the Venge- ment with the other manuscripts of the Gui de Cambrai version, see infra, Chapter 11. I shall not here discuss the relations of F with the other manuscripts of the Venjance, but can state that F presents a version of this poem which does not attach itself exclusively to any one of the existing manuscripts or to either of the manuscript groups recognized by Walberg (Rom., 32, 1903, pp. 154-52; FF, 1906, pp. 5- 30). ,
THE AMALGAMATED VERSION 5
end of the manuscript are lacking. The first folio commences with the third line of Stanza 140 of the RA,! and, after the normal termination of the RA with Branch IV, there are inserted two supplementary stanzas found elsewhere only in manuscript D. Thus far F requires little comment. The men- tion of the authors at the beginning of Branch III is lacking, as are also the episode of Floridas and Dauris and the Iter ad Paradisum.? In connection with the RA it is worthy of note that F presents, except for one or two individual variations, the same order of stanzas.as in what is preserved of manuscript T,? which in its turn shows a close affinity for the group of which D forms a part.’ Lacroix has already called attention to a num- ber of lines which occur in the margin and which he char- acterizes as variants. These are, strictly speaking, not variant readings but individual interpolations, apparently inserted by the scribe of the manuscript, designed to supplement the ori- ginal text and, I believe, found in no other manuscript. After the two supplementary stanzas which terminate the RA and without any indication of a new subject, the Vengement com- mences five lines from ‘the bottom of f° 173 r° with the first stanza of Gui de Cambrai’s continuation. It is not however until the end of the second stanza of the Vengement that the explicit of the RA occurs, f° 173 v®: Ici finist Uestoire dou | bon Roi Alixandre, et comence la ystoire de la venjance ge de lui fist Alior son filz et Phyliparideiis, frere Alixandre, et ses doze per et les autres barons d'ambes dos les traitors. At the top of the next folio an incipit reads: Ici comence la venjance ge firent les barons des dos traitres que oucirent le bon roi Alivandre. In this way the dedication of Gui de Cam- brai is cut in two. The manuscript breaks off at the bottom of f° 190 v° with line 1055 of the Vengement, plus the catch-
. Mich., p. 64, 3.
. See AlGr, II, pp. 220-21; Rom., 441 (1882), pp. 247-47.
. See Rom. wv (4882), pp. 344-16,
. This grouping of D and F in the RA finds its counterpart in the classification of the manuscripts of the Vengement; infra, Chapter 11.
ee OO DD >
VI PREFACE
but wherever possible I supplement these by giving also the corresponding line in the Vengement or in the Venjance. One of the preliminaries to an edition of a text is to seek such light regarding the readings of the original version as may be obtained by establishing the interrelation of the manu- scripts. When, however, can we be reasonably sure that we have some guide to the author’s own wording ? This, for the Vengement, is a question to which I seek an answer in the present study. As one essential step we shall have to weigh
certain evidence which might seem to be in conflict with any and all schemes of manuscript classification. To this evidence, which may be called negative, I devote the first section of my third chapter, where I try to simplify the problem by a con- sideration of some of the mental processes of scribes engaged in copying a text.
Instead of the term laisse, which does not lend itself to adop- tion in English, or of tirade, which as an English word has unsuitable connotations, I employ ‘stanza’, which does not entirely correspond, but which can be conventionally extended to include laisse without undue violence to its meaning.
Chapter I. Chapter II.
Chapter III. Chapter IV.
TABLE OF CONTENTS
Delimitation of the Problem....................... 4 The Amalgamated Version of the Vengement Alixan-
dre and the Venjance Aliwandre................ 4 Manuscript Relations in the Vengement Alizandre... 413 Remarks on the Establishment of the Text......... 47
6 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
words mes pris of line 1056 with which the next signature began.!
An analysis of the Vengement in manuscript F will show to what extent the redactor has made use of his two ver- sions. |
The poem commences with an introductory résumé and the dedication to the Count of Clermont and Symon his brother ; Divinuspater and Antipater have fled into la parfonde Grese under the protection of Marinde king of Jacles; the peers have given Alexander sepulchre and have returned to their
. kingdoms; quarrels.arise between them, and they are like to forget the necessity of avenging Alexander’s death (Stanzas 1-5: Vengement 14-5). —- Caunus and Emenidus concert in calling a council of the peers in ‘‘Alexandra”; a messenger is likewise sent to Val Garnie with a letter for Candace, an Eastern queen with whom Alexander has had an amorous intrigue? (Stanza 6: Vengement 6 plus sixteen lines, the work of the redactor, which introduce Queen Candace and thus prepare the fusion with the Venjance). — Short résumé of Alexander’s relations with the Duc de Palatine and Canda- leon, and also with Candace by whom he has had a son, Alior (Stanzas 17-8: original, with slight resemblances in Stanza 7 to Venjance 3 and 4). — Alior learns he is Alexan- der’s son and resolves to.go to Alixandra to be knighted and to undertake the avenging of his father (Stanzas 9-40: Ven- jance 5-6). — Preparations for departure (Stanza 44: original). — Alior sets out for Alixandra and is. met by the peers (Stanza 12: Venjance 30). — Alior is knighted and recognized as seigneur (Stanza 13: Venjance 21). — The peers take coun-
4. There is nothing to be gained from astudy of the rhyme character- istics of this amalgamation of the two continuations, since stanzas which derive from neither one nor the other are too few to support any definite assertions, and the others show the characteristics of the ver- sion from which they are taken. Thus, for example, rhymes in -ant and -ent are generally kept separate in the stanzas taken from the Vengement (-ant: 24, 50, 63; -ent: 3, 68) while in Stanza 22 taken from the Ven- jance they are confused.
2. Soe Mich., pp. 371-83.
THE AMALGAMATED VERSION 7
cil; two youths arrive bringing news of the traitors (Stanzas 14-22: Vengement 1-45). — Preparations for departure; Aris- tote harangues the troops (Stanzas 23-24: Vengement 11- 48). oo - 2 The Greeks arrive at Rocheflor, a city belonging to Marinde and commanded by Antipater’s son Florent (Stanza 25: Ven- gement 20, lines 548-54, plus eighteen connecting lines, the work of the redactor). — Description of the city; Florent is warned by an oracle that his end is near (Stanzas 26-27: Venjance 36-37). — Alior camps his army before Rocheflor (Stanza 28: Venjance 39). — A spy announces to Florent the arrival of the Greeks (Stanza 29: Venjance 29), — Florent - arms and prepares his men for battle,.as does also Filote on the side of the Greeks (Stanza 30: Venjance 31). — Florent rides into the field with thirty thousand men, leaving fifty thousand to guard the city; Alior encourages his men (Stanza 341: Venjance 40). — Alior and Clicon attack; Alior jousts at Florent who is rescued by his men (Stanza 32: Venjance 34). — Battle; Florent’s forces are driven back into the city (Stan -zas 33-36: Venjance 49-52). — Florent disarms (Stanza 37: Venjance 56). — Night past, Alior calls a council (Stanza 38 Venjance 59 plus a reworking of some lines of 60). — Alior seeks aid from Sapient to prevent the army from revictual- ling; a pontoon bridge is constructed, by which the army _ passes the river Clarence; Florent rushes to the attack (Stan- zas 39-42: Venjance 60-64).! — Florent is again driven back; nightfall obliges a halt (Stanza 43: Venjance 69). — Next day the Greeks fill up the moat and enter the city through a breach in the wall; Florent tries to flee but Alior captures him (Stanzas 44-46: Venjance 712-14). — Florent is tied to horses and dismembered (Stanza 47: seven lines, of which the first is as in the first line of Venjance 16, and of which the last four have a resemblance to Venjance 59, lines 1342-46).?
4. F 40 is a combination of Venjance 64 and 62. -2. Stanza 47 occupies the last seven lines of fo 185 v°, and hence stands on the left hand side of the interior of the middle sheet in the
8 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
Tholomé exhorts the troops (Stanza 48: Vengement 27), — Aristé carries the challenge to the traitors and the two forces engage; Caunus asks from Tholomé the right to strike the first blow and jousts at Agolant, who is rescued by his men; Clicon kills Cesaire (Stanzas 49-52: Vengement 28-34). — Guimart unhorses Aristé but is himself unhorsed and taken prisoner by Alior; Aristé infuriated leads his men into the fray (Stanzas 53-54: Venjance 67-68). — Battle; the trai- tors withdraw towards the moat (Stanzas 55-61: Vengement 32-38). — The battle’ continues (Stanzas 62-64: Vengement 39-41). — Exploits of Perdicas; he is taken prisoner by Anti- pater and his men (Stanzas 65-66: Vengement 42-43). — The traitors retire into the city; their messenger reaches Marinde,
last signature.The seven lines at the top of the next folio contain the end of Vengement 27— an entirely new episode with a different rhyme and commencing in the middle of a speech. All continuity is thus broken between the end of one page and the commencement of another. If this signature were incomplete or if the others showed a tendency to varia- tion in the number of sheets they include, the explanation would be obvious that the middle sheet had been lost, but this signature contains the same number of sheets as all the others. This fact however does not exclude the possibility that, in this signature and perhaps also in the one which followed, an extra sheet had been added. We are justified in assuming that the poem reproduced the ending of the version of Gui de Cambrai (see infra, p. 9), and if so the portion which is lost (Vengement from line 1057 to end) contained approximately 750 lines, as we may gather from an examination of H, D, and J, the manuscripts most closely connected with F (see infra, Chapter III). When we add to this number the 740 lines of the penultimate signature, plus the approximately 150 lines contained in the six stanzas of the Vengement which F in its present state lacks, we get more than 1600 lines, of which two signatures of five sheets apiece could include only 4480 lines. It is possible that before commencing the last two signatures the scribe perceived that he could not finish in the space these would ordinarily give him, but thathe had not sufficient material to justify the addition of an extra signature. He could settle this difficulty by introducing into one or both of his last two signatures an additional sheet. Thus F as we have it may lack, besides the final signature, the middle sheet of the last signature preserved. We may suppose that in this lost middle sheet the redactor closed briefly the episode of Florent and resumed the matter contained in Vengement 24-27 at line 717. — When the story reopens, the Greeks, having left Rocheflor, are en- camped about ArondeJ, and the traitors have sent a messenger to Marinde for aid. |
THE AMALGAMATED VERSION 9
who prepares to go to their aid (Stanzas 67-68: Vengement 44-45).
At this point the Parma manuscript ends, and we must go to manuscript J for indication as to the termination of this version. Here, although we have to do with a manuscript containing only the Vengement, we shall see! that for the final stanzas of the poem the copyist had recourse to a redaction other than that from which he derived the earlier and more considerable portion. In these stanzas there twice occurs’ the substitution, characteristic of the F redaction of the Vengement and of none other,? of the name Alior for that of one of the peers. Hence it is assured that the ending in F followed that of the Vengement. If to this knowledge we add the facts that the redactor has fairly well exhausted the material of the Venjance, and that it is sure from the rubric at the end of Stanza 2 in F that Philiparideiis occupied in the latter part of the amalgamation the same central posi- tion he holds in the Vengement, we may consider it probable that in the lost portion of F there were no interpolated epi- sodes from the Venjance and that the argument followed closely that of the Vengement.
It is possible to form some idea of the manner in which the redactor of the version preserved in F undertook the process of amalgamation. He found himself confronted with two accounts of the vengeance taken by the peers upon the murderers of Alexander. In each the machinery of the action is in broader outline the same: the preliminary council of war, the siege of the fortified place in which the traitors have estab- lished themselves, a series of single combats, the promi- nent part taken by a relative of Alexander, the cruel death which is the punishment of the murderers. Our redactor undertook to fuse the two, in order either to produce a work
4. Infra, p. 34. 2. J 66, 11; 68, 10. 3. See infra, p. 40.
10 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
_ of greater length or to reconcile two conflicting reports of events which he considered historical. In any case, his deci- sion once taken, he was confronted with the problem of choosing a basic version. This problem he settled by taking the Vengement for his fundamental story. That this is the case we may be sure, not merely from the fact that the beginning and end of the amalgam are taken from the Venge- ment and that the Venjance occurs more in the form of epi- sodic interpolation, but also because there are far fewer omissions of stanzas of the Vengement, and because F repro- duces with far greater fidelity its substance and wording than that of the Venjance. By taking the Vengement as base, the redactor was able to escape the somewhat tedious account of the summoning of the peers in the Venjance, but free to include all of the more interesting passages. In one respect, especially, the Venjance greatly surpassed in interest the Vengement : the fact that the avenger of Alexander is not his half-brother but his son. Consequently Alior is made the chief figure in the amalgam. To be sure he suffers somewhat of an eclipse in the stanzas taken from the Vengement, but this the redactor has sought to remedy by occasionally substituting, with a fair degree of skill, the name of Alior for that of some other character.! For the rest, the fusion of the two stories was facilitated by the fact that, in both, the names of the avenging Greeks were taken from the RA, and that conse- quently in both we meet, the same characters. It is worthy of note that in the Venjance neither Calnus or Emenidus are mentioned, while they play an important part in the Venge- ment and consequently in those portions of it which F includes. |
The method of approach once decided, how is the main thread of the story in the Venjance to be attached to the Ven- gement? In the former the siege of the traitors takes place at
1. See F 45, 4, 14 (Vengement 181, 197); 23, 14 (486); 59, 6, 12 (888, 894); 63, 14(971); also the two examples of substitution in J cited supra,
p. 9
—
THE AMALGAMATED VERSION 11
Rocheflor and foremost among the defenders is Florent, the son of Antipater. Now in the Vengement the Greeks on their way to Arondel pass by Rochefort. Given the similarity of the names it is not surprising that the joining on of the prin- cipal episode of the Venjance should have been made here, and since Antipater’s capture must be reserved until the end of the story, it is easy to see how Florent his son should have taken his place. Thus throughout this passage, wherever Anti- pater is represented as present at the siege, F inserts instead a reference to Florent.! But this is not the only change that the redactor has introduced in Jean’s story. In the latter there is a preliminary engagement immediately upon the arrival of the Greeks at Rocheflor (Venjance 34-35). F omits all of this except two stanzas : 34 (F 30) describing Florent’s exit from the city, which F places before Venjance 40 (F 31), where, according to Jean le Névelon, Florent rides out from Rocheflor for the second time; and that part of 34 describing the encoun- ter between Alior and Florent which F places after 40 (F 31). Likewise F joins 29 (F 29) to 39 (F 28), and thus makes the spy bring tidings to Florent only after the Greeks have ar- rived at Rocheflor instead of at the time that the peers recognize Ahor as their seigneur. With these exceptions, and the omis- sion of several stanzas or parts of stanzas not affecting the general trend, the episode follows the outlines of the Venjance. Two of the omitted stanzas, Venjance 67-68, are inserted later between Vengement 31 and 32, and serve to give Alior amore prominent place in the mind of the reader.
So far the main episode has been our sole consideration. It was however necessary for the redactor to work Alior into his story. In my analysis of F, I have shown how in the sixth stanza there occur original passages introducing the name
.
1. See F 29, 2, 7, 15 (Venjance 601, 607, 614); 34, 4 (897); 46, 4 (1799). In one case, 34, 14 (1418), Florent is substituted for the Amirant. The reverse of this process takes place in F §3, 12 ( Venjance 1607), which is interpolated in the Vengement after the completion of the main episode of the Venjance; here Antipater is substituted for Florent.
Go ogle eo
12 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
of Candace. In the next four stanzas the redactor combines elements from the RA and from the Venjance with original material to sum up the facts of Alexander's liaison with Can- dace, Alior’s birth. and the discovery of the identity of his father. In the Venjance Alior is recognized by the peers in Val Garnie, but since in the Vengement the meeting of the peers takes place in Alexandria, F makes Alior go there to be knighted, and, in order to link this event in the Venjance with the story, applies the first lines of Venjance 30, descri- bing Alior’s journey to Rocheflor, to his journey to Alexan- dria. .
By means of these expedients the redactor has succeeded in giving a general unity to his amalgam. At times traces of the fusion are not completely obliterated. For example in F 49 Vengement 2 it is from Tholomé that Caunus demands the nght to stnke the first blow in the fray, instead of from Abhor as would normally be expected. But, leaving aside minor discrepancies, we may savy that the version incompletely preserved in F offered a story logical in its exposition and development. and having the virtue of contain- ing the best points of both versions from which it was taken.
Aside from its intrinsic interest in showing the process and result of a fasion which. if we may judge from the testimony of J. enjoved at least a moderate success in the half century which followed the composition of the two poems, F has but a mediocre value for the reconstitution of either text. This value is almost nil for the version of Jean le Névelon by reason of the extensive alterations his poem underwent in the powess of incorperaiien., while for the Vengement F renders its chief service in offering additiona] evidence for the manu- Spe groupings otherwise indicated.
ized by GOOGIe
t
‘
ee Se ee ee eT
CHAPTER IIl
MANUSCRIPT RELATIONS IN THE VENGEMENT ALIXANDRE
In order to arrive at a satisfactory explanation of the filia- tion of the manuscripts in Gui de Cambrai’s poem it is first necessary to clear the ground by an elimination of those mul- tiple phenomena of concordance which may be regarded as depending upon an intermediary not material but psychologi- cal, that is to say upon similarity in the mental processes of the scribes. Then only is it practicable to consider con- cordances which cannot be explained in the above-mentioned way and which consequently point toa real filiation between manuscripts. Among these I shall call attention to cases of agreement occurring with sufficient frequency to constitute a definite indication of filiation. Next 1 shall take up such con- cordances as remain and atlempt an interpretation of them. Finally I shall endeavour to arrange and systematize the sum of my results by means of that conventional makeshift, the
‘‘family tree.”’
1. — TEXTUAL CONCORDANCES DUE TO SIMILARITY IN THE MEN- TAL PROCESSES OF THE SCRIBES, AND CONSEQUENTLY WITHOUT WEIGHT FOR ESTABLISHING FILIATION OF THE MANUSCRIPTS
In the genre of the epic of antiquity we are not dealing with works whose slightest word must be respected, but with narratives supposedly historical yet where in the details the individual fancy of the author or redactor is allowed great scope — we might even say almost unlimited sway. As a case in point we may examine line 133, where the best sup- ported reading for the first hemistich is ¢int se tiere a fiance. Yet we have in D tint la terre sa suer, in F tint sa terre a seiir,
14 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
in J tint sa terre molt bien, in K tenoit sa terre em pais. In the second hemistich the first word is car in HF, qui in D, si in I, molt in GKL, eé in J, and the rest of the line is in H preus est et hardis, in DGJK fu preus et hardis, in F prous fa et hardis, and in IL fu fiers et hardis.
Above all, the rhyme-words offer a great number of diver- gencies, sometimes for a given word almost as many as there are manuscripts of the poem. Consider for example the fol- lowing phrase in line 793: H sor son elme d’or mier; D sor Velme qu'il ot cler; F sor Vescu de quartier; IG desor lV’elme vergier; J desor V'elme el vergier; K en V'escut de quartier; L en Vescu de. pinnier. Here, in a phrase for which verbal parallels could be found by the score, we have in eight manuscripts five variants of rhyme-word and six of mean- Ing, any one of which offers an acceptable sense. In fact, when by reason of their great similarity to line 785 we set aside the readings of FKL, there is nothing in the remaining four, considered separately, to permit us to decide which was nearer the author’s original intention. The same abundance of variants can be seen in lines 476 (four variants); 556 (four variants in seven manuscripts); 779 (four variants); 885 (five variants); 965 (four variants); 1151 (six variants in seven manuscripts); 1179 (five variants in seven manuscripts).
It is evident then that the scribes of the individual manu- scripts are prone to change their copy. What are the causes productive of such change? Among them we may include the following: the manuscript from which the scribe copied may have been from time to time illegible; a word may have been unfamiliar to the copyist; an earlier line may have produced confusion in expression; the scribe may have substituted for one word another of similar appearance or of similar sound; the scribe may have desired to effect an improvement on the text ’ or to leave there his personal touch. One of the most important factors is doubtless the force of the cliché: bits of epic mate- rial— lines, expressions, words— are afloat in the memory of the scribe, and only a slight conscient effort, or even a meré
MANUSCRIPT RELATIONS 15
lapse of attention, suffices for some of this to be introduced into his copy. Take for instance the completely indifferent. and interchangeable variants in a phrase which occurs seven times in our poem.
753a: I il sist mout bien armés; G il fu tres bien armés; L et fu molt bien armés; J il sist tres bien armés; K et sist tres bien armez.
781: HDIGJK et sisé mout bien armés; FL et fu molt bien armez. >
801: HIGK et sist tres bien armés; Dil sist molt bien armez; F il fu molt bien armez; L cil fu molt bien armés; J et sist molt bien armés. |
902: HDFG bien fu armés Leones; ILJK bien sist armés Leones. | |
. 909: HFG il fu tres bien armés; D omits; IK il sisé tres bien armés; JL il sist molt bien armés..
4175: HG il fu tres bien armés; DILK il sist tres bien armez; F omits; J il sist molt bien armés.
1194: H et fu mout bien montés,; D et sist molt bien armés;
F omits; IK il sist tres bien armés; G il fu tres bien armés; L et seoit bien armés; J il sist molt bien armés. Here we have almost all the possible changes rung on this expression, but the sole and none too certain inference to be drawn from these variants relates to the habits of the individ- ual scribes. Thus we might say that the manuscripts IJ and K show a leaning toward the verb seoir rather than esfre in speaking of a mounted knight, but to trace this leaning to a common ancestor rather than to individual fancy or preference would be absurd.
We may safely conclude that many cases of agreement among manuscripts are explicable as due to coincidence, and so cannot be adduced tooverthrow a manuscript scheme. Nor can they. be adduced, except as supplementary or probable - evidence, for the establishment of a manuscript scheme.
After having sought a general explanation of fortuitous agreement, it remains to study the categories of words and
16 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
ideas which lend themselves to such agreement. For the most part we shall see that the scribes exercise their choice only in certain well-defined general groups of which the greater num- bér belong to the stock phraseology of any martial writing. Under the headings that follow I attempt to list the types of word or of form from whose accord in the passage cited no conclusions can be drawn as to manuscript filiation.
1) Nouns applied to persons: Dameldeus—li cors Dieu: 58: HFIJK—DGL pogneor—contor: 120: HDGL—IJK (F omits.\! ami—nori: 284: HFG—ILJK (D omits.) roi—duc—per: 284: HJK—FIL—G (D omits.) jouvenciel—damoisel: 334: HDFG—ILJK prince—home: 399: HDIJ—GK (F omits; L has individual
reading.) ami—baron: 1025: HDFGIK—LJ Grijois—Greu: 1039: HF—DGJ (I and L each have individual
readings; K omits.) gent—parent: 1054: HDFG—ILJK segnor—baron: 1192: HDK—IGLJ? vasal—varlet: 1345: HDLK—IGJ — frere—signor: 1383: HDILK—GJ
2) Common nouns applied to place of residence or of assem- bly: cité—chastel: 31: HFLJ—DIGK
192; HF—DIGLJK tour (cour)—cité—terre: 189: HDK—I—GLJ (F omits.) freté—crté: 314: HDIJ—GK (F omits; L: castel.) 342: HIGJK—DL (F omits.)
castel—moncel: 341: HK—DIGLJ (F omits.) forterecce—fermeté: 362: HDFGL—IJK chité—regné: 662: HDFGLK—IJ
3) Common nouns applied to instruments of warfare:
4. Note that in 114 poigneor stands in all the manuscripts. 2, Whenever F is not mentioned it is to be understood that the whole stanza is wanting in the manuscript.
MANUSCRIPT RELATIONS 47
pignon—penon: 694: HILJ—DGK 1090: HIJ—DGLK escu—fust: 884: HFL—DIGJK lance—ensegne: 885: HG—ILJK (DF: fust. ) 1610: HIGL—DJ (K omits.) hanste—lance: 915: HFI—DGLJK 1475: IGLK—HDJ | 1663: HIL—DJG (K omits.) siele—arcgon: 1624: HD—JIGL (K omits.) | 4) Nouns applied to material employed in manufacture: Sire—soie: 553: HDIGL—FJK Baudas—Damas: 589: HDIGJK—L 1109: DG—HILJK 4414: GL—HD (IJK: compas. ) cristal—esmal: 860: HDILJ—FGK coral—cristal—esmal: 1076: HIJK—DG —L 5) Nouns markedly similar in meaning or in form: siecle—monde: 192: HDFIGJ—LK 360: HIGL—DFJ (K omits.) ' trinité—maisté: 323: HGK—IJ (D and F, and L, have respect- ively individual readings. ) commant—talant: 405: HL—DIGJK (F omits.) nombre—monstre—ombre: 537: H—DGL—IJK pré—tré: 563: HI~DGLJK | olive—olivier: 735: HDFGL—IJK sart—essart—regart: 847: HIJK—DGL—F iror—dolor: 1126: HIJ—DGL (K omits.) casement—tenemant—mandement: 1340: HIGJ—DK—L orage—oré: 1406: HJK—DIGL nient—rien: 1593: HIL—DJG 6) Nouns and adjectives applied to steeds: coreor—musaudor: 119: HD—FIGJK (L: ambleor.) 4449: H—DIGLJK gascon—arragon: 366: HDFGJK—IL 1016: HGLJK—DFI Baucant —Blancart: 781: HDFG—ILJK 801: JK—L (HDFIG: ceval.)
18 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
ceval—destrier: 1142: HIGL—DJK 1460: HDJ—IGL (K omits.)
fauve—Fauvel—un mul: 4175: HIJK—DG—L
1) Adjectives applied to persons and introduced mainly as emotional epithets: large—preu: 104: DFJ—GK (H and I have individual readings;
L omits.) preu—fier: 133: HDFGJK—IL vallant—poissant: 399: HDG—ILJK (F -omits.)
412: IGLJK—HDF prové—mortel: 4314: HDFGLK—IJ mellé—barbé: 559: HILJ—GK (D has individual reading.) mort abatu—jus abatu: 814: HIL—GJK (DF omit.) hardi—eslit: 882: HDIGLJK—F 885: HK—ILJ (D, F, and K each have individual readings.) 900: D—J (HFIGLK: gentil.)
vallant—persant: 975: HDFJ—IGLK noble—hardi: 1158: HILJ—DG (K has individual ear hardi—plevi: 1171: HG—DILJK
8) Adjectives applied to instruments of warfare: bon—cler—vert: 404: H—IL—GJK (D has individual reading;
F omits.) tout nu—molu: 807: HDFIGK—LJ © frois—bloi: 914: HDFG—ILJK orin—doré—burni: 1007: HI—GLK—J ak have individual
reading.) doubleor—de color—paint a flour: 1444: HIJ—-DG—LK
9) Adjectives applied variously as decorative epithets: bon—fort: 86: HIJ—DFGLK
666: HL—DIGJK
-tierré—fermé: 423: HDGLJK—FI gent — grant : 497; HILJ—FK (D omits; G has individual
reading.) grant—ruite—estrange: 805: HGJK—DIL—F fier—grief—cruel: 1064: HDK—IGJ—L
‘MANUSCRIPT RELATIONS 19
grand—grief—fier: 1077; HDLK—IJ—G haut—cler: 1328: HGIJ—DILK rice—bel—grant: 1347: HJ—DILK—G
10) Verbs markedly ‘similar in meaning: noumer—claimer: 87: HFGIK—LJ (D: apeller.)
1052: HI—GLJK (D omits; F has individual reading.) dire—faire: 163: HDIGLJ—FK plorer—larmoier: 381: HDFJ—IGK (L has individual reading.) respondre—dire: 395: HIGL—DFLK sigler—errer: 495: HIGK—DL (F and J omit.) enpuisouner—envenimer: 523: HDI—GLJK (F omits.) enficier—refichier—afichier: 688: HIJK—-DL—G trebucier—craventer—abatre: 897: HDFL—IK—GL 1447: D—HIGJ—LK
jouster—rejoster: 945: HGK—DFIJ (L has individual reading.) coisir—veir: 980: HDIGLJ—FK desmentir—dessartir: 1001: HFILJ—DGK ' cesser—finer—arester: 1095: HDI—GJK—L croire—cuider: 1296: HGLJ—DIK soi partir—partir—soi departir: 1396: H—DK—IGJ (L omits.) encaucier—chacier: 1555: HDIG—LJ 's’en partir—soi departir—departir—soi partir: 1566: HG—DJ
11) Verbs markedly similar in form: esmet—esmai: 473: DFGL—IJK (H has individual reading.) s’entrevinrent—s entrevirent: 783; HGL—DIK (F and J have
individual readings.) a josté—a esté—ai esté: 1271: HDGJK—I—L' recreti—retenu: 1459: HK—DIGLJ |
1463: J—DIGK (H: detenu; L omits.)
recovra—retorna: 1476: HIGL—DK (J has individual reading.)
4. Since i and j are not differentiated in the manuscripts, and since e and o are frequently indistinguishable, it is easy to see how the copyist of I, to avoid an obvious absurdity, should have corrected what he took for a first person into a third person.
20 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
12) Verbs markedly similar in application to a given con- cept: monter—errer: 222: HGIK—DFLJ prendre—trover: 369: HDIL—FGJK acesmer—aprester—achiever: 446: HI—DLJK—G (F. omits.) desarmer—desancrer: 547: HK—DIGLJ (F omits.) traire—tenir: 807: HG—DFILJK
4201: IJ—HGLK (D: prendre.)
troer—percier: 828: HIJK—DFG (L omits.) hanir—foir: 996: HILJ—FGK (D omits.) rendre—raiembre: 1038: HGL—IJK (D has individual reading.) doner—dire: 1224: HIGJ—DL (K omits.) contenir—maintenir—bien tenir: 1464: HIG—DJK—L
13) Alternative tenses and persons of the verb:
a) Present—past definite:
a—ot passim. est—fu (iert) passim. tiennent—tinrent: 109: DL—HFIGJK oent—oirent: 210: HDFIK—GLJ se reloge—se loja: 686: HIJK—DL (G: s'est logiés.) apelle—apela: 1168: DILJK—HG
b) Present—future: avons—arons: 379: DFIGLK—HJ
c) lmperfect—past definite: _tenoit—tint: 70: HDJ—FGLK (1 has individual reading.)
d) Past definite—past indefinite: se loja—s’est logiés: 686: DL—G (HIJK: se reloge. )
e) Past definite—future: | prisierent—priseront: 143: HDJ—F IGLK
f) 3rd singular—3rd plural: commande (commanda, commencha) — coumandent : 770 :
D (JK,L)—HFIG
g) ist plural—2nd plural: pensons—pensés; 504: HG—ILJK (DF omit.) porons—porés (pensez): 1314: HDG—IJL (K)
h) 1st plural—3rd plural: puissons—puisent; 214: LJI—HDFIGK
MANUSCRIPT RELATIONS 21 14) Adverbs of:
a} Manner: biel—bien: 72: HL—DFIGJK $19: FGK—HDILJ 7178: DK—HFIGLJ a envis—et anuis: 99: HFIGL—DJK si—tant: 383: HDFI—GLK (J: maint.) bien—tost: 503: IJ—GLK (x has different reading.) plus isnel que le pas—assés plus que le pas: 574: HDIGK—LJ molt—tres: see supra, p. 15 bonement—bellement: 1035: HFIGLK—DJ auques—assez: 1044: HFIGK—DL/(J has individual reading.) 1410: HIGJK—DL granment—forment—durment: 1099: HI—DGLK—J a force—ensamble: 1107: HDIGK—LJ 1257: HDIGLK—L plus—molt: 1139: HIGK—LJ (D: trop.) 42144: GLJIK—HI (D has individual pening: ) bien—mius: 1296: DGL—IJK (H has individual reading.) a poi—por (per) poi: 1426: HDIGK—LJ par force—a force: 1680: HG—DJIL b) Time: donc—lors: 215: HIGL—DFJ K 541: HJIK—IGL (DF omit.) ja mais—il ja: 293: HFIGJK—DL puis—il: 848: FIGLJ—HD (K omits.) sempres—anqui: 1449: HDGLK—IJ c) Place: , contre vent—contre val—contre mont: 769: HGJ—DI—LK (F omits.) | 15) Prepositions: desor (desore)—desus—dalez: 86: HIJ—-G—K (DF, and L, have individual readings.) sus—ens—sos—sor—en: 88: HF—D—IK—GL—J tenir a—tenir por: 445: HIGK—LJ (DF omit.) desor—sor: 572: HDIK—GLJ
92 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
amont sor—desor son—sor son: 808: HDIJ—GLK—F avoec—a tout: 873: HILJ—DG (F has individual reading; K omits.) de delés—par delez—par dejoste: 1029: HD—FIGLJ—K 16) Pronouns and articles: li—l’en: 46: HIGLJ—DFK cy—cist—cil: 56: D—FIGJK—L (H has individual reading.) s’en—en: 83: HDFIGJ—LK cel—cest: 128: HDFGJ—IK (L omits.) vous—nous: 167: HFGL—DIJK li—lor: 349: HGJ—ILK (DF have individual reading.) nus—on: 369: HDILJK—FG.- nostre—nului—le mien: 569: HDL—IK—GJ (F omits.) le—son: 967: HDIL—FGJK le refiert (le fiert)—refiert lui: 1138: HI (L)—DGJ (K omits.) lor—les: 1164: HIL—DGJK ne les—nes—nous: 1220: HG—DIJK—L le—au: 1263: HIGJ—DLK 17) Conjunctives, asseveratives, negatives: et—ce: 32: HFI—DJK (GL have individual reading.)
197: DI—HGL (F and JK have individual readings.) s’en—si—et: 121: H—DFIJ—GK (L has individual reading.) et—qui: 174: HD—FIGLJK quant—que: 177: HD—FIGK (LJ: de.)-
178: HDFJ—IGL (K omits.) en—si; 210: HDFGJ—ILK por—et: 231: HDIK—GLJ (F omits.) n’en—ne: 373: HL—DFIGJK que—com: 379: HFL—DIGJL or—ce: 1229: HIGK—DLJ mie—pas: 1717: HG—DJI (L omits.) 18) Expressions of time and number: ensanle—andoi: 305: HFIGK—LJ (D has individual pending: ) a quinsaine passant—a quinze (D: au .xv.) jors passant— ains le tiers jor passant—ains quinsaine passant: 421: HI— DG—L—JK (F has individual reading.)
‘MANUSCRIPT RELATIONS 1 93
a l’entree d’avril—ce fu entrant avril—ce fu el mois d’avril— ce fu el tans d’avril—ce fu el mois de may: 754: H—DFJ—
_I—GK—L ,
vint mil—set mil: 759: HD—FGILJK
19) Inversions of order: \ tant li rois—li rois tant: 102: HDIGLJ—FK sebelins mantiaus—mantiaux cebelins: 365: HIGLJ—DK (F
omits.) | onques par eus bastis—por aux bastiz onques—onques bastis
par eus—onques fu bastis entr’aus: 479: HDG—F—IJK—L biaus jors, clere vespree—clers jors, bele vespree: 704: HDK
—IGLJ
20) Omissions of single lines unnecessary for the general sense of the passage:
33: FG omit. 142: DFI omit. 445: HF omit.
846-47: HK omit.
21) Certain miscellaneous phrases which fall into none of the above categories, but in which coincidence can obviously - be attributed to chance:
384: HIGJ On pora tres bien dire.
DFLK Dont (LK: or) poura on (F: l’on) bien dire.
392: HILK De quel tiere estes né.
DFGJ De quel terre estes vous: 496: HIJ | Grans fu li asanlee. DFGK Mont fu grans l’asamblee. L -_ Ensi fu li aiinee. 338: HILJK Sil (L: qui; J: se) ensi (I: issi) sont esclos (ILJK: estors). DG Qui si lor sont estors. 820: HDG Nus hom nes escriroit. (F omits.) ILJK Nus hon n’escriroit tant (JK: tout). 1183: HL ~~ A terre l’abati. (1G omit.) DJK Que il (J: tout) mort l’abati. The foregoing list is assuredly not exhaustive, and from
24 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
other texts other categories might be deduced. Again, all the words under discussion might be classed, as has been done in the case of the verbs, according to similarity in meaning, form, or application. It is for the sake of clarity that I have introduced certain more purely formal subdivisions. In all cases a comparison with the context will clearly show the idea involved, and explain why such and such a concordance can have no weight. To sum up, I would repeat that wher- ever there is possibility of independant accord, there we must avoid any argument as to the filiation of the manuscripts ; and in determining accidental coincidence common sense is the surest guide.
2. — TEXTUAL CONCORDANCES POINTING TO FILIATION OF THE MANUSCRIPTS °
a) Concordances occurring with sufficient frequency to con- stitute a definite indication of filiation.
The eight manuscripts in question fall clearly into two groups, for which evidence occurs in such quantity that there is no stanza which does not contain corroborative examples.' These two groups we may call x and x’ ; x comprising the manuscripts HDF and, from the latter part of Stanza 64 on, J;? x’ comprising the manuscripts IGLJ and K.
In the following cases x’ is clearly at fault:
1) 246a, x’: Aristotes li maistres qui mout fist (JK: fait) a loer. —
4. Cases in which it is possible to declare with assurance two or more manuscripts in common error are infrequent. We usually have to do with readings which, however much they may differ one from the . other, offer, from considerations of sense, acceptable versions. In con- sequence of this infrequency of assured common error, when I have based my deductions upon cases in which one group of manuscripts is clearly at fault, I add supplementary evidence pointing towards the same grouping. | .
2. From now on, J, inthe absence of specific statement to the contra- ry, will be used to designate this manuscript in its relations with x’, and J, to designate it in its relations with x. Thus J will normally refer to J, lines 14-1463, and Je will refer to J, lines 1464-1749, |
MANUSCRIPT RELATIONS 25
By the insertion in x’ of this line, which does not occur in x, Aristote is made the protagonist of the next scene. From the agreement of all the manuscripts in 288-89, it is clear that it was Calnus who first saw the two knights. This is in accord with the earlier passage as treated by x. The insertion of this erroneous line ‘in x’ is due to a misreading in 215: cascuns for Calnus. |
2) 243, x:Qi le vigne planta (D:aporta) premerains el pais.
(F omits.)! x’: Et le vigne planta premeraine (JK: premerains) es (GLJK: el) pais. | |
In x this line refers to Denis (Dionysus), the reading of H.? This interpretation is in accordance with all we know of Greeco-Roman legend and its diffusion in France. On the other hand, x’ places the line after 246, where it ineptly refers to the Queen Semiramis, of whose viticultural exploits we have no record, while in Ovid andelsewhere she is designated as the constructor of the murs d’araine bis as in 246.
3) 592-651, x’ omits the passage.
The absence of this passage from F is a part of the omission by F of the whole of Stanzas 24-26, and so is of no interest in the present connection. We may be confident that this pas- sage belongs to the original, not merely because it contains a statement necessary to the plot,? but because of the treatment of the line immediately following the break. This line appears in ILJ and K as the last line of Stanza 24. Regardless of
whether the omission in x’ is the result of a deliberate conden- —
sation or of a gap in the source of x’, it is evident, bothfroma consideration of sense and of rhyme, that in ILJK the last line of Stanza 21 is misplaced. In fact G, recognizing that something is wrong, omits it altogether and inserts three lines found in
4. By reason of the condensation in F of the material drawn from the Vengement we must expect to find many cases where the readings _ of x will be preserved only in D and H.
2. The reading Denis in H is supported and confirmed by the reading deuis in x’.
3. Compare 630ff, and 1033ff.
26 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
no other manuscript. IL reproduce the omission as it exists. JK insert before the last line of the stanza one line to connect it with 591. The conclusion then must be that the common error of omission points to an x’ grouping.
The poem does not offer equally conclusive examples of error in x, but in the following cases the evidence is strong:
1) 157-59, x: Aristés et Clincons (DF: dans Clins), li sires
d’Orcanie
DF insert: Vinrent en celle cort (F: I furent tuit venu) a molt grant compaignie
Et Tolomés i vint a la ciere hardie (D: et l’autre baronnie), (F omits.)
Et tout li autre per o mout (F omits mout) grant baronnie (F: chevalerie). (D omits.)
x/; Et (K: Dans) Tholomés i vint, a le ciere
hardie,
Clicons et Aristotes (G: Aristés; L: Aridés) li sires (IJ: le signor) d’Orquanie
Et tout li autre per (J omits per; I: i vie- nent) a (GL: par) mout grant baronie (GL: segnorie).
Although it is here only a question of the inversion of lines 157 and 158 in x, the strain thus put on vint by making it ‘support plural subjects in both preceding and following lines is too great for us to suppose that the lines originally stood in this sequence. Nor can we argue that the extra line which D and F offer as solution is original, and that we have to deal with a case of common omission in H and x’. For one thing, variants of the formula a molt grant compagnie are already found in this stanza in lines 159, 161 and 174, and the inclu- sion of yet another would be to accuse the author of an almost total lack of the literary sense which we perceive elsewhere in the work. Also, cases of repetition of the same rhyme-word ina stanza aré, while not wanting, rare, and here conpagnie 1s demanded in line 173 by the agreement of all the manuscripts. We are thus forced to the conclusion that HD and F concur
MANUSCRIPT RELATIONS 27
in giving an order of lines not in the original, and although agreement in inversion might possibly result independently it seems more logical to trace 1t to a common ancestor. 2) 343-47, x: Et mout sont ricement batillié li cretiel (F: crenel; D omits the line.)
D’escus as (D: de) cevaliers adoubés de nouviel. (F omits this and the remaining lines.)!
Les lances sont brandies, li estage sont biel (D: De lances brandeices sont li estache fiel),
Et les breteskes plaines (D: belles), et (D: souz) levé li rastiel.
x’: Et mout sont ricement batillié li cretel (K. Mout i sont hautement entaillet li cretel),
Desus (L: Desor) a cevaliers adoubés (L: arobés) de novel,
Les lances ont (K: sont) drecies, si ont (K: a) maint pignoncel;
Les bretesces sont hautes, li estage sont (K: mout) bel (L: et li estage bel),
Les portes coleices u pendent li rastel (J: fastel; L: rainsel).
It is easy to see how in the second line of this passage one scribe should have altered desus a cevaliers to d'escus as cevaliers? and subsequently modified — and botched — what followed in an attempt to reconcile it with his false start. Compare the reading ofx’. Here nothing is wanting for a clear and logical presentation of the scene. It would be difficult to comprehend how such unity could have resulted from arework-
4. Since F will be shown later (pp. 30ff.) to belong with D, the existence of an x group will be sufficiently confirmed, in cases where F is lacking, by H and D.
2. This is all the more comprehensible since the hanging out of shields as cognizances is not unknown to mediaeval poets. Cf. (ruil-
_laume de Dole [SATF), 2125ff.
28 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
ing of an original as in x. Therefore we are again led to assume common error in the latter group.
The classification of the manuscripts into these two groups is confirmed by the following specimens of supplementary evidence, which give an idea of .the frequency with which the separation between x and x’ repeats itself throughout the poem: |
84, .x: un fort castiel basti.
x’: un castel establi.
87, x: priés estoit d’Alixandre.
x’: el destroit Alixandre (K: d’Alixandre).
x: Dauri (DF: Davi).
x!: anti (J: senti).
x: ne perdi (F: ne vi).
x’: ne servi. (L omits.)
xX: s'apercoit.
x’: percoit (J: paroit) il.
x: tres bien avés parlé. x’; ne me vient pas a gré (L: ce ne m’est mie agré),
731, x: ciertes (DF: sire) molt volentiers. x x Xx X x X
88,
's certes, dist Aristotes. : a l’afaire (F: le respons) conté. ’; a tout issi (LJK: ensi) conté (G: a trestout ce conté). x: n’1 quisent (D: quistrent) soujour.! : n’i fisent (LK: firent) sejour (G: n’en feront sejor). : Perideis s’esploite (D: esploite) et li autre pio (D: Grijoir). x’: Lors ont bien (GLJK: tant) esploitié li nobile con- tour (L: li noble poignedt): 1448, x: conoisent. x/: coisirent. 1450, x: Por bienvengier (D: por conjoir) le frere au noble pogneour (D: au roy Macedonor). x’: Desireus (G: devisant; LJK: desirrant) de (G: a) vengier le tort (GLJK: mort) de (L: a) son signor.
4. In the case of Stanza 63 (= 1437-50) which comes after jthe point where F breaks off, x = HD. 3 |
MANUSCRIPT RELATIONS 29
4513, x: Caunus de Menalite (DJ: et Caunus de Melite).! _ x’: Aristotes hi maistres. 1520, x: et Divinuspater apiela (D: apele). x’: et Marindes apele son home. | 4521, x: amis alons descendre (DJ: alons nos en): x’: Sinadon or enten. 1526, x: nenil gou dist li siers (D: nenil fait dont li sers).
U
x’: nenil ce dist Marindes. (K omits.)
Differences of this sort where neither set of manuscripts is definitely in error are not in themselves of sufficient weight to constitute the basis of a classification. The resemblances in either set may be due merely to the fact that all of its manuscripts have preserved the correct reading; one of the sets does constitute a group, but we have no way of deciding which one. However, once we have established cases of com- mon error for both groups or even for only one of the two, numerous examples following the same lines of division con- stitute, by their mass, corroborative evidence.
It is unnecessary to give additional evidence of the same character, examples of which could be drawn from every stan- za of the poem. It is however important to know whether there is any testimony opposed toa grouping x and x’. I have found only the following example:
842, HIGL: lia le cief fendu.
DFJK: l’a trestout porfandu. Here the distribution of the manuscripts points away from the two groups we have seen recurring above, without giving us the possibility of saying that either reading is erroneous. The only explanation I can suggest is that, in the second hemistich of the line which follows the passage cited, all the manuscripts read: li a molt cier vendu; and that the force of the verbal resemblances between the two lines, together with the occurence of cief two lines before, may have led to an identical alteration of line 812 by two copyists having no
4. For Stanza 66 (= 1502-26), x= HDJ,; x’ = IGLK.
30 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
connection one with the other, Since this is the sole example and since it is at least possible to explain it by coincidence, I do not consider that it should be accorded any weight in comparison with the manifold testimony in support of a dif- ferent grouping.
My conclusion is that in the present work the quantity and quality of the testimony establishes the existence of the group x’. Likewise the probable error of x in the examples cited above (page 26), and the constant opposition of x to x’ throughout the poem, give substantial support to the view that it too forms a group.
The above division recognised, it remains to be seen whether within either group we may observe any subdivisions. A circumstance which we cannot fail to observe, and which hardly yields to the main division in its clear claim to atten- tion, is that D and F often agree against the reading of the other manuscripts. Our preceding classification would lead us to expect DF to bein error when they present a certain reading and H and x’ another. In the following cases the evident
inferiority of the DF version coincides with this expectation:
1) 86, DF: que je vous chant (F: cont) ici.
H: de sor ot un fai.
x’; par desore (G: desous; J: deseur; K: dalez) un
(G:y) fai. (L gives, as often, an individual
reading: et mout-bel fu basti.) 2) 444, D: si com il m’est avis. | F; ensi cum est avis.
H + x’: en (L: i; K: la) fu (G: ont) li jors assis. (I omits.)
In the above examples D and F substitute for the correct
readings, as evidenced by H and x’, obvious rhyme-fillers.
3) 409-11, DF: Des sers dont vous parlez, et nous dites
commant. Sire, molt volentiers vous dirons nous commant (F: lesemblant), = H -+- x’: Qi des sers de put aire nos (IJK: vos)
MANUSCRIPT RELATIONS 31
diroit (ILJK: dist ne; G: a dit) tant ne quant,
Et de le mort le roi dont nos sommes dolant. (K omits.)
Et (K: ce) respondi-li uns (J: li uns respondi): Ja sarés le (IJK: ja en serés; G: ja en serons; L: bien me solés) creant. |
From self-evident reasons of style, and from the absence of any transition between the two speeches, it is evident that the version offered by D and F is not so good as that found in the rest of the manuscripts. |
There are also numerous cases! where D and F are the sole manuscripts to give a line; in none of these cases is the line indispensable, and this concordance is too frequent to be the result of fortuitous independant omission in H and in x’.
I have mentioned that, in the course of Stanza 64, J goes over to x, and that thereafter J always gives the x reading when it isa question between x and x’. The first intimations we have of this change are after 1464, where x’ (that is, IGL and K) inserts an additional line, and in lines 1481-82 which HD and J are alone in giving. In line 1483, the substitution with D of gloton for traitour in H and x’ gives us the indica- tion that it is toward the DF group that J is tending, and in 1512 and again in 1555 the substitution of rois Alior for Anti- gonus (x’: Antiocus) in one case and for Liones in the other is proof that the source of J, was one of the manuscripts amalgamated like F with the Venjance.? From this point on to the end of the poem there is nothing which contradicts the group xJ,. One example will suffice to show how definite is the passage of J to D(F).
1587-92, DJ: Certes, dist li vaxaus, ci a (J: c’est) povre reansons |
1. 10a; 158a; 524a; 748a; 787a; 842a; 864a; 904a; 932a. 2. See supra, Chapter u.
32 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
Contre la mort le roy qui tant par estoit bons. Com (J: car) je ses freres sui, si con- ceut (J: consaus) ces (J: les) felons Qui por sa mort bastirent si perileux (J: orible) poisons, Dont sui je recreans, quant ges ai en ‘prisons, Se par moy en est faiz ne rachaz ne pardons. ! H + x’: Certes, dist li vasaus, c’est povre raencons Et por (x’: Contre) la mort au roi (IL: le roi; G: mon frere) qi tant estoit preu- dons (x': qui tant par estoit bons). De moi meismes voir (x’: seroit ce deraisons | Quant jou ses freres sui et tu sers a ban- don (x’: ses hom ses abandons)— Encor soit auques povre ceste conpa- risons— — Que (x’: Ja) par moi n’en fu (x’: ert) fais ne racas ne pardons. In a consideration of x’, the first striking phenomenon is that L often deserts x’ to take readings from x and particu- larly from F. This brings up the problem of contamination.
meismement)
Contamination consists in a normal agreement of a manuscript with one redaction, upon the readings of which are occasion- ally grafted others from a different redaction. It differs from amalgamation, whether as found in F, where two separate sto- ries are combined into one by a form of dovetailing, with a_ section of one poem succeeding a section of the other, or as in J, where the last portion of the Vengement is copied from a manuscript representing a branch different from that on which the first part is based. An amalgamated version presents the
4. Here x = HDJ,; x’ = IGL (K is preserved only through 1524).
MANUSCRIPT RELATIONS 33
errors of the group which the scribe is at that time following; a contaminated version generally avoids obvious errors, or at» least those occurring in the secondary manuscript, and takes ‘therefrom only readings which the scribe considers better than those of the main manuscript or indifferently interchangeable ‘with them. The motives and the procedure connected with contamination are difficult to determine. It is possible that the copyist, in the desire to give a new colouring to his redac- tion, perhaps even with his supplementary manuscript on the other side of his revolving lectern, may have occasionally glanced thither to compare or to rectify. The considerations which led him to prefer one manuscript to the other as the base of his edition or to take precisely one word or expression instead of another are usually impossible to determine. In any case, from what we know of the production and copying of manuscripts, when a scribe has deviated from his basic manuscript in favour of a supplementary one, it is improbable that he will erase the reading he has recorded and substitute another.
As has been said, in a manuscript which has undergone contamination we shall not find many examples of agreement in error with the secondary manuscript, and since the proba- bilities are against our being able to establish such agreement | in error the cases of agreement will have little individual weight for establishing contamination— indeed I have attempt- ed to demonstrate that a great number of these agreements could have arisen by chance. However, if examples of the same grouping occur in sufficient number to constitute a trend, we are justified in assuming that they can not all be due to chance. In the specific case of. contamination under discussion, L is nearest to F when F contains the passage; to D when F does not.!
In the following cases, in which F is present, L departs from the reading of the x’ group and resembles F:
4, Where both D and Fare lacking, L still shows by similarity to H
its contamination from the x group. 3
34 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
98,
122,
138,
FL: : que petit l’ont joi.
: petit en ont joi. (G omits.) : retour.
: ator (G: tor).
: il enragera vis.
: il en esragera vis.
qe petit l’ont servi.
: mout en sera maris.
: le mestre sale.
: la grande sale.
> conraé.
: atorné. (G has individual reading.)
: del (F: de) parlement juré (F: doné). : del sairement juré.
: li fait el cerviel frire.
: lia fait el (J: le) cief frire.
: les langues (L: pendoient) contre val. : les langes de cendal (D: cristal).
; et mis en vostre terre. |
: et mis dedanz ta terre.
: et mis ens en sa ta tiere.
; et as mis en ta tere. (K omits.)
In the following cases, in which F is not present, L departs from the reading of the x’ group and resembles D or H.
309,
HDL: x:
DL: H-+- x’: HL:
monté. entré. |
cité.
freté. |
et mainte rice coupe. (D omits.)
: et maint (G: mainte) rice doner (G: dame;
JK: donné).
> errer.
: sigler. (J omits.)
: maint divers tres i tendent.
: tant divers tres tendirent.
: Mais (x’: tant) divers tres i a (x’: ot). : le chief dou bu tolu.
MANUSCRIPT RELATIONS 35
H + x’: le cief sevré del bu. (K omits.) 1078, HL: es escus coumunaus. (D omits.) x’: ens (JK: es) escus a (J: as) esmaus. 4218, DL: li traitor se tiennent. H + x’: cist (x’: li) traitour s’enbatent. 1263, HD: sil envoions. L: cil lenvoient. x’: sel trametons. 1345, DL: di, quiest cilz (L: cil) vasaux? | H +x’: amis, qui'st (J: dist) cis (J: cil) vasaus (IGJ: varlés)? 1350, D: qui soit de tieus parans. HL: nede si grans parens. x’: ne de si grant despens. 1403, HD: qu'il desirent forment. L: dont il sont desirrant. x’: dont cil (J: il) seront dolent. (K omits.) 1405, D: sanz nul restorement. HL: sans nul delaiement. x’; sans nul arestement.
1505, HDJL: meesme li poilent.
x’: li plument meisme(GK: meisme li plument).
I have omitted certain minute points of agreement in which it is unnecessary to suppose contamination, and reserve for later treatment certain cases which might constitute further corroboration but which require detailed investigation. Even leaving these aside, the evidence indicates that L or the origin- al of L took occasional readings from an ancestor of F after the F branch of DF had detached itself from the common source of DF’, but before all the individual variants of F had appeared.
One other fact stands out from an examination of the manu- scripts: a relation of some sort between K and L. Let us exam- ine the evidence before attempting to draw conclusions as to the exact nature of this relation.!
4, Here for considerations af brevity x’ is used for the three manu- 3*
36 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
75, x + x’: richement.
103,
134, 195,
68),
LK: DFLK: H:
: et par verté le di. X-++x': LK: X-+ x’: LK:
X + x’: LK:
f
largement. | pour (FLK: par) verté le vous di. par verité le di.
devint mout (D: touz; F: puis) ses amis.
fu dolans et maris.
hi traitour felon.
li encrieme felon.
l’autres (D: li autre) de siglaton. (F omits.)
li autre (K: li autres) de coton (K: d’auqueton).
753a (in x’ only),
173,
7193,
896,
932,
IJ: : et sist desor ferrant.
: sor un destrier ferrant.
; cil del castiel esploitent (F: chevauchent;
desor un auferrant.
I: s’esploitent; J: s’acesment).
: cil du chastel en (K: s’en) issent. : desor l’elme vergier.
: desor l’elme el vergier.
: sor son elme d’or mier. |
: sor l’elme qu'il ot cler.
: sor l’escu de quartier.
: en l’escut de quartier.
: en l’escu de pinnier.
: tot droit.
: devant.
; maint riche (F: vassauz) poigneors. (H omits.) ; mout de lor poigneors.
LK:
mains (K: mout) de lor compaignons.
963, x-+x!: a loi (F: fuer) d’oume sacant. L: com chevaliers vaillant. K: com chevaliers sachans.
scripts (IGJ) of the group which remain when LK are considered sepa-
rately.
iteea ty Google
MANUSCRIPT RELATIONS _. 37
4482, H: un de ses drus coissi. |
D: un des siens a chosi. (F omits.)
x’: un des autres coisi.
L: un des Grieus si feri.
K: un des siens si feri.
1326, x-+-x’: or saciés (I: soiés) a (D: de) fiance. (F omits.) LK: s'il ot (K: a) paor de mort.
1477, x+x’: que il li (J: ili) mosterra. (F omits.) LK: qui (K: qu’'1l) li mousterra Ja.
From these examples we see that K is connected with L, to such an extent that in 932 they agree in giving an incorrect rhyme. Also 103 and 793 indicate that the relations of K with L date from a stage in L posterior to its contamination with F. Why can we not, then, assign L and K to a subordinate group in x’? First, because we have as yet no light on the position of L itself in the x’ group, since all we know is that it shows signs of contamination from the other group. Second, because K, in order to be placed in a group with L, would normally include all the examples of the Lx contamination, since it contains some of them. This is not the case. Therefore we are forced to postpone the final placing of K until we have exhausted the other evidence which the manuscripts may contain.
b) Residual concordances and their interpretation.
Up to this point the investigation has led to the following conclusions: 1) The manuscripts fall distinctly into two groups: DFHJ, and IGJKL. 2) D and F form a sub-group. 3) L tre- quently leaves x’ to attach itself to x and more especially to F. 4) K has many readings in common with L.
It remains to be seen if from the manuscript agreements it is possible to determine any other manuscript relations. Since many cases of agreement can be attributed to chance, and since it would be unwise to attempt to establish any group- ings solely from such cases, all argument for further group-
38 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
ings must rest on examples of accordance which coincidence would not normally explain. The following list will therefore contain all such examples which remain, with the exception of the groupings x-x’, DFJ,, FL, and LK, since these are already sufficiently substantiated.
1) 4134, H:
DF:
G:
IJ:
: mout en sont en freor.
2) 2415, HDF: : d’umilité plorer.
:mout tenrement plorer.
3) 230, HD:
4) 264,
6) 418,
povre en ont le pior.
li povre ont le piour.
il en ont le pior. (L omits.) forment sont en-esror.
de Milette plorer.
Et les piés es estriers par devant tra- verser. (FILJK omit.)
: Et lor piés et lor jambes molt sovent
regarder.
: mais or est il peris. (DFK omit.) : mais or est empiris. ; si va de pis en pis. : or va de mal en pis. 5) 407, H: D: nostre gent paisant. : de gent si (J: ci) paisant. : estre gent paisant. : vilain ne paisant. : et sa gent avenant. : et sa gent l’avenant. FJ: IGLK: 7) 437, HDFJK: IGL:
de no gent paisant.
et sa gent la vaillant.
et sa gent en avant.
car le nos creantés.
et car (L: cor) vos en hastez.
8) 450a, JK insert:
A [i
Li serf seront ochis, nostre en iert l'iri- tés. |
K: Se li serf sont ocis vostres ert li regnez.
MANUSCRIPT RELATIONS 39
9) 479a, x’: inserts a line. JK: insert the same line but as 478a. 10) 481-82, D: Et fermé forteresce et se sont mis dedans. : Mais je vous di por voir, dou garir est _ neans. (HF omit.) I; Le castel ont fermé, poi lor valra lor sens, . GL: Dongons ne forteresce, rouge (L: blanc) ors ne blans argens. Le (L: Se) castel ont fermé, poi lor vaura lor sens (L: poi i ara deffens). JK: Dongons ne fremetés ne lor vaura (K: fera) deffens, Nes porra racater rouge or ne blanc argens. 41) 594a, JK insert: J: Ou il a maint pingnon de paile et de cendals. K: Ou il a maint penon de paile de Damas. 12) 658, H: lor estor lor rendés. D: lor estor endurez. IJ: l’estor en endurés. . GL: tres bien vos deffendés. K: mout bien vos deffendez. 13) 668, | H: pruec que il fuscent coi. — D: por tant qu'il fussent coy. _IGL: par si qu'il fussent coi. JK: s'il ne fussent que troy. 14) 731a,. GLJK insert: i G: Irai as traitors volentiers et de gré. L: Et je irai tantost volentiers et de gré. JK: Volentiers i irai, par moi n’iert refussés (K: renfusé). 15) 889-94, IGL omit. 46) 920, HIJK: Mout sont grandes les os.
40
17) 970,
18) 1019,
19) 1094, HDJK:
IGL:
20) 1098,
21) 1102, HDIJK: GL:
22) 1135, I:
Kx:
: hi rois el missaudor.
: et Clicon son parent.
: et Caulus son parent.
; Jusges au mont de Cus.
; Jusqu’au bonus Artus.
: dusq’as bonnes Atus.
: deci as (J: au) puis agus.
: dusques as puis agus.
23) 1215,
24) 1233,
MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
DF: : Molt fu grans la bataille.
: el pré mort et (F: ou) sanglent. : mort ens el pré sanglant.
: mort el pré tout sanglant.
: mort ou pré verdoiant.
: par desour le sablon.
: par desus le sablon.
: desore le sablon.
: del destrier el sablon.
: del destrier arragon.
: dou bon destrier gascon.
: de son doré archon.
Mont par sont grans les os.
Or cevauce Marindes, mout fu grans ses
barnés.
Mout fu fiers (G: li) rois Marindes, et (GL: plains) de grans crueltés.
Or (G: il; L: et) cevalce a grand (L: par) force, mout fu grans (G: et il-et) ses barnés (L: plains de molt grans fier- tés).
: Clincon et Lincanor apiela Tholomés. : Clicons et Licanors parla et Tolomés. : Clicons et Tholomés par sont molt
acorés. et li regnes gastés. ains que vos en partés. el destrier missaldor. (HDL omit.) el destrier coureour.
MANUSCRIPT RELATIONS 41
25) 1254, H: mander par amisté. : mandent par amisté. : mandons par amisté. ; comment par amisté. : que on li ait mandé. : que il l’aient mandé. 26) 1344-15, IL reverse order. | 27) 1331, HD: nesun(D: nul dieu) plus droiturier. IG: nul plus droit iretier. LJ: nul meillor iretier. (K omits.) 28) 1530, HDJ: Parideiisles (DJ: le) voit. I: Voient Parideiin. G: Divinuspater voient. L: et Divinuspater. 29) 1654, H: amont el hiaume amont. | L: amont en I’elme amont. DJIG: amont en l’elme agu.
Do these twenty-nine examples follow any lines of agree- ment sufficiently marked to aid us in grouping the manu- scripts? In Case 15 IGL omit six lines which were in the ori- ginal, otherwise the un des lor of 896: would attach itself to nothing in the stanza. Acting upon this indication, if we review our list of examples to see whether a grouping IGL
Qot nm =
will explain any other cases, we find that Cases 4, 7, and —
49 square with this hypothesis.
In Case 21 GL have a reading which, besides departing from the combined testimony of all the other manuscripts, has the air of being a stop-gap and which thus constitutes a link between the two versions. This connection will serve also to explain Cases 16 and 28.
_ We have now to consider J and K. If we group these two manuscripts together we reach an explanation for Cases 8, 9, 41, 13 and 17. We have previously seen K showing affin- ities with L. Judging by the comparatively greater weight of the testimony fora JK group as compared with that for a KL group, and by the fact that L itself shows a strong lean-
42 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
ing to an IGL group, it is more logical to suppose that the cases where K gives the readings of L are the result of con- tamination in K. That the contamination does not work the other way and that L does not go to K-for certain readings is indicated by the fact that K sometimes contains readings shared only by G and L, but that L never contains readings , peculiar to JK.
Likewise a contamination of G from the x group will explain why, in Case 3, Gis the only one of the x’ group to give the line; why, in Case 20, it is the only one of the group to have what is apparently the better reading; and why, in Case 1, it has the x reading. In Case 25, if we suppose amisté to have stood in the original, it is possible to see why G in search of a more correct rhyme should have taken the reading of D. |
_A contamination of J from L will explain Cases 23 and 27; also, if we suppose that the contamination went back to an earlier copy instead of to L itself, Case 22, where J has the reading of G against that of IK while L omits the line.
Up to this time only those cases have been considered where a complete solution depends solely upon the uncondi- tioned action of the above groups. Let us now, by various combinations of points in the foregoing suppositions, see if it is possible to strengthen these theories of agreement by reference to other examples. Granted that GL and JK form two groups, we can understand the distribution of readings in Cases 10 and 14. Granted that GL form a group, and that K is contaminated from L, we have an acceptable explanation | of Case 12. Granted that IGL form a group, and that J is contaminated from L, we can attempt an explanation of Case 2h, where the correct reading is possibly asin D and K. Granted that IGL form a group, and that G is contaminated from x, we can see why only I and L have an order of lines in Case 26 contrary to the testimony of all ‘the other manu- scripts, !
4, In this case it is possible that G might have become conscious of
MANUSCRIPT RELATIONS 43
At this stage we are left with a residue of five cases: 2, 5, 6, 18 and 29. Of these the first two present testimony which is capable of various interpretations, and consequently leads to no definite conclusions. I incline to the belief that in both cases I and K have preserved the original readings of the x’ _ group, rather than that there is any connection of contamina- tion between these two versions. The phrase Calnus de Melite occurs many times throughout the poem, and it is possible that in Case 2 an ancestor of GL inserted this expres- sion which subsequently became distorted as it appears in the group, and which by contamination from L found its way into J. In Case 5 the reading of GL may have been originally as in G. If so, J took its version from an ancestor of L, and then L took from F its reading which approximates that pre- served in D. In Case 6 it is fairly sure that the reading of the X group was avenant and of the x’ group en avant; F and J accord in giving la vaillant. When we consider however that - in the stanza vaillant occurs as rhyme word twice in DFHI and K, three times in G and L, and four times in J, we are justified in admitting that it is possible for the two manu- scripts to have introduced the expression independantly. Likewise in Case 29 L joins H in a reading which at first sight it would seen impossible to attribute to chance, but as in 1478 we see H making the identical departure under circum- stances nearly identical, the case loses some of ifs weight, and Wwe may say that it is not beyond the bounds of possibility that this isa chance agreement. Equally if to chance and the force of the cliché we accord the resemblance of D to JK in Case 18, we have here but another example of the group. JK, with K introducing arragon independantly or perhaps J taking the rhyme word from L.
What then from these last five examples are our conclu- sions regarding the manuscript scheme ? Hardly anything definite. I consider that inx’ we have to do with a group of
the lack of continuity resultant from the inversion and have corrected ' without recourse to the x group.
44 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
manuscripts in regard to whose position, after having demon- strated their general and usual relations, we must content ourselves with calling attention to the various and conflict- ing evidence, a statement of what we consider the most prob- able lines of contamination, and the declaration of our fail- ure to find the links in the chain of the other divergencies.
In terminating this chapter, it may be worth while to note that Stanza 40 furnishes a possible indication that all our man- uscripts are derived from a common source.already reworked. Such an inference is suggested by the length of the stanza— 67 lines, as against 40 lines in the next longest stanza and 23 lines as the average length of the remaining seventy-three stanzas— and is borne out by an analysis of the content. _ After seven introductory lines treating of the relation of clothes to character, the narrative is resumed with a fifteen- line account of the lineage and proceedings of the two bache- lers who have come in search of Alexander. Thereafter for thirty-three lines the reader is subjected to a charivari of medizeval commonplaces on the degeneracy of the present age contrasted with the times of our forefathers, concluding with a laboured résumé of antecedent events obviously intend- ed as a link to what follows: the whole quite lacking in coherence and stylistic finish. Since this digression is of a type absent from the rest of the poem and since Gui de Cam- bral normally evidences a fairly clear-cut style, the authenti- city of these thirty-three lines seems far from assured. Should we omit them the sense of the stanza is in no wise disturbed - and is even bettered: the twelve remaining lines, which deal with the joy of Calnus at the arrival of the two young men and with the details of their landing, form a smooth sequence to the first twenty-two. The resulting stanza of thirty-four lines becomes one of normal length for the poem. In the absence of definite proof that the passage is interpolated, I shall retain it in the text, but here record my doubt of its authenticity.
MANUSCRIPT RELATIONS 45
3. —= GENERAL CONCLUSIONS
From a study of the manuscript relations of the Vengement - the following results are obtained:
‘ 4) From the great quantity of individual variations which can be remarked in each of the manuscripts, it is possible that many variations in which two or more manuscripts agree are due not to a connection between the manuscripts but to chance concordance in individual variation. Arguing from this principle I have indicated in a general way the nature of such agreements as have no positive import for determining manuscript relations.
2) We find recurring cases of agreement which, due to their frequency and to the difficulty of explaining them all as the result of chance, enable us to recognise some connec- tion between certain manuscripts— that is, to place them together in a group. Among these groupings the following are the best assured: HDFJ, (x), IGLJK (x’), and the sub- group DFJ, in which F and J, show signs of still closer con- nection. We are probably justified also in dividing the x’ group into two sub-groups: IGL, which itself contains a sub-group GL, and JK.
3) Several manuscripts of the x’ group slags an affinity for readings in manuscripts with which, to judge by the group- ings already established, they should have had no relation. We are thus led to suppose contamination for those cases. I have pointed out that it is quite possible for a scribe working with a supplementary manuscript to take from it an occasion- al word or phrase while in the main he follows his basic text. The two contaminations which we can assert with the greatest degree of probability are that of L to a manuscript closely related to F, and that of K to a manuscript closely related to L after L had received its contamination from F. For the other contaminations we cannot allege cases equal in number or in weight, but is seems possible that G is contam-
46 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
inated from the x group, possibly from a manuscript related to D, and that J is contaminated from a manuscript closely related to L. before L had received its contamination from
4) A very few instances of agreement which go contrary to the proposed groupings have been noted, and I have sought, wherever possible, some explanation. Any inadequa- cies I impute tomy own limitations rather than to the incor- rectness of the groupings previously postulated.
In graphic form my conclusions are as follows:
CHAPTER IV
REMARKS ON THE ESTABLISHMENT OF THE TEXT
After having attempted to classify the manuscripts of the Vengement Alixandre of Gui de Cambrai the next step in our study is to see if there is any manuscript to which we may give preponderant weight in establishing a critical edition of the poem. In discussing this question two procedures are at our disposal. We may consider the manuscripts separately, and by means of stylistic and technical criteria reach a con- clusion as to which manuscript presents the best version from a literary point of view. Or we may consider the manu- scripts collectively, and, having once established the grouping of the manuscripts, by a process of elimination arrive finally at that manuscript which can be said to be freer from faults than the others.‘ For our present problem I prefer the second method, as offering less scope for the personal element.
In the Vengement we possess eight manuscripts which fall into two groups. In view of this fundamental division, we may greatly simplify our problem if we find it possible to - peject one or the other of these groups and to assert that any manuscript found therein, if only by reason of its inclusion of all the errors proper to the group, is unsuitable as base for an edition. Our examination of the two groups has shown us that in two definite cases x’ (i.e., IGLJK) is unquestionably
4. In this connection it is necessary to distinguish fault or error from difficulty of interpretation. It is possible that a manuscript chosen by the second method may present many cases where the reading offers difficulty, while in other manuscripts the passage offers an acceptable sense, but where nevertheless the first manuscript can be shown to be nearest the original and the others to be later reworkings of an imperfectly understood passage.
Google =
48 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
in error and that in one other case it omits some sixty lines which are necessary to the development of the story. It also omits the dedication of the poem and the name of the author (lines 48-55). These errors and lacunae are in themselves suf- ficient to lead us to prefer the x group. There is another argu- ment against the majority of manuscripts contained in the x’ group. Traces of contamination appear in four out of the five manuscripts, most strongly marked in L and K, and present also in G and J. These manuscripts then may be rejected, and out of the whole x’ group we find but one manuscript, I, capable of serving as base. But [hasthe abovementioned group faults ,and unless the three manuscripts of the x group present omissions, errors,andindividualreadings suchas to fall below Lin the matter of trustworthiness, it is to them that we must turn. Of these manuscripts we may at once reject F because of its fragment- ary nature and its frequent errors. Of the two that remain, D and H, H presents a far smaller number of individual errors than D. For example in Stanza 43, if the text of H were the sole to be preserved, no hesitation would be felt in- printing it as it stands. If we had only the text of D, in three places we would be sure that the copy was at fault: in line 369 the first hemistich l’eve est entour is unmetrical; in lines 375 and 376 donjons is repeated as rhyme word; in line 382 the omission of the negative renders the following passage unintelligible. The concordance of the H reading with that of the majority of the x’ group in all three examples shows clearly that H has not brought individual corrections to the aid of a corrupt passage but that it preserves instead the ori- ginal. In H the proportion of obviously incorrect readings is much less thanin any other manuscript save perhaps in I, which for other reasons we have rejected. In the first thirty stanzas of H there are only two cases where we may, even before comparison with the other manuscripts, say definitely that H is at fault: in 499 where the reading dist for ont is clearly absurd, and in 683 where the faulty metre shows H to be in error. As if in compensation, in 242 H alone of all the
THE ESTABLISHMENT OF THE TEXT 49
manuscripts has the correct reading.! My conclusion is, there- fore, that H should be used as the basic manuscript in an edition.
Having chosen the manuscript which we consider nearest to the best tradition of the text, it now remains to decide how closely, in a critical edition, we shall follow the readings of this manuscript. Wherever possible I shall follow the basic manuscript and shall substitute other readings only: (1) where the basic manuscript presents an obvious error or absurdity, (2) where it presents a reading fundamentally dif- fering from that attested by the agreement of the other manu- scripts. Even here those cases are to be eliminated in which, as seen above, ? concordance among several manuscripts may be due to chance. Thus, to take the first examples that occur, in 26 the reading of H que il of is to be preferred to that or DIGJK qu'il avoit since manuscript testimony of this sort counts for nothing. On the other hand, in 13 the reading of _ H irés is not to be retained since all the other manuscripts give retez which is not close enought to irés either in appear- ance, sound, or meaning to justify the theory of independant substitution. A critical edition produced in accordance with these principles will follow one single manuscript in the vast majority of cases, but will not fail to weigh the testimony of other manuscripts for readings which seem to represent more closely the author’s original conception.
In certain cases where the concordance of the manuscripts against H leads me to adopt the reading of the majority, there occur slight divergencies among the individual manuscripts, even when the general trend of these manuscripts points con- clusively towards the main outlines to be followed. For example in 20 the reading of H dont onparlast tousjors is con- tradicted by the reading of DF li miaudres c’onques fust and by that of x’ li meudres qui ainc fust. But in adopting a read-
4. See supra, p. 25. 2, Pp. 413-24.
50 MANUSCRIPTS OF THE VENGEMENT ALIXANDRE
ing other than that of H shall we print c’onques or qui ainc? In view of the fact that H itself belongs to the x group, the fundamental principle will be to take the necessary variant readings wherever possible from this group, and in general from D, since F is available only in a limited number of cases. Sometimes the x’ group can be definitely shown to have a reading preferable to that of the x group: the general rule in these circumstances will be to take the reading from I, since this is the sole manuscript in the group not showing signs of contamination, but these cases are so infrequent as to admit of individual discussion whenever they occur.
BIBLIOGRAPHY AND ABBREVIATIONS
AlGr. — Meyer, Paul, Alexandre le Grand dans la littérature francaise du moyen age. Paris, 1886. 2 vols. |
Appel, Carl, ed. — Gui von Cambrai, Balaham und Josaphas. Halle, 1907.
EM. — Elliott Monographs.
FF. — Filologiska Foreningen i Lund. Lund, 1897 ff.
. Jacob, P. L. [Lacroix, P.] — Dissertations sur quelques points
curieux de Uhistoure de France et de Vhistotre littéraire, n° vu.
Paris, 1839.
- Meyer, Paul. — « Etude sur les Manuscrits du Roman d’Alexandre ». Rom., 11 (1882), pp. 213-332.
Mich. — Michelant, Henri, ed.: Li Romans d’Alizandre par Lam- bert li Tors et Alexandre de Bernay. Stuttgart, 1846. [BLVS.., Vol. 13.1 |
Migne, Abbé J. P., ed. — Dictionnaire des manuscrits par M. X*™*, Paris, 1854. 2 vole. [Nouvelle Encyclopédie théologique, Vols 40-41,]
RA. — Le Roman d’ Alexandre.
- Sachrow, Karl. — Ueber die Vengeance d’ Alexandre von Jean le Venelais. Halle, 1902.
Schultz-Gora, Otto, ed. — Die Vengeance Alizandre von Jehan le Nevelon. [Privately printed, 1902. 50 copies only.]
Vengement. — Le Vengement Alixandre of Gui de Cambrai.
Venjance. — La Venjance Alixandre of Jean le Névelon.
Walberg, E. — Review of the above-mentioned works of Sachrow and Schultz-Gora. Rom., 32 (1903), pp. 150-60. |
Walberg, E. — « Classification des manuscrits de la Vengeance d’ Alexandre de Jean le Nevelon. » FF, 1906, pp. 5-30.
VITA
Epwarps, Bateman. — Born: Bangor, Maine, April 24, 1898. Educated: Bangor High School, Bowdoin College (B. A., 1919), Princeton University (graduate student 1919-20, 1922- 24. Fellow in Modern Languages 1922-23, Proctor Fellow 1923-24), Sorbonne 1924. Instructor in Romance Languages, — University of Western Ontario 1920-22. |
PROTAT FRERES, IMPRIMBURS, MACON (FRANCE). — MCMXXVI
Original from
on fe ~N Digitized by Google UNIVERSITY OF MICHIGAN
+75 eee* & ©
. — aa TEreietsy 2) . : TVs LOVETT GT ae TT eer eee ere re eee Tees 12 Pits Pa ee eee eT ET eee eat eres +3 . ri 'y - ; e ect . Lak;
; — ° ° 7 > - , ~ - . — . > — o-® > . ° * e«* 2+ ee er oo oe ae oe eee oe 8 ~ . : ese : - . ‘ 7 . ~ thas ives neee4 -
° ee @- * -*
~ a ee ee ee e+@ -%=+4 - - . a ee . . . . . o- e-b +e ee 5s * . i
° ~ . . ° ee ae ee a eh ot bk et ee. aoe DK ee ek ok Ok Be ae Be ee ee bes woes oe eee a oa noe eee. 3 -* o-* - Bet we) Oe 68 2-2 O86 OO Ere e-< -- : : : :
e ee ’ ot -+- Oe @ ~ .
. - ee Oe oO OO OO Oe Ome ETT ST Fore @ yes * pees $2 ere Soo = bts oe © —0-e Om 0 tee G+ -6 0-8 GeO
e” &-§ >see
IMPRIMEUSNS, N
¥
. a | wi -FRERES, . . als | e
* - ee i. Z “ “ “ae > : “5 = 4 4 bs > > a >>. ¢ ‘, 4 be - of Pm, a -- + * n a. A » —" > Yo” — é a - 7 7 _ * » a “ * 9 “ eee FF + . m - <— i” we ef * Lh - . : A -. a to a > ie a “d és oe? ed + ales . . , » : a > , r ry ’ —, a 3 ° = - _ « rp Ses “a. : - ¥ OE ne a- ~ ; ; ~~ « » _ *> . j A th “ao en, ae Ue >) } « a *“ hee iy = ‘ * zi % ree ” “ 7 { say ¥ * ae oo |) » "A a ¥ ‘ 4 a, estes - oe S| 74 i a a ' ea - ~—_ = BA Se ine a at Soe ite & y per. 4 tae hye alae rs of © ul, , 2 A ye 4 — - 7 a » ~ + i ao Be-% 4 - : ; : . ‘ i] w Ke ms a ‘. iss ae, y > < my te ay hse ae eee 2 ey - a Tc 5 mo 19 Je ‘ » SS » + ® sa s _% ro ia é