PQ 1463 G413Zc Cowper The Sources, Date, and Style of "Ille et Galeron" by Gautier d* Arras THE LIBRARY OF THE UNIVERSITY OF CALIFORNIA LOS ANGELES Ube Taniversits cf Gbicaao THE SOURCES, DATE, AND STYLE OF "ILLE ET GALERON" BY GAUTIER D'ARRAS SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF ARTS AND LITERATURE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF ROMANCE LANGUAGES AND LITERATURES BY FREDERICK A. G. fcoWPER Private Edition, Distributed By THE UNIVERSITY OF CHICAGO LIBRARIES CHICAGO, ILLINOIS Reprinted from MODERN PHILOLOGY, Vol. XVIII, No. n, March 1921 and Vol. XX, No. i, August 1922 TCbe TUmverstts ot Cbicago THE SOURCES, DATE, AND STYLE OF "ILLE ET GALERON" BY GAUTIER D'ARRAS A DISSERTATION SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF ARTS AND LITERATURE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF ROMANCE LANGUAGES AND LITERATURES BY . / FREDERICK A. G. COWPER Private Edition, Distributed By THE UNIVERSITY OF CHICAGO LIBRARIES CHICAGO, ILLINOIS Reprinted from MODERN PHILOLOGY, Vol. XVIII, No. u, March and Vol. XX, No. i, August 1922 7 - PREFATORY NOTE The following papers are adapted from a study made in partial fulfilment of the requirements for the degree of Doctor of Philosophy in the University of Chicago. The study deals with the sources, date and style of the Ille et Galeron, and with the patrons of Gautier d'Arras. I hereby take the opportunity to thank Professors W. A. Nitze, T. A. Jenkins, E. P. Dargan, E. H. Wilkins, and Karl Pietsch for their encouragement and suggestions in regard to form and contents, and Professor A. Coleman for securing me a rotograph of the Paris manuscript of the Ille. I am indebted also to Lord and Lady Middleton for their courtesy in allowing the Wollaton manuscript to be photographed, and to Mr. W. H. Stevenson for taking the manuscript to Oxford and superintending the making of the roto- graph, and also for his valuable information in regard to the text and miniatures. F. A. G. C. DURHAM, NORTH CAROLINA April 1922 DISS.1T" I THE NEW MANUSCRIPT OF ILLE ET GALERON The poem of Ille et Galeron by Gautier d'Arras has been known only from the very defective Paris manuscript (fonds frangais, 373). In 1911, Mr. W. H. Stevenson made a report to the British Manu- scripts Commission upon the manuscript treasures found at Wollaton Hall and quoted liberally from the prologue and epilogue of a new text of the poem. 1 A brief notice of this discovery was made in an obscure corner of Romania, in 1913 (XLII, 145). So far as I know, the only other mention of this find is in Professor Sheldon's article, "On the Date of Ille et Galeron," Modern Philology, XVII, 1919. 2 Through the kindness of Lord and Lady Middleton and Mr. Steven- son, I have been fortunate enough to secure a rotograph of the new text. A comparison of this with the Paris manuscript shows inter- esting and important differences. I shall here briefly indicate these differences and shall also discuss the conclusions reached by Professor Sheldon. Mr. Stevenson states that the new manuscript is in an early thirteenth-century French hand and in the Picard dialect. A careful examination of the new text indicates that it is in the hand of at least two scribes. The past participles of the first conjugation end in -t, as do nouns like gret. The Picard features differ as between the earlier and later folios of the text and certain Anglo-Norman features have been introduced. The only indication of the history of the volume is the name "John' Bertrem, de Thorp Kilton" (County York) in a fifteenth-century hand (fo. 347*0 . The text is in two columns of forty-seven or forty-eight verses each. It contains illuminated initials and seven miniatures in colors. 3 Practically 1 Report on the Manuscripts of Lord Middleton, preserved at Wollaton Hall, Notting- hamshire, Hereford, 1911, pp. 221 f. 2 Since this was written Brandin's edition of the Chanson d" Aspremont ("Classiques francais du moyen-age, " Vol. XIX), which is made from the Wollaton manuscript, has come to my attention. ' 157r. Lamb with banner of Cross; 158r. Boy Ille with dragon; 160r. Rogelyon in armor on horseback; 164r. Ille and the Roman emperor; 170r. Ille and Ganor; 175. Ille and the emperor again; 185r. Ganor. It is interesting to note that neither Duke Conain nor his sister Galeron are pictured, while Ganor and her father appear twice. The illustrator, at least, was more interested in the Roman part of the romance than he was in the Breton. C THE NEW MANUSCRIPT OF "!LLE ET GALERON" all of the text is legible. A few letters are blurred here and there, but almost all can be restored with the aid of the Paris manu- script. Two words are frequently written as one. The scribe used damaged parchment in several instances, for one page which bears stitches and several with holes show the text intact, written around the damaged spots. The poem contains 5,835 verses, 757 less than the Paris manu- script. This is the net loss, for 1,182 lines of the Paris manuscript are missing, while there are 425 new lines. Necessarily there are important differences. The losses of lines are mainly in the prologue, in Ille's earlier battles, and in the account of the courtship of Ille and Galeron. The chief additions are in the kidnaping and rescue of Ganor, and in the epilogue. There are innumerable minor changes of letter, word, or word-order, almost all of which clear up contro- verted points. The larger part of Foerster's notes are now obsolete. In many cases, Loseth's emendations are justified by the new text. The rhymes are generally exact. Identical rhymes and two couplets on the same rhyme-syllable are more frequent than in the Paris manuscript. There is but one lacuna, the rhyme-pair to verse 1255 (after 1938, Paris) being lacking. The verse does not occur in the Paris manuscript, and it very clearly does not belong where it stands. 1 Seventy-three lines of the old prologue are missing. The allu- sions to Germany are lacking and the eulogy of Beatrice is reduced in other ways. 2 Of the 13 new lines, one fills the lacuna after 117, 3 two are added to the discussion of Envy, 4 and after 131 are added the ten following: W. 63 Molt par me torne a grant anui Quant ainc ma dame ne conui; 65 Molt me fust encor plus soef . Or m'estuet sigler a plain tref Por c.als ataindre qui ains murent Et qui ainc (1. ains) de moi le conurent. Tols les premiers volrai ataindre; i P(aris) 1255 "Icil i vint molt erramment. " P. 8-19, 23-54, 79-102, 107-10, and 132. s P. 46 " Tant come honors loe et conselle. " 4 W(ollaton) 57 "Li drois d'envie est une ardors Qui li fait hair les mellors. " THE NEW MANUSCRIPT OF "!LLE ET GALERON" 7 70 Car molt a entre faire et faindre. Servir le voel si com jo sai; Car a s'onor voel faire .i. lai De Galeron, etc. These lines might well be taken into consideration in connection with any argument regarding the date of the poem. Does Gautier mean that he did not know Beatrice until after the coronation at Rome, and that he wishes to enjoy as much of the new Empress' favor as those who had known her before she had risen to her full height of fame ? This would seem to favor, for the beginning of the poem, a date somewhere near August 1, 1167. Line 72, if it is Gautier's, is very important, for in it the poem itself is referred to as a "lai." Unfortunately we cannot compare it with the famous passage (P. 929-36) criticizing lais, for that passage does not occur in this version. If this passage belongs in the original manuscript, it clearly does not refer to any possible source in a lai d'llle et de Galeron: the lais which Gautier is criticizing are those of Marie de France, which were probably then enjoying great popularity in the French courts. In the description of the first battle fought by Ille against Hoe'l, his traditional enemy, when Ille returns from exile in France (P. 277-546) , we find many lines in changed order. While 20 new lines appear, 291 are missing, including all the plays on the numbers of knights and those where the French knights show a certain nervous- ness (P. 447-63). The 100 lines recounting the exploits of Bruns d'Orleans and Estout de Langres (P. 578-677) are absent, and the role of Hoe'l is greatly abbreviated. In the episode of the battle with Rogelyon, the rejected suitor and nephew of Hoe'l, 62 lines are missing, while 9 are added. In the courtship of Ille and Galeron 185 lines are dropped, 13 added. The monologues of the two lovers are entirely omitted, as is the pretty scene in which Conain drags from his 'sister the confession of her feelings. In this more primitive version, Conain offers Galeron to Ille, and, when the offer is accepted, goes and tells the girl to get ready at once for the wedding. We are here closer to the spirit of the chanson de geste than to that of the courtly epic. The faulty connection at this point indicates, however, that s THE NEW MANUSCRIPT OF "!LLE ET GALERON" at least some of the lines in the Paris manuscript belonged in the original. The important episode in which Ille lost an eye is quite different. The 32 lines (P. 1625-56) which tell of his triumph in the tournament and his unlucky decision to try just one more tilt are missing and in their place are the six following: W. 981 Un jor estoit en une guerre; Si prist le segnor de la terre Devant le castiel qu'il avoit. Mais uns de als que il tenoit Al rembarer la forteresce Retorne al pont et si s'adrece. De la lance qu'il porte en destre Fiert Ylle, etc. This decided difference between the two versions suggests the possibility that Gautier himself made two versions of his poem, one for Beatrice and another for Thibaut. The absence of the tournament scene from the Wollaton manuscript recalls the opposi- tion to this form of sport. It had been forbidden by a papal decree of 1131, renewed in 1139, 1 participants were threatened with excom- munication, and ecclesiastical burial was to be denied anyone who might be killed. The episode in which the wounded Ille slips away to a castle so as not to see his wife, and she succeeds in getting into his presence, is much improved by the insertion of the following lines after P. 1754: 1069 A bien petit que ne se tue; D'uns dras a home s'est vestue. In the catalogue of countries visited by Galeron in further pur- suit of her elusive husband, there are some marked changes. In verse 1295 (P. 1988) Bresaliande replaces Nohuberlande; in 1297 (P. 1990) Auvergne is replaced by Norouerge and Normendie is added; in 1301 (P. 1994) Esclavonie is replaced by Bougerie; and two new lines (1306-7) after P. 1998 bring in Borgoigne. In Ille's first battle for the Roman emperor against the Greeks there are only minor changes; 33 lines are added, 23 subtracted. 1 Young Henry of Champagne and the king's brother Robert held a great tournament at Easter, 1 149, in spite of the very vigorous efforts of St. Bernard to induce the Abb6 Suger, regent of Prance, Count Thibaut (father of Henry and our patron Thibaut), and other notables to forbid it. See Arbois de Jubainville, Histoire dea dues et des comtes de Champagne, III. 21-24. THE NEW MANUSCRIPT OF "!LLE ET GALERON" 9 The second battle, in which the seneschal is killed, is substantially the same in both versions. In the third battle, where Ille com- mands as acting seneschal, the differences are more numerous, though of little importance: 135 lines are dropped, 16 added. After P. 3504, the following new lines add clearness to the emperor's offer: 2674 Ma fille aura a son deport Et tolt 1'empire aprie"s ma mort. The next important changes are in the scene in which the messengers report their vain search for Galeron. The Wollaton manuscript omits the entire speech in which Ille laments his loss (3897-3938), as well as the 26 lines in which is related Ganor's eagerness for a speedy wedding (3956-79). In the account of the festivities on the eve of the wedding, one adds to the list of quota- tions attesting the popularity of the Breton lais: 3094 (P. 3984) Cil jogleor harpent et notent, Vielent et cantent et rotent Ces lais bretons entros qu'en son. In the scene at the church door, the Wollaton manuscript omits the 31 lines (P. 4225-55) in which Galeron expatiates upon the prayers she will offer for Ille if he will place her in a convent, and substitutes for them five of a more worldly and realistic type : 3344 "Se tos li mondes ert a moi Ne me valroit il rien sans toi Ne me poroie joie atendre." Cil le voit bele et blance et tendre Et voit le cors bien fait et gent. Ja le baisast devant la gent (P. 4256). In the account of Ille's second visit to Italy, several scenes are amplified. The messenger who informs him of the abduction of Ganor gives him directions as to the best means of waylaying the abductors. The attack and the rescue are described in greater detail, 60 new lines appearing. Twenty-eight additional lines by way of summary, and 28 in further description of the joy of the newly wedded pair and their court, mark the remaining important additions to the body of the poem. The 30 new lines of prologue will be men- tioned in connection with Professor Sheldon's article. 10 THE NEW MANUSCRIPT OF "!LLE ET GALERON" In his interesting and illuminating discussion, Professor Sheldon attacks the generally accepted dating of Ille. He criticizes Foerster's statement that the poem must have been composed shortly after the Roman coronation, August 1, 1167, mentioned in verse 69, 1 and pleads for a later date. He considers that the critics who have given 1167 or 1168 as the date of the poem have failed to prove their point. I agree with him that the only points absolutely fixed are 1164 as the earliest date for Eracle, 1167 the earliest for Ille, 1191 the latest for Eracle, and 1184 the latest for Ille; but I do not quite follow his argument for a later date for Ille. He says first (p. 385) that the poet's reference in the prologue to the coronation does not preclude a much later date than 1167, as the coronation was important enough to be mentioned at any time; second, that it is doubtful if Gautier would have written his prologue, or retained it if written, while the Empress was in Italy or during the flight from Rome, because she would not have been in a receptive mood for the poet's offering then or for some years after, perhaps not till 1174, or even until after 1178. Professor Sheldon himself is not fully satisfied with the validity of this argument, for he says (p. 391) : The tantalizing lines 9*-18*, with what may have immediately preceded them, seem to allude to something that caused an absence which led him to consider another patron, though he had not lost hope of some recognition from the Empress. 2 Did he perhaps begin his poem while the Empress was in Italy, hoping for her return before or soon after its completion, and then because this return was delayed (in which case we should naturally think of her stay of nearly four years in Italy, 1174-78), or because he had some other reason, whatever it was, did he finish with praise of the count as well as of her? Whatever had happened, it looks as if a fairly long interval elapsed between beginning and end. This latter point of view (except for the dates 1174-78) seems to me the more nearly correct. The poet was about to compose his work in honor of the new Empress. He was determined to win as much favor as those poets who had known her longer, but, for some reason, in his epilogue he changed his dedication to another patron, Thibaut, whom he applauded as her equal. Was not this action eminently appropriate to the black days after the coronation and 1 W. 25, " Rome le vit ja coroner. " 1 The lines are less obscure if 7* and 8* are placed before 5*, according to a suggestion made by Professor T. A. Jenkins. THE NEW MANUSCRIPT OF "!LLE ET GALERON" 11 the flight from Italy ? If Beatrice had been in a mood for generosity, the poet would have had no reason to seek another patron. But, as Professor Sheldon himself says, there is nothing conclusive about any of this argument. There is, however, a possibility of narrowing down somewhat the question of the date. Professor Sheldon is convinced from verses 6592-1 * : W. 5805 Gaiters d' Arras qui s'entremist D'Eracle ains qu'il fesist cest uevre, that Eracle was written before Ille. I think that the definitions of the verb s'entremetre given in Godefroy will bear me out in my claim that the only thing proved by these lines is that Gautier began Eracle first. The text of Eracle bears the marks of having been written in three different parts, of which at least one was written after Ille. This opinion is strengthened by evidence in the Wollaton manuscript. The argument for the conclusion that Ille was finished before Eracle may be briefly stated as follows: For Eracle there were three patrons or three phases of patronage : (1) Thibaut V of Blois, alone mentioned in the prologue; (2) Thibaut and Marie of Champagne, his sister-in-law, mentioned at the begin- ning of the epilogue; (3) Baudouin of Hainaut, mentioned in the epilogue as cause of the poem's completion and the person to whom it was being sent. For Ille there are two patrons : (1) Beatrice of Burgundy, Empress of Germany, alone mentioned in the prologue; (2) Beatrice and Thibaut, mentioned in the epilogue, the former as the cause of the beginning of the work, the latter as the cause of its completion. If we accepted the theory that Eracle was completed before Ille, we should be obliged to take with it not only the conclusion that Gautier broke off with Thibaut and Marie, and finished Eracle for Baudouin, but also that he thereupon began a work for Beatrice, deserted her, and returned to his former patron Thibaut. Is this probable ? I am convinced that Ille must have been completed during the period when Gautier was working for the Champagne-Blois group arid before he attached himself to Baudouin, consequently that Ille, while begun later than Eracle, was completed before it. We are now confronted with the question as to whether the last patron is Baudouin IV or Baudouin V: if the former, both poems 12 THE NEW MANUSCRIPT OF "!LLE ET GALERON" must be placed before 1171, the year of his death. That would allow a margin of four years from 1167, the year of the coronation, when Thibaut's interest in Ille had not yet been sought. The closer Era.de is placed to the later date, the farther may Ille be removed from llt>7, but at the extreme outside it could hardly be later than 1170. In case Baudouin V is the patron, the problem is no nearer settlement than before, 1184 for Ille and 1191 for Eracle being the limits. Foerster preferred Baudouin IV, considering Baudouin V as too young. 1 Professor Sheldon states that this choice "is of doubtful correctness," but does not give his reasons. I hope I have shown that his whole plea for a later date for Ille depends upon that choice being incorrect. In my own investigation of the subject, I have preferred Baudouin V, largely for the reasons that he was known as a patron of letters, while his father was not; that he was brother- in-law of that well-known literary patron, Philip of Flanders; and that he was in decidedly close relations with the courts of Cham- pagne, Blois, and France. But the whole matter still rests upon too slender a basis of evidence to be at all satisfactory. 2 Ille und Oaleron von Walter ton Arras, herausgegeben von Wendelin Foerster, Halle, 1891, pp. xv-xvi. 2 The following errors occur in the report of the British Manuscripts Commission and were reproduced by Professor Sheldon in his paper: Page 388, verse 3. com, manuscript con. pens, MS reads pens. "Our poem begins on folio 158 recto, etc." It actually begins on folio 157 recto and ends on folio 187 verso. Page 389. P. 6579 (W. 5790) o non. MS anor. Page 390. 6* En vie, MS Envie. 18* me, MS m =m'en. 11*. 15*. and 22* MS reads q =que. Page 391. 25* MS reads liu. Par and never pur is found in the manuscript wherever unabbreviated. M' It is never written out, but is found once in rhyme with tolt, i.e., tout. THE SOURCES OF "!LLE ET GALERON" 13 Much as has been written upon the sources of the Ille et Galeron of Gautier d'Arras, there is room for additional work. Along what line it can be most profitably directed, I hope to point out in this article. The sources of the Ille may be conveniently divided into two groups: (1) written literary sources and (2) contemporary historical events or court gossip. The written sources have been extensively discussed by Paris, 1 Foerster, 2 and Matzke, 3 and I shall merely sum up their conclusions with my own deductions added. Gautier himself gives his source as an estore* and claims to have followed it faithfully. But scholars are inclined to accept two works as written sources, the Chronique de Nantes 5 and Marie de France's Lai d'Eliduc. 6 Ferdinand Lot 7 argues that "le commencement, les 1500 premiers vers environ, repose sur un fondement historique tres de'figure', re"el cependant." He then relates from the Chronique de Nantes the inci- 1 Gaston Paris, La Poesie du moyen-age, 2 sfirie (Paris, 1895), pp. 109-30; Histoire litteraire, XXX (1888), 9, 600; La Litterature francaise au moyen-age (1888), p. 113; Romania XXI (1892), 275-81; Journal des savants (1901), p. 706. J Wendelin Foerster, in the introduction to Ille und Galeron von Walter von Arras, Rom. Bibl. 7, Halle, 1891. 1 J. E. Matzke, "The Source and Composition of Ille et Galeron," Modern Philology, IV (1907), 471-88; "The Lay of Eliduc and the Legend of the Husband with Two Wives," ibid., V (1907), 211-39. P(aris) 6590 "Ne en 1'estore plus n'en aut; Ne plus n'i a, ne plus n'i mist Gautier d'Arras," etc. W(ollaton) 5803 "Ne en 1'estorie plus n'en ot, Ne plus n'en a, ne plus n'i mist Gaiters d'Arras," etc. 6 Ren6 Merlet, Chronique de Nantes, "Collection de textes pour servir a l'6tude et a 1'enseignement de 1'histoire." XLX, Paris, 1896. Karl Warnke, Die Lais der Marie de France, Bibl. Norm. Ill, Halle, 1900. 7 Ferdinand Lot, "Une source historique d'lile et Galeron," Romania, XXV (1896), 585-88. 14 THE SOURCES OF "!LLE ET GALERON" dent of the assassination in 981 of Hoel, count of Nantes, by Galuron, emissary of Conan, count of Rennes, and concludes: En somme, les comtes de Nantes et de Rennes et leurs vassaux out 6t6 1'objet, au X* siecle, de remits ^piques en langue fransaise tout comme ceux des autres provinces. Le d6but d'llle et Galeron me parait en avoir consent un souvenir loin tain. Galuron, he"ros d'un de ces remits, aura du h sa cele 1 - brit mme d'etre gratifi6 du role bien different d'Eliduc. 1 Lot's opinion has been quite universally accepted. It is not so with the question of Marie's Eliduc as principal source. On this subject there are two main schools, the French, which follows G. Paris' opinion that a lost lai of Eliduc served as common source for Marie's poem and Gautier's; 2 and the German school, headed by W. Foerster, 3 which maintains that Ille et Galeron is based directly on Marie's Eliduc and is intended to be a moral re-working of the theme. Gaston Paris 4 bases his opinion on the fact that the episode of the injured eye, upon which hinges the whole action of the Ille, is absent from the Eliduc. Paris is undoubtedly right if we can accept at full value Gautier's statement about his estore. If he is faithfully following it, Marie's Eliduc cannot be the direct source. And if we do not admit Gautier's statement, we do not exclude the possibility that he had another source either in common with or different from the Eliduc. But Foerster's view has great plausibility. The two poems have great similarity of plot and incidents, though the motive for leaving court seems at first thought to be quite different. Eliduc leaves court through loss of favor, and, tired of idleness at home, goes abroad to find military activity. Ille loses an eye (in a tourna- ment in P., in a war in W.) and will not return through a mingled pride and humility. He seeks military service abroad to re-establish his prestige and self-respect. The loss of the eye is a symbol of loss of prestige at court, and according to the Ovidian love casuistry as expounded in the Courts of Love 5 a symbol of the loss of ability to love. The loss of prestige in each case brings about the temporary 1 Op. cit., p. 588. 2 La Literature franyaise au moyen-age, p. 113. Foerster, op. cit. * Romania, XXI, 278. 'Andreas Capellanus, De Amore, ed. E. Trojel, Havniae, 1892. THE SOURCES OF " ILLE ET GALERON " 15 separation of husband and wife and leads to the introduction of the second woman. The other important incident which makes possible the hero's second marriage is quite different in the two poems. Galeron's accomplished vow after recovery from the perils of child- birth replaces the shipwreck and resuscitation of Guilliadun; the realistic replaces the miraculous and magical, and civilized morality is strictly observed. Without going more fully into this question, I give a tabular comparison of the similarities, and also, that it may be seen that the Ille is not merely an expanded copy of the Eliduc, of the differences as well. Matzke did this to some extent in his articles, but since he was bent upon upholding G. Paris' theory, he did not draw the fullest possible conclusions from the material. ELIDUC 1. Source stated as a lai, verses 1-4 2. Eliduc is hero 3. The wives are Guildeluec and Guilliadun 4. Eliduc is vassal of the king of Bretagne 5. Eliduc is seneschal of the king of Bretagne 6. Eliduc leaves for a foreign country 7. He goes to Lougres in Britain (Alliteration and similarity of spelling) 8. Eliduc offers his services to an aged king who has a daughter as sole heir and is being attacked by a rejected suitor 9. Eliduc defeats the enemy and is made guardian of the land 10. King's daughter falls in love with Eliduc ILLE ET GALERON 1 Poem called a lai (W. 73) . Doubt- ful whether reference to lai in P. 929 is to source. Ille, the hero, is son of Eliduc (P. 134) The wives are Galeron and Ganor (epic alliteration) Ille is vassal of the duke of Bretagne Ille is seneschal of the duke of Bretagne Ille leaves for a foreign country He goes to Langres (MS Lengres) in Burgundy on way to Rome Ille offers his services to an old emperor who has a daughter as sole heir and is being attacked (2001- 5238). In the second war this enemy is a rejected suitor (5403) Ille defeats the enemy and is made seneschal Emperor's daughter falls in love with Ille 1 I omit mention of Ille et Galeron to vs. 927 (Paris), as this part must have had another source, as Lot suggested. 16 THE SOURCES OF "!LLE ET GALERON" ELIDUC 11. Report that lands at home are laid waste 12. Eliduc informs king that he must return home 13. King offers part of his inheritance 14. Eliduc promises to return with aid in case of need 15. Eliduc goes to take leave of Guilliadun 16. She faints in his arms 17. When she comes to, he prom- ises to return if she sends for him 18. He kisses her on leaving 19. He is received at home with great joy 20. He pacifies the land 21. He returns to the land of the second woman 22. He marries Guilliadun (at his home) 23. Guildeluec enters a convent (at end of story) ILLE ET GALERON Report that lands at home are laid waste Ille informs emperor that he must return home Emperor offers part of his posses- sions Ille promises to return in case of need Ille goes to take leave of Ganor She faints in his arms When she comes to, he promises to return if he hears that she is in danger He kisses her on leaving He is received at home with great joy He pacifies the land He returns to the land of the second woman He marries Ganor (at Rome) Galeron enters a convent (before Ille thinks of loving Ganor) The principal differences are as follows : ELIDUC 1. Length 1,184 lines Length 2. Hero already married. Ban- ished from court through slander of enemies 3. Liege lord is king of Bretagne 4. Wife's family not mentioned 5. Hero does not wish inactive life at home 6. Hero goes to Britain 7. Goes with 10 knights 8. Guildeluec stays at home and governs estates 9. Enemy is rejected suitor ILLE ET GALERON Paris MS 6,592 lines Wollaton MS 5,835 lines Treats of courtship and marriage of Ille and Galeron. Hero success- ful over enemies (after childhood banishment) Liege lord is duke of Bretagne Ille marries duke's sister Hero loses eye in tournament (or war) and leaves through pride and humility Hero goes to Rome Goes alone and incognito Galeron searches for her husband Enemy is seemingly political in first campaign, but is suitor in second THE SOURCES OF " ILLE ET GALERON 17 ELI DUG 10. Eliduc binds himself for one year. 11. Accepts advances of princess, conceals fact that he is married 12. Is offered third of kingdom 13. Eliduc summoned home by his lord to help him 14. Eliduc sad at leaving Guillia- dun, kisses her while she is in faint. Refuses to take her merely because he cannot honorably. Promises to return at her request. They exchange love tokens on parting 15. Eliduc sad at home. Wishes to end war and return to Britain as soon as possible. No children of first marriage mentioned 16. Eliduc goes to Britain to abduct Guilliadun 17. Guilliadun leaves home when sent for by Eliduc to come to his ship 18. She learns of his marriage through incident in storm at sea 19. Trance of Guilliadun. Brought to by Guildeluec through red flower. Agreement of wife to enter convent makes second marriage possible 20. Eliduc desires second marriage 21. Eliduc abducts Guilliadun 22. No fighting on Eliduc's second visit to Britain 23. Eliduc does not gain in rank 24. No children of second mar- riage mentioned 25. Eliduc and Guilliadun go to convent to end their days 26. Ages and lapses of time left indefinite ILLE ET GALERON Ille binds himself for an unlimited time Declines advances of princess. Tells of his lost wife. Consents to marry only when messengers have scoured country in vain for Galeron Is offered half of empire and Ganor, and all after emperor's death Ille returns because of recovery of Galeron Ille sad for Ganor, but would not give up Galeron for her. Promises to return if he hears she needs protec- tion. Kisses her on parting through pity only Ille happy as duke of Bretagne. Wife has three children, but makes vow to become nun and therefore can no longer be his wife Ille goes to Rome on hearing of a Greek invasion Ganor comes to Bretagne to ask Ille's aid against the Greeks Ganor has known of Ille's mar- riage since his first visit to Rome Fear of death in childbirth causes vow to go to convent which makes second marriage possible Ille becomes ill over loss of Galeron Ille rescues Ganor from abductors Ille defeats Greeks and drives them from Italy on second visit Ille becomes emperor Four children by Ganor Ille and Ganor left at height of their happiness Ages and lapses of time definite 18 THE SOURCES OF "!LLE ET GALERON" The close similarity of plot and incidents, with the connection of the name Ilk, le fil Eliduc, points definitely to the lai of Eliduc as the chief source of the Ille. I do not think that the two poems come from a common original. No poem has been found which contains more than a general similarity of plot, nor which contains the ship- wreck and resuscitation scenes in addition to the incidents common to the Ille and the Eliduc. The differences are partly substitutions of realistic elements for the fantastic, partly the result of using certain elements of the Eliduc lai twice in order to expand the poem to a suitable length, and partly from the use of contemporary historical and other source material. Moreover, Marie states in the prologue of her Eliduc that she is giving the cunte et toute la raisun of a very ancient Breton lai that it may not be forgotten. Gautier says that he is following his estore closely. 1 If we believe both Marie and Gautier, the poems cannot have a common source and Paris' theory is untenable. Gautier in all probability knows the lais of Marie, for in lines 929 ff., 2 he objects to the popularity of the lais which smack of unreality, and Marie's collection is the only one known to have been in existence early enough for him to have used. If Marie's Eliduc is not Gautier's source, it (i.e., the estore) must be a re-working of her poem made by another contemporary poet, for it was only in the latter third of the twelfth century that it became the style to revise the old tales thus. It is much easier to believe that Gautier made an adaptation of Marie's Eliduc than that he copied a contemporary, unless he were translating from the Latin. In that case, he would probably have boasted of the fact, as in the Erode, 3 where he says that he is going to tell a story en romanz* and later, introducing Part III, he states 5148 Signeur, nous lisons en latin 1 This has always been the interpretation. But his lines taken literally mean merely that he ends where his source ends, i.e., he does not add anything to the ending. 2 P. 929 " Li lais ne fust pas si en cours, Nel prisaissent tot li baron. Grant cose est d'llle a Galeron: N'i a fantome ne alonge, Ne ja n'i troveres menconge. Tex lais i a, qui les entent, Se li sanlent tot ensement Con s'elist dormi et songieV' ' Oeuvres de Gautier d'Arras, Tome I, published by E. Loseth, Paris, 1890. < Vs. 103. I hesitate to use this, as it is questionable whether the reading should not be el TO mam. THE SOURCES OF " ILLE ET GALERON " 19 In addition to these two important written sources, there is evidence of probable influence in (1) the proper names, (2) the Ovidian love psychology, and (3) the features of style and versi- fication. Those of Gautier's proper names which do not occur in the Eliduc and the Chronique de Nantes may possibly be taken from Wace's Brut and Ron, Geoffrey of Monmouth's Historia Regum Bri- tanniae, the Eneas, Troie, Thebes, Tristan and the earlier Chansons de Geste, including especially the Roland and the Antioche. The itineraries are the regular routes for trade and pilgrimage from France to the Orient. The love psychology, derived from Ovid, 1 developed by the Provengal poets and popularized in the north of France in the Eneas, is in full flower here. The source of Gautier's versification and style is probably directly in Wace, indirectly in the Chansons de Geste and the Provencal love lyrics. 2 These are the obvious and more easily demonstrable sources, but we may well ask ourselves whether the origin of many episodes over which scholars puzzle and wrangle may not profitably be sought in the milieu of the poet. In 1842, Massmann 3 first called attention to similarities in Gautier's two poems between incidents in the lives of Eleanor of Poitou and Louis VII and of Beatrice of Burgundy and Frederick Barbarossa. Later critics were inclined to scoff at Massmann, but a careful study of the known facts in the lives of these persons with the more accurate historical information which we now have available, tends to justify Massmann in his conviction and to convince us that if these lives were not drawn upon either consciously or unconsciously, the similarities form a striking series of coincidences. Now what was Gautier's milieu? He was court poet of (1) Thibaut of Blois, who was husband of Alix, second daughter of King Louis VII; he wrote (2) for Marie, sister of Alix and wife of Count Henri I of Champagne, brother of Thibaut; (3) for Baudouin V of Hainaut, brother-in-law 4 of Philip of Flanders, regent of France; 1 See Wilibad Schroetter, Ovid und die Troubadours, Halle, 1908. 1 See F. M. Warren, "Some Features of Style in Early French Narrative Poetry," Modern Philology, III. 179-209 (October, 1905), and 513-39 (April, 1900): IV, 055-75 (April, 19O7). 1 H. F. Massmann, Eracliue, Deutsches und framOsisches Gedicht des 1%. Jhts, etc., Quedlinburg and Leipzig, 1842. We are not sure whether the patron was Baudouin V or his father Baudouin IV. See above, pp. 11-12. 20 THE SOURCES OF "!LLE ET GALERON" and (4) for Beatrice of Burgundy, second wife of Frederick Bar- barossa. From his connection with all these patrons, Gautier had excellent facilities for keeping up with the popular form of poetry and the latest scandal of the great. Is it not probable that he made use of a considerable amount of court gossip, and also of facts taken from the lives of those to whom he dedicated his poems ? Both Alix and Marie were true daughters of the famous Eleanor of Poitou. Granddaughter of William IX, the first troubador, Eleanor brought her southern literature and her poets to the court of France when she married Louis VII in 1137. 1 She established and presided over the so-called "Courts of Love." Since Eleanor encouraged in every way the poets and troubadors, it is not surprising that the incidents of her fervent prayers for the birth of a male heir, 2 her alleged adultery, 3 and her attempted abduction by Thibaut V of Blois and Geoffrey Plantagenet 4 after her divorce from Louis 5 should be touched upon and developed in literature. The allusions to Eleanor, more open in the Eracle, are rather veiled in the Hie. The most obvious is that of the attempted abductions of Galeron (P. 1539-71) and Ganor (P. 6125-6530). 6 Much more evident are the incidents which correspond with events in the lives of Thibaut, the patron, and Barbarossa, husband of the patroness. The Ille was begun for Beatrice and finished for Thibaut. 7 What more natural for Gautier than to laud in thinly disguised language the exploits of his patroness' imperial husband or those of the Grand Seneschal of France ? Thibaut was Grand Seneschal of France, Ille of Bretagne and later of Rome; both were twice married and had seven children. Thibaut's grandmother became a nun because of grief over the death of a daughter and a sister also took the veil; Ille's first wife, because of a vow, became a nun after the birth of a daughter. 1 Alfred Richard, Histoire des Comtes de Poitou, 778-1204, Vol. II (Paris, 1903). 2 Arbois de Jubainville, Histoire des Dues et des Comtes de Champagne, II (Paris, 1859-66), 379. Richard, op. cit., II, 93. Ibid., p. 108. ' Ibid., p. 107. Cf. Cliges, 2859-70 and 3621-3816. i Wollaton 5828 " Por qant por li le commencai Et por le conte le flnai." THE SOURCES OF "!LLE ET GALERON" 21 The parallels are still more numerous with the life of Frederick Barbarossa. 1 BARBAROSSA 1. Father known as "one-eyed" 2 2. Two successful campaigns against Greeks: the first in Third Crusade when, a youth, he accom- panied his uncle Conrad III to Holy Land; the second as Roman em- peror at Ancona and Tusculum 3. Repudiated first wife Adelaide von Vohburg, either for adultery or consanguinity 4. Second wife was Beatrice, heiress of Burgundy, whom he rescued from her uncle, who planned to deprive her of her estates. Her father was dead 5. Historical Greek emperor wished to rule Rome and unite the Greek and Roman churches. Had been married to aunt of Barbarossa 6. Hohenstaufen arms three lions ILLE Lost eye in a tournament or war Two successful campaigns against Greeks: the first as an unknown and laughed-at young esquire with one eye gone; the second as Duke of Brittany (at first incognito) Repudiated first wife when she became a nun Galeron Second wife was Ganor, daughter of Roman emperor, whom he res- cued from Greek emperor, who wished to marry her for her estates. Her father was dead Greek emperor of romance wished Rome, to unite the two empires. His first wife was a relative of Ganor and died from cruel treatment of husband Ille's arms a lion (ducal arms of Brittany a gold lion) Ille's army contained Germans and Romans 7. Frederick had Germans and Romans in his army in his second campaign These points of similarity seem to me to show that Gautier intended Beatrice to see her noble husband in Ille. He tastefully softened down the sensitive points in Frederick's career lack of children by the first marriage and the cause of his divorce so as to make them inoffensive, though these second marriages after separa- tion (the ostensible cause of separation was usually forbidden degrees of consanguinity) were so common among the nobles at that time as to make this seem unnecessary. Whether Gautier obtained his impressions of court life, history, and geography from direct observation or by hearsay, we cannot tell. His descriptions of the places he mentions are too sketchy to See Hans Prutz, Kaiser Friedrich I, 3 vols., Danzig, 1871. Also Massmann, op. cit., pp. 544 ft*. ! Prutz, op. cit., "Sein Sohn Friedrich der Einaiigige "I, 0. 22 THE SOURCES OF "!LLE ET GALERON" enable us to determine. Assuming that Gautier lived at Thibaut's court and not in his own home or in a monastery, we may reason- ably assert that he accompanied Thibaut on some of his frequent visits to the courts of Louis and Henri. The poet may even have gone on the Crusades with the Champagne nobles or he may have been with the embassy which Henri sent to the court of Bar- barossa in Italy in 1167-68. The Ille may very easily have been begun or in large part composed on that occasion. The prologue mentions the coronation at Rome, August 1, 1167, and the epilogue names Thibaut as well as Beatrice. The poet had one or both of these patrons in mind all the time he was composing his poem. This fact must have had an influence upon his work. The literary sources have been practically exhausted unless some new manuscript of romance or chronicle is discovered. In the history of the nobles and courts mentioned above lies our best oppor- tunity to add to the knowledge of the background and sources of Ille et Galeron. UNIVERSITY OF CALIFORNIA LIBRARY Los Angeles This book is DUE on the last date stamped below. Form L9-Series 4939 PAMPHLET BINDER UC SOUTHERN REGIONAL LIBRARY FACILITY t iiT iU) jl