This dissertation has been 6 5—3847 microfilmed exactly as received

DUGGAN, Joseph John, 1938- FORMULAIC LANGUAGE IN THE OLD FRENCH EPIC POEMS LE SIEGE DE BARBASTRE AND BUEVON DE CQNMARCHIS.

The Ohio State University, Ph.D., 1964 Language and Literature, general

University Microfilms, Inc., Ann Arbor, Michigan FOEMUIAIC LANGUAGE IN THE OLD FRENCH

EPIC POEMS LE SIEGE DE BAHBA5TRE

AND BUEVON DE CONMARCHIS

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By

Joseph John Duggan, B.A.

******

The Ohio State University 1964

Approved by

kiti/MCX (illlihUIaL m h/ ffaa Adviser Department of Romance Languages ACKNOWLEDGMENTS

The completion of this dissertation was made possible by the Woodrow Wilson National Fellowship Foundation which granted me a Dissertation Fellowship for 196h. The Ohio State University

Computer Center allowed me over two hours of free computer time for the execution of the computer program.

I wish to express my gratitude to Mr. G. Miller Clarke,

Assistant Programming Supervisor of the Chio State University

Computer Center Data Processing System, who patiently guided me through the preparation of the computer program; and especially to Professor Eleanor Webster Bulatkin, my adviser, whose encour­ agement and criticism were invaluable to me in this undertaking.

ii VUA

September 8 , 1938 Born - Philadelphia, Pennsylvania

I960 ...... B.A., Fordham University, New York, New York

1960-I96I .... Woodrow Wilson Fellow, Department of Romance Languages, The Ohio State University, Columbus, Ohio

1961-1962 • • • . University Fellow, Department of Romance Lan­ guages, The Ohio State University, Columbus, Ohio

1962-I963 • • • • Residence in Paris, France, for research at the Bibliotheque Nationals

1 9 6 4 . Woodrow Wilson Dissertation Fellow, Department of Romance Languages, The Ohio State University, Columbus, Ohio

FIELDS OF STUDY

Major Field: French Language and Literature

Studies in Medieval French Literature. Professor Alexander H. Schutz

Studies in Mo dem French Literature. Professors George Havens, Don L. Demorest, Stanley Aston, James Doolittle, Charles Blend and Hugh Davidson

Minor Field: Italian Literature

Studies in Italian Literature. Professor Luigi Borelli

Minor Field: Spanish Language and Literature

Studies in Medieval Spanish Language and Literature. Professors David Griffin and Kenneth Scholberg iii CONTENTS

Page

ACKNOWLEDGMENTS...... ±i

VITA, FIELDS OF S T U D T ...... iii

Chapter

I. INTRODUCTION ...... 1

II. STROPHIC STRUCTURE AND INTERVENTIONS BY THE POET . . 25

III. E P I T H E T S ...... 50

IV. PROPER NAMES ...... 92

V. DISCOURSE...... 106

VI. BATTLE SCENES...... 139

VII. OTHER ACTIONS...... l6l

VHI. CONCLUSION ...... 185

APPENDIX ...... 196

BIBLIOGRAPHY...... 200

iv CHAPTER I

INTRODUCTION

One hundred and twenty-seven years have elapsed since

Francisque Michel first published the text of the Chanson de

Roland, and yet it has only been within the last ten years that the formulaic character of the Old French epic has begun to come to light. For centuries popular epic was thought to have been composed in the same manner as the epics of Virgil, Goethe or Ronsard. It was not until 1927 that a young American scholar,

Milman Parry, while conducting research on the versification of the Homeric poems, discovered that these works were composed in a highly stylized idiom which greatly facilitated their reci­ tation before an audience. The Harvard scholar found that there was a striking economy in the choice of phrases in Homer: that the same expressions occurred time and time again under the same metrical circumstances, and that for each recurring set of semantic and metrical conditions there was usually only one phrase employed. Parry came to the conclusion that the diction of the THad and the Odyssey, with its curious mixture of dialect elements, was of a traditional nature, and that no one man could possibly have invented the hundreds of fixed phrases which are found in substantially the sane form many times over throughout the two poems. Virgil repeated 92 full verses in the Aaneid. a number ■which is dwarfed by the 1804- repeated verses of Homer* s two works.-*- Such an elaborate system of diction could only be the task of generations of poets each of whom contributed to the art and then passed it on to his successors. Parry assigned the name

"formula** to the stock phrases, and he defined the formula as

"a group of words which is regularly employed under the same metrical conditions to express a given essential idea" (p. 80),

Gradually, through the study of the contemporary Serbo- croatian epic, which shares many of the characteristics of diction of the Greek poems, he realized that the method of composition of the poets who had molded the Homeric language must have been one of oral improvisation like that of the Serboeroatian singers who put their poems together, verse by verse, at the very same time as they are performing them before an audience. This art of

spontaneous creation is only made possible through the use of for­ mulas, for the pace of oral performance is too rapid to allow a poet the time he would require to compose each verse extempora­ neously.

Parry and his disciple, Albert B, Lord, collected the epic

1 Milman Parry, "Studies in the %>ic Technique of Oral Verse- Making. I. Homer and Homeric Style," Harvard Studies in Classical Philology. XLI (1930), 13*K texts of Serboeroatian singers, 'who at the time were very active, and stored than on phonograph records in what was to become, after Parry's death, the Milman Parry Collection of South Slavic

Texts at Harvard University,2 They discovered that the formulas occurred within more or less fixed narrative patterns that they called "themes," which corresponded to the stock actions of the characters of the epic. To sing ah epic poem the author did not have to memorize it word for word, nor on the other hand did he create it in a purely original way each time he performed it.

The process was rather one of becoming well acquainted with the order of the themes of a given tale, usually by hearing it sung by another poet (the traditional poet is almost always illiterate) and then re-creating it by means of formulaic expressions but not necessarily in exactly the same words. Once a singer, in the course of his apprenticeship, had learned the formulas for such stock themes of the epic as holding a council, fighting a battle, sending a messenger from one camp to another, and so forth, he no longer had to woriy about reproducing the song exactly as he heard it but could inproviso at will, expanding some scenes,

%f&ry of these poems are edited and translated in Albert B, Lord, ed,, Serboeroatian Heroic Songs, collected by Milman Parry (Cambridge and Belgrade: Harvard University Press and the Serbian Academy of Sciences, 1953-5*0• Vol. I: Novi Pazar: English Trans­ lations. with musical transcriptions by Bela Bartok and prefaces by John H, Finley, Jr. and Roman Jakob son. Vol. H : Novi Pazar; Ser- bocroatian Texts, with introduction and notes by the editor and with a preface by A, Belie. shortening some, and adding or subtracting others.3

Although Parry's views were set forth in two Sorbonne theses,

they wore not immediately put to profit by scholars working in the

field of Old French literature.** Bedier had published his monu­ mental Legendes enidues between 1908 and 1913» and in the twenties his theory that the chansons de geste had originated from a close

collaboration between monks, who wished to attract pilgrims to

their shrines, and jongleurs, who were the means of propagandizing

this "tourist*1 trade, still prevailed among the vast majority of

critics. Between 1926 and 1928, however, Ferdinand Lot devoted

a series of articles in Romania to a critique of his friend

Bedier*s theories, pointing out that although a majority of the

chansons did indeed contain references to the monasteries which

dotted the pilgrimage routes and were undoubtedly closely linked

to them, the oldest of the poems, that is the Roland of the Oxford

manuscript and the Guillaume, reflected no knowledge of the routes

whatsoever.5 Seeing no other way to explain the presence, in

works of the eleventh and twelfth centuries, of historical per­

sonages from the eighth and ninth centuries, Lot advocated a

3a full treatment of formulaic language is found in Albert B. Lord, The Singer of Tales (Cambridge: Harvard University Press, I960).

^Milman Parry, L'Ebithete traditionnelle dans Homere and Les Formules et la Metrique d*Homere (both Paris: Societe Edition "Les Belles Lettres," 1928).

^Ferdinand Lot, Etudes sur les Legendes enidues franoaises. with an introduction by Robert Bossuat (Paris: Honore Champion, 1958)* return to the romantic theory of oral transmission from age to age which had been espoused by Gaston Paris and against which

Bedier had constructed his system.^

Kany other critics contributed to the attack against Bedier* s position in the years that followed Lot's articles. In 1933

Robert Fawtier, primarily a historian, also proposed a return to the theory of continuous transmission from century to century beginning shortly after the tine of the events recounted in the chansons de geste.? In 1946 and 1947 Rene Louis, in his three- volume work De l*Histoire & la Legende. held that the survival of the name of the hero Girart de Vienne in the epic can only be due to a continuous poetic tradition which must have begun in the g ninth century. Rita Lejeune, first in her book Recherches sur le

Thane I Les Chansons de Geste et l*Histoire (1948) and then in various articles, affirmed the historicity of the events which form the nucleus of the chansons de geste Basin et Basil a, Mainet.

Berthe aux grans pieds and Qgjer le Danois. discovered seven examples of the onomastic couple Roland-Olivier which antedate

^Gaston Paris, Histoire noetique de Charlemagne (Paris: E. Bouillon, 1865). 7 Robert Fawtier, La Chanson de Roland (Paris: E. Boccard, 1933).

8 , , . Rene Louis, De l*Histoire a la Legende: Girart. Ccante de Vienne et ses Fondations monastioues (Auxerre: Lnprimerie Moderns. 1946): De 1 'Histoire a la Legende: Girart. Comte de Vienne, dans les Chan­ sons de Geste Girart de Vienne. Girart de Fraite. Girart de Rous­ sillon (2 vols.; Auxerre: Lnprimerie Modems, 1947). the Oxford version of the Chanson de Boland, and suggested the possibility that the Old French epic originated in the South of

France.^ Jules Horrent, in La Chanson de Boland dans les fe­ te natures francaise et espagnole au mqyen age, examined the manu­ script tradition of the Chanson de Roland, concluding that it contains obvious signs of oral transmission and that the variants of the later versions are sometimes literarily superior to the early texts.-^ Studies by Martin de RLquer,^- Paul Aebischer,^

Dima so Alonso,^ Pierre Le Gentil^ and others contributed to the growing strength of the traditionalist position.

But it was not until 1955 that Parry’s theories were applied to the chansons de geste. Jean Rychner, in his work La Chanson de

Rita Lejeune, Recherch.es sur le Theme: Les Chansons de Geste et l ’Histoire (Paris: Societe d*Edition "Les Belles Lettres," 1948); "La Naissance du Couple litteraire ’Roland et Olivier,*" Annuaire de l*Institut da PHilol.o^ie et d’Histoire orientales. X (1950), 376-hOO; »«Pne Allusion meconnue a une Chanson de Roland.H Romania. LXXV (195*0, 145-6^; "Actualite de la Chanson de Roland.w La Nouvelle Clio. V H - I X (1955-57), 207-227.

Jules Horrent, La Chanson de Roland dans les fefeanatures fran­ caise et espagnole au moven age (Paris: Societe d*Edition ”Les Belles Lettres," 195l).

•^~Les Chansons de Gesfce francaises. second edition translated by Irenee Cluzel (Paris: Mizet, 1957).

• I A Rolandiana bore alia (Lausanne: F, Rouge, 195*0.

■^"La primitiva epiea francesa a la luz de una nota emilianense, •» Revista de Filologia Espanola. XXXVII (1953), 1-9*K 1h. La Chanson de Roland (Paris: Hatier-Boivin, 1955)• Geste: Essai sur j»Art epique des Jongleurs, found that the style of these compositions corresponded in many respects to that of the 15 Homeric poems as analyzed by Parry. There were the same kinds of

stock themes, the same foraulaic language, the same lapses of memory on the part of the poet that pointed to the oral origin of the Greek and Slavic literature."^ For Itychner the real chanson de geste existed only in actual performance; as for the texts which we possess,

••nous n»en tenons la qu*un reflet,*1 From what he was able to dis­ cover about the conditions under which the chansons de geste were

composed, episode by episode, with the .jongleur deciding whether to leave out or to include each succeeding adventure, fychner concludes that the works which are extant may be simply "le re suit at presque fortuit de la mise par ecrit d*un assemblage, entre dfautres pos­

sibles- (p. 155).

Scholars in the field of Old French literature have in general

accepted the analysis which Hychner made of the genre, although there was the inevitable reluctance to accept the notion that oral compo­

sition could produce masterpieces like those of the great writers such

as Hacine or Flaubert. %chner»s principal opponents have been the

neo-bedieristes who believe that each poem is the result of the

1C ^Jean Rychner, La Chanson de Geste: Essai sur l»Art epique des .Tnw^lmrs (Geneva: Droz, 1955)*

^Ifcrchner, p. 137f cites an example from La Prise d*Orange in which the jongleur mechanically inserted a motif in the wrong place: Guil­ laume and Guielin fight the Saracens in the stairwell of a castle, but only after the poet has aimed them with lance and gonfanon as if they were on horseback. A literary poet would have seen his error upon re-reading and would have corrected the text. moment sacre of the individual artist's personal creativitySo great was the interest generated by fychner's work that most of the papers read at the colloquy on the literary technique of the chansons de geste held at Liege in 1957 either took his theories as their starting point or else attacked him vigorously.

Ifyohner was not, however, alone in seeing the Old French epic m as traditional oral poetry. Ramon Menendez Pidal published, in

1959, a book which summarized his own and other scholars* views on the origin of the chansons de geste and placed him at last, in the eyes of French scholars, in the position of principal exponent of the traditionalist theories: La Chanson de Roland v el neo- 18 tradicionaH .«ann (OriCgenes de la Epica rom&iioa). Despite the

efforts of Bedier* s partisans to discredit Menendez Pidal* s con­

clusions,'1'^ it appears that individualism in the field of epic

criticism is on the wane. In this study the traditionalist hypothe-

■^An example is Maurice Delbouille in "Les Chansons de Geste et le Livre," La Techr^qn** -Htteraire des Chansons de Geste: Actes du Col­ logue de Liege (Septanbre 1957) (Paris: Societe d*Edition **Les Belles Lettres,** 1959), pp. 295-^7. 18 Madrid: Espasa-Calpe, 1959. This work was translated into French by Irenes Cluzel, with revisions by the author: La Chanson de Roland et la Tradition enioue des Francs (Paris: A. et J. Picard et Cie., 19S0T. *^See, for example, Andre Burger, **La Question rolandienne, Faits et Hypotheses, *• Cahiers de Civilisation Medievale. IV (1961), 269- 91. sis is accepted as the most accurate account of the origin and transmission of the chansons de geste.

In one of the studies read at the Colloque de Liege, "Sur 20 l'Originalite de Raoul de Cambrai," Ctaer Jodogne attempts to

discern the relationship of the various epic traditions as revealed by an examination of ten of the most commonly found formulas in

the works of the twelfth century. He notes:

Le jeu dn hasard me paralt axclu ici et la pre­ sence slmultanee de plusieurs cliches dans des chan­ sons differentes de sujet et d'esprit prouve une parents de lenir existence. XL suffit qu'un des cliches se rencontre une seule fois dans une ahanson: l'eoproht est prouve et l'accointance des auteurs est des lors attestee. (p. ^7)

He concludes that, because they share many of the same "cliches

epiques," La Chevalerie Ogier. Aliscans and Raoul de Cambrai are

related, and that the songs of the cycle of the rebellious barons

are descended from the cycle of William of Orange (p. 51)*

This conclusion seems excessive, since for two songs to share

the same formulas all that would be necessary is that the .jongleur

■who sang one of them should have heard the songs of other poets

•who were, more or leiss remotely, cpnnected with the .jongleur of

the second work. Lord recommends that one begin the study of the

living epic with the works of one poet, and then go on to study

those of the singers who lived around him and who have influenced

20 La Technique litteraire.... pp. 37-58* 10 him, and then, in ever-widening circles, those of the other singers of the language area.^- It should be obvious that a formula could be transmitted from the singer -Who invented it, by the intermediary of others, to a singer with whom he has never come directly in contact. Thus the only result of a survey, such as that contained in Jodogne»s article, of scattered formulas in ten chansons de geste is to prove that the poems were sung by poets who shared the same tradition.

It is precisely because incomplete samplings of formulas from different works will not produce solid proofs of connection or in­ fluence that this study undertaken. Duncan McMillan, reviewing the Actes du Colloque de T.iago- says of the study °f formulas.

"Bien que sterile sous la forme embryonnaire pronee par M. Jodogne, qui pourrait dormer tout au plus des resultats negatifs..., la methode seralt peut-etre fructueuse sur une echelle bien plus com- plexe."22 In order for the scholar to be able to say with assurance that a given author was either original or perfunctory in the manner

in -idiich he used the formulas of the tradition, one would first have to be in possession of a fairly complete repertory of the works

which are contemporary with the poem in question.

However, to know how to recognize a formula one would have to

begin with one chanson de geste and read it over and over to find

which phrases were constantly repeated. In other words, an analysis

Singer of Tales, p. 49.

^ Cahiers de Civilisation Msdievale. IV (196l)» 91, note. 11

of the tradition must begin with the denouillement of the formulas

of a single work, and then be extended to others until it encompasses the whole corpus of the genre, A formula might appear only once in

some works, and it would not be recognized as belonging to the

repertory of fixed phrases until it was seen in several poems. Thus

it would be necessary to return once again to the poems with which one began and to re-read them in the light of the knowledge then acquired,

* * * * *

The process of analyzing the formulas of a chanson de geste. then, is difficult in the extreme. Apropos of the Homeric style

Parry wrote*

A full description of the technique is not to be thought of, since its complexity , , . is altogether too great. One must either limit oneself to a certain category of formulas, and describe their more frequent uses, , , , or one must take a certain number of for­ mulas of different sorts which can be considered typ­ ical, 23

TOLth either procedure, however, the results would be incomplete,

and so it seemed necessary to develop a new method of research which could properly cope with the complexity of the problem posed by the analysis of oral poetry.

This task was completed through the use of the IBM 7092

computer which was made available by the Numerical Computation

23"StuxHes the Epic Technique , , , ," p. 126, Laboratory (now the Computer Center Data Processing System) of the

Ohio State University. The verses of two Old French poems, Le

Siege de Barbastre and Buevon de Conmarchis. were punched out on

IBM cards and fed into the core of the computer. Since the spelling of Old French texts is not standardized, it was not possible simply to program the computer to compare each verse, letter for letter, with all the others, for the discrepancy of a single letter would cause the two verses to appear to the computer as totally unrelated.

For this reason each verse was reduced to its basic grammatical and lexical components, and a card bearing a codified version of these components was associated with each verse-card. The code words and their meanings are as follows: nouns:

KAML animal PPSN person (proper noun) NBDG edifice NPLC place (common noun) NBDT body or part of body PPLC place (proper noun) NEAT food, meal NRGN religious act or. object NGEP group, array, battle NEMR armor, clothing, furniture NHLP help, rescue NSAY speech, sound, word. NKJLL death NSHP ship or part of ship NKNW thought, judgment NTIM division of time NMHL moral quality NOUN any noun not included in NNTR natural object the above categories NPSN person (common noun) verbs:

VAWT to be obliged, ought VGO to go VBE to be VHAV to have VBRG to bring, guide, protect VHER to hear VBEK to break VHIT to strike, wound VCAN to be able, can VHLP to help, rescue VEAT to eat VEIL to kill VFER to fear, be angry VKNW to know, think, judge VFND to find VLIV to live, be born VGIV to give VLOV to love 13 VMAK to make VSPR to spur VPUT to put VTAK to take VRGN to believe (or any act VTHW to throw of religion) VWSH to wish VSAY to say VERB any verb not included in VSEE to see the above oategories VSHP to sail others:

ADJ adjective ANBR number AVB adverb

The computer was programmed in such a way that all hemistichs whose grammatical and lexical patterns were the same would be printed out as a group. Each of these groups was arranged in alphabetical order so that identical hemistichs would appear either in juxtaposition on the printed page or else in close enough proximity as to be immediate­ ly recognizable. It was thus possible to tell at a glance whether or not a hemistich was formulaic. Those hemistichs whose patterns were unique, that is which could not be formulaic, were printed out at the end of the data. Most of the body of this study consists of 2k a description of the results of the computer program.

It is therefore the aim of this study to describe all the formulas of the two Old French poems Le Siege de Barbastre and

Buevon de Conmarchis. Since the oral poet is presented at each moment of his recitation with the possibility of using a formula which he has learned from the poetic tradition which preceded him

2k The computer program, in the SCATRAN language (a dialect of IBM,s FORTRAN language), is found in the Appendix, written out just as it appeared on the output sheets. l*f or of fabricating a new verse to fit the unique needs of the nar­ rative, the study is also an exploration of the possibilities of formulaic composition within the framework of a certain plot, for the two works narrate the same events. It thus becomes possible to compare the treatments of similar situations in the two chansons de geste and ultimately to deteimine whether there is a major stylistic difference between them.

* * * * *

Le Siege de Barbastre has been dated in the last third of the twelfth centuiy. Its 7392 verses are divided into 203 laisses of riming alexandrines, and each laisse ends with a non-riming hemi- 25 stich or petit vers. Its author is unknown, and it is composed in a literary.language in common use for narrative verse at this period which reveals nothing about the geographical origin of its author.

Buevon de Conmarchis. by Ad©net le Roi, a poet at the court of

Brabant, dates from around 1280. It is thus separated in time from

Le Siege de Barba stre by approximately 100 years. Its 132 laisses

25 Toward the middle of the poem (laisses 7k to 90) many verses are assonanced. They may be the remains of an earlier, entirely assonanced version. The petit vers is found in many of the poems which make up the cycle of 'William of Orange. 26 J.-L. Perrier, ed., Le Siege de Barba stre (Paris: Librairie Honore Champion, 1926), p. iii. All references are to this edition. 15 contain 3947 rimed alexandrines, and as In the earlier poem each 27 laisse is terminated by a petit vers.

The plot of Le Siege de Barba stre is as follows. Aymeri de

Narbonne and his wife and sons are celebrating the feast of Eente- cost in the castle of Narbonne when a messenger arrives and tells them of the approach of hostile Saracen forces, led by the a-wH rant of Spain and the aamstant of CordLova, who are determined to capture the city. After a brief skirmish during which Buervon de Conmarchis, son of Aymeri, and his offspring Girart and Guielin are taken pris­ oners along with several other French knights, the defending forces retreat to their fortifications and the Saracens lay siege to the town. Buevon and the other captives are led to Barbastre, a Spanish stronghold ruled by the amirant but rightfully belonging to the disinherited pagan Clarion de Vaudune. The French are thrown into a dungeon and beaten by their captors, but that evening Clarion, feeling his resentment toward the amirant and guided by God, frees

Buevon and his men. In the subsequent battle for control of the castle Buevon kills the amoraive. a lieutenant who was left in charge of Barbastre by the amirant. and all the Saracens surrender, submitting themselves to mass baptism. In the meantime Corsout de

Tabarie, a Saracen hero, manages to escape the conquerors and rides to Narbonne to tell the news to the army. The amirant raises the

27 ' All citations from Buevon de Conmarchis will be taken from the edition of Albert Henry, Les OEuvres d*Adenet le Boi. Vol. U s Buevon de Conmarchis (Bruges: "De Tenrpel," 1953)* siege and leads his men to Barbastre where they meet Beuvon and his followers in a battle before the walls of the city. The French finally have to retreat and the amirant lays siege to Barbastre.

At this time the amustant. in search of reinforcements, sends a messenger to Malatrie, his daughter, at Cordova, telling her to lead an army to his aid. She hears from the messenger about the bravery of Girart and falls in love with, him before ever having seen him. She arrives at Barbastre with her troops, and the amirant intro­ duces her to his son, Libanor, to wham she is to be wed. The next day Malatrie rides past Libanor*s tent on her way to take some sun­ shine and shyly induces him to accompany her, hoping that Girart will see them and come out from the castle to fight Libanor. This occurs, as she wishes, and Libanor is disgracefully defeated. Mala­ trie and Girart are about to flee to the protection of the castle together when the other Saracens arrive to rescue Libanor. Guielin has to come to his brother’s assistance, and the French retreat without the girl. Malatrie is scolded by her father for her friend­ liness toward Girart (as Girart is rebuked by Buevon for failing to bring her back with him), but upon returning to her tent she arranges to have one of her men, Malaquin, go to Girart and invite him to a rendezvous for that evening. Girart and several other knights go secretly to the women’s quarters of the Saracen camp where they spend a few hours with Malatrie and her ladies, but as dawn approaches the pagan sentinels discover Guielin and his knights who are protecting

Girart’s path, of retreat. After a brief encounter the amirant hears 17 of the French incursion and leads his men against them, A battle begins.

The thirteenth-centuxy Buevon de Conmarchis follows the plot

of Le Siege de Barbastre up to this point, and then its unique 28 manuscript conies to an abrupt end. In the older poem the French

retreat to Barbastre with Malatrie after inflicting manor losses on

the enemy. Clarion goes as a messenger to Naibonne and then to

Paris, and William of Orange and King Louis eventually lead an

enormous army to Barbastre where, after many battles and intrigues,

they defeat the Saracens. Girart and Malatrie marry and Le Siege

de Barbastre comes to a close.

The point at which Buevon de Conmarchis ends corresponds to

a place around the end of laisse 86 of Le Siege de Barbastre.This

laisse ends with verse 2696, and it is the section before verse 2696

which will be compared to Buevon de Conmarchis.

Menet does not pretend to have invented the story he is telling:

he freely admits that he took the maters of his work from a previous

poem.

Tout droit en cel teimine dont je parle vous ai, A Saint Denis en France la droite voie alai. A un moine courtois sagement m*acointai, Dans Nicholas ot non, car je li demandai; D'estoires ancleimes enquis et encerchai. De la geste Aymeri quis tant et reverchai Que je la vraie estoire avoec moi enportai,

28 Albert Henry, Adenet's editor, believes that he did complete the work but that an accident in the transmission of the text has deprived us of the ending. Les OEuvreg d*Adenet le Roi. Vol. I: La Tradition manuscrlte. Biographic d»Adenet (Bruges: "De Tempol," 1951), P. 174. 18

Hors d*un moult tres biau llvre la mat ere en puisal; Maint autre grant estolre avoec 1 esgardai, Hals pour la plus eerbaine a cestl m ’apoiai. Pour ce qu’est mal rimee, la rime amenderai SI a droit que 1*estolre de rlens ne fausserai, Mengonge ne oiseuse ja n'l ajoustera! Mais parml la mat ere droite vole en irai. (11-24)

The claim that the original poem -was found at Saint-Denis is not to be taken seriously. Statements of this sort are simply conventional topoi of the epic genre and are meant primarily to add authority to 29 the narrative. As for the rime, Adenet did "amend1* it through the technical feat of having every laisse in a masculine rime fol­ lowed by a laisse in the corresponding feminine rime (rime deriva­ tive).

Adenet does not introduce any new episodes into the plot as he found it in his model, but he does make some changes. Erich

Roll, in a dissertation entitled Untorsuchungen uber das Verhaltnis des Siege de Barbastre zum Bueves de Conmarchis von Adenet le Roi. und 30 die Stellung der Prosafassung. discusses these modifications and concludes that they can all be traced to the courtly environment of

Adenet. For example when Girart defeats Libanor in single combat he unhorses him and the Saracen hero falls into a nearby pond. Adenet has Girart help his adversary out of the water, whereas in the earlier work the poet is concerned only with the capture of the defeated man’s horse and lady. Later when Girart visits the tent of Malatrie, Adenet

29 a See L.F. Flutre, "Etudes sur le Roman de Perceforet." Romania. i m (1950), 374.

•^Greifswald: Buchdruckerei Hans Adler, 1909- has Malaquin precede him there, while the anonymous poet allows

Girart to enter without an intermediary.^ These and other changes

in the narrative are slight, and Boll is probably correct in his

assessment of their cause. He posits three divisions of Buevon

He Cermarehis according to the degree to which Adenet alters the

details of the story. From the beginning to laisse 46 there are

few changes; from laisse 4? to laisse 77 there are many differences;

from laisse 78 to the end Adenet again follows his model fairly

closely. From this Boll concludes that at the beginning and at the

end of his task Adenet had before him a copy of the original poem,^

but that for the middle section there must have been a considerable

time lag between the reading of the original and the act of com­

position. Albert Henry takes the more plausible position that this

hypothesis is not needed to explain the varying degree of imitation.

He sees, in spite of many almost word for word repetitions in the

last part, an effort by Adenet, progressively more successful, to

liberate himself from the mere slavish paraphrase of the older poem.33

* * * * *

Since the purpose of this study is to compare the formulas of

the two works, one must first arrive at an adequate idea of just

31 ^ A succinct account of the differences between the plots of the two works is found in Henry, Buevon de Conmarchis. pp. 17-35#

version used by Adenet was probably very similar to that of ms, C, upon which the mo d e m edition of Le Siege de Barbastre is based.

^%uevon de Conmarchis. pp. 32-3^# 20 what constitutes a formula. Parry, it will be recalled, defines the formula as "a group of words which is regularly employed under the same metrical conditions to egress a given essential idea."

He goes on to say:

The essential part of the idea is that which remains after one has counted out everything in the expression which is purely for the sake of style. . . . The word group is employed regularly when the poet uses it without second thought as the natural means of getting hiis idea into verse. The definition thus implies the metrical usefulness of the formula.-^

If a phrase is repeated, but not because it is useful in telling the story or in fulfilling the conditions of the versification,

then it is not a formula. Thus literary references, such as, to use an obvious example from contemporary literature, T.S. Eliot's use of verses from Baudelaire, Shakespeare, Spenser, Marvell and

others in The Waste Land, are not formulas. As for what Parry

refers to as the "essential part of the idea” which the formula

conveys, this is difficult to define. In this study two phrases will be considered formulaic if the only difference between them

is the substitution of one preposition, conjunction, interjection,

pronoun or adverb for another. If a noun, adjective or verb is

changed, however, the "essential idea" can hardly remain the same

unless the substitute word is synonymous with the word it replaces.

This will usually occur only in cases where the last word of a b

hemistich is replaced by a synonym because of the needs of rime.

Therefore a phrase will be considered formulaic if it is found two

Oh "^"Studies in the E^pic Technique . . . ," p. 80. 21 or more times in the same poem in substantially the same form, that is without any changes in nouns, verbs or adjectives unless it be for the rime alone.

It sometimes occurs that the context will require a word to be replaced by another one which has a different syllabic value.

In this case there must be another change made in the hemistich, and this is done by the alteration- of some other word or phrase without any modification in the essential unity of the formula.

Let us take as an example the phrase Es vous un Sarrazin. found twice in Buevon de Conmarchis (2067a, 2311a). "When the poet wishes to point out more than one Saracen (Sarrazins without any article in

Old French) he must modify the expression es vous to a_tant_es,

recovering the lost syllable which was taken up by the article in

the original hemistich. The new phrase is A tant es Sarrazins (3913a).

A tant es and es vous do not differ in meaning, but they do provide

the poet with a choice between two different syllabic combinations.

This process I have called "metrical compensation" because a compen­

sating change is made by the poet so that the hemistich may be kept

in the right meter. Phrases which undergo such changes will be con­

sidered variant forms of the same formula.

Phrases which are found only once in each poem may merely be

evidence of borrowing on the part of Adenet, and they cannot be con­

sidered formulas.

In deciding on the formulaic nature of certain phrases I have

been, forced to make some judgments which are perhaps arbitrary. But

the judgments had to be made, and I have endeavored to apply my 22 criteria to the two poems in the same way.

Of the three elements of which any formula is composed, meter, syntax and mean^i^, it is the last which is the most important. For the purpose of the ^jongleur is, after all, to tell a story, and all else depends; on this ^ Expressions which are syntactically alike but \ which differ in meaning are counted by Parry and Lord as playing a very important role in formulaic composition, parry called them

"formulaic expressions" (in the sense of "expressions similar to formulas"). The phrases Ring sessi le cheval and Einz sessi la contesse would be considered by Parry to belong to the same "for­ mulaic" pattern. But so would Bt baisa la contesse. It is difficult, however, to see any meaningful relationship between Bt baisa la con­ tesse and EH w e sessi le cheval. How far could one deviate from the original phrase while still remaining in the same pattern? How do the syntactic patterns of oral poetry differ from those of ordinary speech? Parry and Lord do not answer these questions, and until more work has been done on syntactic patterns in the epic, they will remain, it seems, rather unconvincing evidence for proving that a poem is formulaic. Undoubtedly most of the hemistichs of a poem will fall into certain syntactic patterns, but this should be equally true of oraj. and non-oral verse. Vhile syntax and meter may be factors in determining just what form a poet will adopt for his idea, the thought itself is the most important feature,

35 'Jackson J. Campbell in "Oral poetry in The Seafarer." Speculum. XXXV (I960), 90n, rightly points out in regard to syntactic patterns that "such adjustable 'systems* . . . certainly existed, . . . but besides those which are clear-cut and self-evident, there are those whose ambiguity makes them bad evidence." 23 "What is important in this study is not so much how many formulas are found (although this is certainly of interest), but the rela­ tive amount of each poem that is formulaic.

The results of each chapter will be summarized briefly and a final conclusion will be drawn, at the end of the dissertation. These results will be relative, it is true, for no other data are available for the other chansons de geste. this being the first study of its kind. There have been other works comparing the formulas of several poems, 36 but no effort has been made to find out exactly what amount of a given Old French poem is formulaic. Since the subject of a poem determines in great part what formulas will appear and which ones the poet will have no need of, it is necessary to begin the depouil- lement and comparison of the chansons de geste with two works such as the ones dealt with here which have the same subject.

The classification of formulas which is necessary for this study has been drawn up with special regard to the usefulness of the for­ mulas to the poet. It is evident that certain sets of formulas are called upon again and again to fill ever-recurring needs in the

Eight theses of this type were done for the degree of licencie es lettres at the University of Louvain in i960, all under the direc­ tion of Professor Ctaer Jodogne. They are found in the Bibliography under the names of their authors: C. De Cock, J.-P. Desert, S. Ros- meulen, M. SLootmaekers, A.M. Van de Walle, H. Troeh, F. Laaibotte, and C. Lejeune. One concerns the earlier of the two poems treated here: M.-C. Fripiat, "Les Motifs et les Formulas dans quatre Chan­ sons de Geste: All scan s. la deuxieme Redaction en Vers du Moniage Guillaume. La Mort Aymerl de Narborme et Le Siege de Barba sire." This work concludes that there are certain similarities of expres­ sion in the formulas of the four poems, but it does not treat all the formulas of any one poem and so is of little use for this disser­ tation. 2b process of story-telling. These needs are of several types: the conditions which lead a poet to use the formula of discourse Bele. dit l«amirant (SB; 1709a) are not the same as those which give rise to the epithet Qui tant fet a prisier (SB: 1893b). The former is needed to further the action, to introduce a sequence of conversa­ tional statements which will contain part of the story-line of the poem. The latter expression tells us something about a particular hero, but it does not advance the story in any way.

Among the formulas which are, one might say, "conveyers" of the narrative, those which are involved in the presentation of dis­ course form an important group. Next in size is the block of phrases which the poet uses to describe the actual fighting which is the main subject of any epic poem. After these come formulas which are attached to the less commonly found themes, such as the arrival and departure of messengers, the testing of new horses, or, in the inanimate sphere of events, the transitions from day to night and from night to dawn.

The "non-conveyers'* include formulas which are made up primarily of the proper names of characters and those which are purely orna­ mental (epithets). Also under this general heading one would have to include those by which the poet inserts his own personality into the fabric of the narrative by pointing out things to the reader’s

■attention or by making comments on the happenings.

Each of the following chapters deals with one of these types of formula, or with a sub-division within one type. CHAPTER H

STRQPHXC STRUCTURE AND INTERVENTIONS

BY THE POET

In this chapter will be considered those formulaic elements which are found in the introductions to Le Siege de Barbastre and

Buevon de Conmarchis and those ■which are specifically related to the strophic structure of the two poems, as well as explicit signs of transition from one episode to another, and phrases which may be taken as evidence of oral presentation.

Introductions

Le Siege de Barbastre begins in this manners

Plest vos o£r chanson bien faite et conpasee? Toute est de vielle estoire estraite et porpansee. Molt fait bien a o£r, piega ne fu contee; Tote est de la lingniee qua Deus ot tant amee, De la geste Aymeri qui prouece a duree. 2. Ce fu a Pantecoste, cele feste ennoree. . . . (SB: 1-6)

The first laisse goes on to tell of the feast at Aymeri's court at

Pentecost, and then these verses take up again the announcement of

SB and BC are, of course, abbreviations for Le Siege de Bar­ bastre and Buevon de Conmarchis. Throughout this study the letters a or b after a line number indicate that the reference is, respec­ tively, to the first or to the second hemistich; p designates a petit vers.

25 26 of the beginning}

Ci conmence ohanyon de bien enluminee, D*amor et de bataille et de bien avivee; De grant chevalerie s*ele vos est contee, Et il soit qui la die et el soit esooutee; Ne euit qu*ele vos faille de ci a l'avespree; Estrete est del parage. (SB: 41-46)

Later, after the capture of Buevon and his sons by the pagan invaders, Corsout de Tabarie takes the prisoners to Barbastre where, we are told, they are cruelly beaten.

N'i a paien ne tiengne ileques sa corgie; Dieusl con ill ont no gent batue et ledengie. Or conmence chanyon a merveille enforcie, Einsi conme Aymeris secorut sa mesnie, Et Looys li rois an ot France voidie, Juque au perron seint Jasque fu la terre essillie, Einz que venist en France. (SB: 494-500)

These verses may be evidence of the division of Le Siege de Bar- bastre into songs of lengths suitable for recitation in one evening.

M. Jean ftychner estimates that a .jongleur could sing from 1000 to

2000 verses during an evening•s seance, but of course this number 2 could vary greatly according to the circumstances of the performance.

This passage, which states that the "song" is beginning whereas actually it is 18 laisses long at this point, may constitute the first few verses of the second seance of the poem. Five hundred verses do not seem to be an adequate amount for one seance, but it is interesting to note that the only other passages of this type which appear in the work (Or conmence chancon. s*il est qui la vos die. SB: 5^92, and Ci comence chancon de bien onluminee. SB: 6105)

occur a little over 400 verses apart. This would suggest that the

La Chanson de Geste (Geneva: Droz, 1955), p. 49. 27 length of seances for the .jongleur of this work may well have been in the range of from 400 to 500 verses. Even if many of the

seances ran on for more than 500 verses, the first performance could well represent a sort of advance demonstration for the audience on 3 the part of the jongleur.

The verses which begin Le Siege de Barbastre and those which

signal the actual start of the narrative and the relancement of the song after a pause are themselves formulaic. It is, of course,

impossible to test the formulaic nature of the opening verse of a

chanson de geste by comparing it with the rest of the verses of the

same poem, for the poet will never be in the same circumstances

of having to introduce himself to the audience again. One can, however, compare these verses with the initial passages of other

epic poems in the same tradition.

3 The author of Huon de Bordeaux interrupts his song at one point with the following laisse: Segnor preudomme, certes, bien le vees, Pres est de vespre, et je sui moult lasse. Or vous proi tous, si cier com vos m*aves Ni Auberon ne Huon le membre, Vous revenes domain apres disner; Et s'alons boire, car je l»ai desire. Je ne puis, certes, mon coraige celer Que jou ne die $ou que j*ai empense: Moult sui joians quant je voi avesprer, Car je desire que je m'en pulse aler. Si revenes domain apres disner, Et si vous proi cascuns m*ait aporte U pan de sa chemise tune maille noue, Car en ces poitevines a poi de largete; Avers fu et escars qui les fist estorer Ne qui ains les donna a cortois menestrel.(4976-4991) If the story is an exiting one and the jongleur skilled at singing it, his listeners will return the next day with their coins and the song will be taken up at the point at which it was interrupted. 28

Plest vos o£r chanson bien faite et conpasee? SBs 1 Plaist vos o£r de granz batailles et de forz esturs? Chanson f^miaune: 1 Plaist vos ofr d'une estoire vaillant? Couronnetnent de Louis; 2 Bone chanjon plest vos a escouter? Charroi de Mimes: 3 Can chon de fiere estoire plairoit vos a o3n? Aiolt 5~ Plaist vos o£r bone eanson valiant? Chanson d*Aspremont: 1 Bone chanjon plest vos qne je vos die? Girart de Vienne i 1 Plairoit il vous o£r .III. vers de baronie? TCHa Saint-Gille: 3 Boine canchon plairoit vous a oir? Moniage Guillaume II: 1 Chanson de geste plaroit vous a entandre? Enfance« nmiianmei 1 Plaist vous o£r, bonne gent houneree, Bonne canchon de bien enluminee? Bueve de Hantone H : 2 Plaist vos olr canchon de grant baraage? Chevalerie d*Qgier de Danemarches 3

None of these verses is exactly the same as the initial verse of

Le Siege de Baibastre. but the idea of beginning a song by asking the audience, rhetorically, if it would like to hear a good story is widespread. The verse which contains this invitation usually has plest vos plus the verb o3!r (or entendre), and the phrase canchon de bien erTum-inee from Bueves de Hantone is also to be found in SB; hi.

Adenet le Roi does not open his work in this manner, but rather employs a nature description for his initial verses.

Ce fu ou tans d*este, si comme ou mois de mai, K'en maint lieu resplendissent cler dou soleil li rai Et que arbre florissent et pres sont vert et gai, Lors m*est pris volentes que tous jours maintenrai — De cele volente ja ne me partirai, Se Dieu plaist et ses sains, tant com je viverai — Ce est que des preudomes volentiers parlerai. . . . (BC: 1-7)

One of his other chansons de geste. Berte aus grans pies, opens in like fashions

A.l'issue d'avrill, un tans dou$ et joli, Que herbeletes pongnent et pre sont raverdi Et aibrissel desirent qu*il fussent parflori. . . . (1-3)

Apart from these two works of Adenet, there is no other chanson de geste. from any period of Old French literature, which begins in this way. It is a common practice to open lyric poems with an evocation of nature, and Adenet may be borrowing this device from the lyric genre. In this respect he seems to be closer to the courtly culture to which such lyric conventions belong than he does to the relatively unlearned culture of the singers of chansons de geste.

Strophic Structure

The laisses of Le Siege de Barbastre vaiy in size from 7 verses (laisse 5 8 ) to 117 (laisse 6 9 ), the average being about 31- verses per laisse. The division of Buevon de Conmarchis into laisses is not as varied! it ranges from 12 verses (laisses 13 and 108) to

75 (laisse 1 0 0 ), but averages out to 30 P©** laisse, or almost the same as the earlier poem. The size of a laisse depends in part on the rime, as can be seen from the fact that the laisses which are the longest in each poem are composed in relatively easy rimes, laisse

69 of the anonymous work being in -ier and laisse 100 of Adenet*s poem in -er. On the other hand the small laisses are generally composed in difficult rimes! laisse 13 of Buevon de Conmarchis is in

-oing. laisse 58 of Le Siege de Barbastre in -in. That Adenet is a virtuoso at the art of riming becomes evident when one considers that he set for himself the task of composing his poem in rime derivative. And yet in Le Siege de Barbastre. where there are no

h. , See Manfreid Gsteiger, "Note sur les Preambules des Chansons de Geste," Cahiers de Civilisation Medievale. H (1959), 219. 30 challenging criteria apparent, there are eight laisses ■which are shorter than the shortest of Adenet*s (nos. 19, 22, 2 3 , 27, 32, 58,

83 and 8 6 ).

Formulaic expressions appear among the initial verses of laisses and also among the half-verses 'which end laisses*

Most laisses in both poems open with a mention of the place where the action occurs, with a general evocative verse describing the setting, or with the proper name or designation of a character.

The last of these is the most frequent practice, for it is important for the poet to situate the action clearly by placing a character at the beginning of each new division of his work.

Nature appears in some of the descriptive verses opening

Adenet*s laisses;

Ce fu ou tans d*este, si comme ou mois de mai. BC: 1 Eh este quant li jour sont bel et lone et cler. BC: 53 Droit en ou point d*este que pre sont plus herbu. BC: 1533 Li jours fu biaus et clers, sans vent et sans pound.ere. BC; 1783 Pres de l*aigue de Sore, en un molt tres biau pre. BC: 2052 Moult fu cele ajoumee clere et plaisans et bele. BC: 37^1

All but one of these verses have to do with the time of the year.

Natural phenomena also make their appearance in the initial verses of three laisses of Le Siege de Barbastre. but there they are couched in a formula;

Es prez desoz Nerbone sont Franc a la quintainne. SB: 67 Es prez desoz Nerbone sont Franyois el sablon. SB: 105 Es prez desoz Nerbone es vos paiens logiez. SB: 3^+8

This setting is more a bit of rigid scenery in the background than an evocation of nature, as will be seen when it is compared with those of Adenet. The formula of the older poem does occur once in 31 Buevon de Conmarchis i Ea pres devant Nerbonne sont Franc a la quin- talne. BCt 134, But this verse is found in the same position in the narrative as verse 67 of SB and is probably a borrowing.

Among the initial verses which situate the action that is about to be recounted in the body of the laisse one set of for­ mulas stands out in Le Siege de Barbastre1

Franjois sont en lachartre, corrode et destroiz. SBi 501 Francois sont en lachartre, corrooie et destroit. SBl 524 Paien sont en la sale, lie et joieus et baut. SBi 538 Francois sont en lachartre, dolant et corroceus. SBi 547 Franjois sont en lachartre, corrode et dolant. SB* 554

The first two and the last two verses of this series are obviously similar, and their first hemistichs are identical to the letter. The middle verse of the set, while it is similar in syntax to the others, provides a contrast between the freedom and good spirits of the Sara­ cen captors and the desperate straits of the Christian prisoners. The verses occur at the beginning of laisses 19f 21, 22, 23 and 24. Laisses

20 and 21 are parallel in part, as they both tell of Corsolt's order that the captives be placed in the tower of Barbastre, but laisse 21 ends with the appearance of Belinais, a terrible serpent, who has been placed in the chartre to harry the French. Laisses 22 and 23 repeat the same material, with a little more embellishment on the fearsome qualities of the serpent, and laisse 24 tells of Buevon*s unsuccess­ ful attempt to kill the beast. The French climb onto a ledge to

escape death, and they remain there until Clarion comes to their aid in laisse 26, The parallel verses at the beginning of the five laisses

serve to dwell upon the plight of the French heroes and to set the

stage for their rescue. 32

The first five laisses of Le Siege de Barbastre are also laisses In the first laisse, after the feast, Aymeri and his court go out onto the plain before the city where the knights amuse themselves at Jousting.

Mes, ainz que il soit vespre ne la nuit oscuree, Li plus hardis des lor avra mestier d'espee, Que 1* ami rant d»Espangne a sa gent aunee. (SB: 28-30)

The second laisse again begins with the account of the court and the feast, the jousting and the games, and then the ominous note is heard again.

Mes ainjois qu*il soit vespre avront il a plorer, Que l*amirant d*Espangne fait ses oz ajoster, Et l,amustant de Cordres, et par terre et par mer. (SB: 60-62)

In the next laisse the court which Aymeri held inside the castle is not mentioned, but the jousting (Es prez desoz Nerbone sont Franc a la qulntaine. SB: 13*0 a**d the courtly gathering of knights and ladies is described, only to be interrupted for the third time by the same refrain.

Mes ja ne verront vespre, ne la nuit prernerainne Que grant poor avront nostre gent segureinne, Que l'amirant d'Espangne, le fiz a 1 *amulaine, Fu entrez en Gascongne, bien a une semaine. (SB: 84-87)

Laisse 4 opens with a verse which is similar to the one found at the beginning of the previous one: Es prez desoz Nerbone sont Francois el sablon (SB: 105). There is mention of the joy of the revelers, but this is brought to an end by the arrival of the messenger.

Mes ainz que soient vespre, changeront lor reson. Es vos un chevalier brochant a esperon, Et fu el cors navrez, si se tint a 1*arson. (SB: 124-26)

He delivers his message in laisse 5, a*kl 'the poet reflects on 33 the fate of the Christian knights and ladies.

Mes ainz qu'il voient vespre^ ne soleil esoonsant, Li plus hardiz des nos n'avna talant qu'il chant, Que 1* amir ant d'Espangne apela Madiant. . . . (SB: 157-59)

The laisse ends with the two forces joining combat. The poet has

skilfully interspersed descriptions of the tranquillity of Narbonne in a time of peace, and the account of the messenger's arrival, with anticipations of the troubled events which will be the main subject of the work. It is thus a device for creating suspense through which the audience is made to look forward continually to the advent of the pagan airoy.

I have dwelt on these details from the opening laisses of Le

Siege de Barbastre both because they provide a good example of the

effective use of formulaic expressions to create an atmosphere that

is dramatic in its forebodings and almost lyrical in its refrains, and

also because, unfortunately, this is the artistic high-point of a poem that becomes progressively less interesting.

Most of the verses which begin the laisses in the two poems bring into focus the actions of one of the characters or else show

him or her reacting in some way to the action of another which closed

the previous laisse. For bringing characters into the scene several

methods are available. The poet may say that a person is arriving

(SB: laisses 45, 65, 68 and 78), or leaving on a trip (SB: 18 and

74; BC: 4, 5 and 8 6 ), or leaving one character to go somewhere else

in the vicinity (SB: 28, 30 and 42), or riding from one locality to

-*See too Einz qu'il soit avespre. SB: 57*&>, 1358b• 34 another (BC: 67, 118, 121 and 122). He may have a person stand up to speak to someone else, as with the phrase La pucele se drece which begins laisses 6l and 62 of Le Siege de Barbastre. or simply have him speak without any posture being indicated, as with Sar- razjn. ce dit Bueve3 . which opens laisses 27 and 29 of the same poem.

For reactions to the situation with which the previous laisse ended, many laisses (nearly one-sixth of those in Buevon de Conmar­ chis) have in their initial verses verbs of seeing and hearing, as some character takes cognizance of the action of another and acts himself in turn. This procedure tightens the structure of the poems by linking the end of one division with the beginning of the next in a sort of chain relationship.

The temporal conjunction quant is found in the first verse of many laisses, especially in Buevon de Conmarchis t. Qant paien entendirent la voiz de Clarion. SB: 1331 Qant li dus Girart fu sor Ferrant remontez. SB: 1350 Qant la pucele ol les paiens de Nubie. SB! 1537 Qant ce a antandu la pucele ennoree. SB: 1558 Qant Girart entandi le cons el de Guion. SB: 2491 Quant css nouveles oent nostre gens de Nerbonne. BC: 203 Quant Guillaumes d* Orange vit si grief le besoing, BC: 306 Quant Franpois retornerent contre les Arrabis. BC: 484 Quant l'amustans entent Corsolt de Tabarie. BC: 620 Quant Bueves entendi si parler l'amustant. BC: 670 Quant li Sarrazins ot la chartre deff ermee. BC: 845 Quant Barbastre ont conquise Franpois, moult en sont l±e7~BC: 1398 Quant l'amustans de Cordres ot la nouvele oye. BC: 1503 Quant l*amustans de Cordres a nostre gent veue. BC: 1566 Quant nostre Franpois virent 1 * ami rant aprochier. BC: 1746

^In this list and in the others which are found in chapters H through V H , all of the verses given are not necessarily formulaic. The actual number of formulaic hemistichs found will be tallied at the end of each chapter. 35 Quant Bueves ot veus les cuvers desfaes. BC: 1867 Quant Gerars rot son bran, merveilles en fu lies. BC: 1969 Quant la gens l*amirant ont Buevon perceu. BC: 1998 Quant li aatirans ot la parole escoutee. BC: 2090 Quant Gautiers de Tolouse le conmant entendi. BC: 2125 Quant nostre Franpois voient l'amirant aprochier. BC: 2843 Quant Gerars vit son pere de corrouz alumer. BC: 3010 Quant Guis de Conmarchis qui n*ot pensee avere. BC: 3085 Quant Gerars ot que Guis ses freres le reprent. BC: 3098 Quant Gautiers ot parler Gerart en tel maniere. BC: 3293 Quant Gerars entendi Navari le sa chant. B C : 338O Quant Gerars entendi de Guion la maniere. BC: 3^59 Quant Guis de Conmarchis a son frere escoute. BC: 34-77 Quant Gerars ot son frere qui est de tel afaire. BC: 3566 Quant Malatrie entent que Guis de Conmarchis. BC: 3754-

The first verse of a laisse sometimes describes the fierceness of a battle -which is about to be recounted. 7

Molt fu grant la bataille, li estors esbaudis. SB: 1238 Molt fu fors li estors, ce sachies, et mellez. SB: 2206 Molt fu granz la bataille et 1* est or fu plenier. SB: 2688 Molt fu granz la bataille et li estors pleniers, SB: 2596 Le Jour de Pentecouste, qui maint home atalente, Fu moult grans la bataille seur l*yaue de Clarente. BC: 285-86 Fiers fu la bataille, moult fist a ressoingnier. BC: 343 La bataille fu grans et perilleuse et fiere. BC: 401 Au retraire des nostres fu grans li fereis Et li estours morteus et fiers li chaplets. BC: 1590-91 Fiere fu la bataille et li estours pesans. BC: 1942 Fiers fu li pongneis plus que ne vous devise. BC: 2820 Fiere fu la bataille et asprement enprise. BC: 3779

Molt fu granz 3 a faatai 11 a in Le Siege de Barbastre and Fiere fu la bataille in Buevon de Conmarchis thus constitute formulas which are used only to introduce laisses, but expressions similar to them do

occur in other positions (SB: 278, 1240; BC: 1602, 1615, 3928). Ano­ ther verse -which opens a laisse and is similar in its inverted word order to the ones quoted above, but which does not have to do with bat­ tle, is: Molt fu bele la chanbre. gente et de grant richor. SB: 991*

?See Ity-chner, La Chanson de Geste. p. 72. To these formulas may be added several verses which tell of the joy or sadness of the characters.

Dolant furent paien qant virent mort Gaudras. SB: 1215 Dolant fu Libanor qant s*o3! menacier. SB: 2013 Dolanz fu Libanor qant se santi en l ,eve, SB: 20**8 Joiant furent Franyois dedens la tour antie. BC: 1034 Grant joie ot Malatrie quant de Gerart oy. BC: 2675

The poet of the anonymous work uses substantially the same expres­ sion in the first hemistichs of his three verses (with the neces- saiy change that the plural subject and verb require) and the second hemistich is, in the three cases, a tenqjoral clause beginning with qant. Adenet, in the two cases There he comes up against the same problem, constructs his two verses differently.

For the final, six-syllable -petit vers of each hemistich four formulaic uses are found in the two poems.

Whenever the laisse ends in a prayer to God for help, whether this be by the poet himself or in the mouth of one of the heroes, it terminates in a petit vers which continues this prayer. It ' sometimes happens that the prayer is limited to the petit vers, in which case it may be at the end of any laisse in which the French are in danger.®

When the Saracens are involved in a battle with the French, the laisse can end in the words gent sarrazine. combined with two other syllables in whatever pattern is required by the syntax of the sentence, such as de la gent sarrazine. sor la gent sarrazine o or contre gent sarrazine.

®See below, pp. I3I-3I*. ^See below, pp. 66-6?. 3 7 Other expressions sometimes found as petits vers ares

A nul jor de sa vie. SB: 104p A nul jor de ma vie. SB* 650p, 2672p and

Si dirons les noveles. SB: 1068p Ja dira ses noveles. SB: 1103p BC: 3320p Si ont dit lor mesage. SB* 153%) Ja lor dira noveles. SB: 2449p Ces nouveles li eonte. BC: llOlp

The first of these is found after general expressions as a sort of hyperbolic intensifier, the second after a laisse -which ends with a messenger motif

Transitions

If the ability to make smooth transitions from one subject to another in the course of a composition were the mark of a good author, Adenet le Roi would not qualify for that honor. Some of his transitions are merely statements that he is going to stop talking about one subject and begin another.

D’Admeri vous lairai ore a parler ici, D'Ermengart de Pavle et de Guillaume aussi, Si dirons de Buevon qui ot le cuer mari Et de ses deus enfans, de Gerart et de Gui, Et des autres barons que paien ont saisi. BC: 596-600

Other occurrences include Ja lor dira noveles. SB: 98a; Ja dira son mesage. SB: 1105b; Si dlrai mon mesage. SB: 2425a; Cis lour dira nouveles. BC: 150a; Si haut dist ses nouveles. BC: 1483aJ Et dire mon message. BC: 3280a; La novele conter. SB: 1187b; Mes noveles conter. SB: 2444b; Des nouveles qu»il conte. BC: 1813a; Ces nouveles non- cier. BC: 3283b. De Corsolt ci endroit a parler vous lairons, Quant tans et lieus en iert, moult tost i revenrons. BC: 1378-79

Mais d,aus vous laisserons ore a parler ici, De Corsolt vous dirons qui ot le cuer mari. BC: 1472-73

Mais d'aus vous laisserai ore a parler ici, Si vous dirai conment Persant et Arrabi Bt Turc et Achopart et Conmain et Luti S'en vont droit vers Barbastre, de guerre aaati. BC: 1821-24

Des Franjois vous lairai ore a parler ici. BC: 2131

De li ne vous iert ore plus parole contee. BC: 2104

Mais d*aus vous laisserai ore a parler a tant, Si dirai de Gerart le hardi combatant. BC: 2428-29

Un petitet de chose vous ai en oubli mis Que je vous dirai ja, n*en sui pas esbahis. BC: 2782-83

De li ne vous iert ore plus raisons racontee, De Buevon vous dirai cui moult foment agree De ce que sagement a sa gent ramenee. BC: 2980-82

Des Franjois vous lairons ore a parler a tant, Si dirons ci endroit de la gent Tervagant. BC: 3153-54

De ce ne vous iert ore plus parole acontee. BC: 35 H

Mais un petit de chose i avoie oublie. BC: 3808

None of these verses appears at the beginning or at the end of a

laisse, ■which is "where one would have expected thepoet toplace them.

This fact alone shows that Adenet gave only perfunctory attention to

the concurrence of f o m and content in his strophic structure.^

Laisses are, as has been noted, linked with each other in that the

^ I n justice to Adenet it must be said that the poet of Le Siege de Barbastre once employs this obvious transition, but in a passage ■which is beyond the part of the poem with which we are concerned here: SB: 4488. 39 opening verse of one refers in some way to the closing verse of the previous one, but the subject sometimes changes abruptly right in the middle of a laisse, and one wonders if for Adenet the avail­ ability or scarceness of words of a particular rime were not the criterion by which the size of laisses is often determined.

Another rather clumsy practice of Adenet is that of cutting off the subject which he is speaking about by merely saying that there is no reason for dragging out the account.

Que vous diroie je?^ BC: 596a» 973&, 3036a, 3264a Que vous feroie je de ce lone parlement? BC: 1333 Que vous iroie je plus la chose alongier? BC: 2363 Que vous iroie je la chose parloignant7 BC: 3704

One cannot help feeling that these verses signify a temporary drying up of the poet's imagination.

Signs of Performance

Neither the anonymous poet Of Le Siege de Barbastre nor Adenet le Roi could” ever be accused of being impersonal in their storytelling as both frequently interject their personalities into the narrative and address the audience directly or comment on what is happening.

The intervention may take the form of a verse calling the atten­ tion of the audience to what is to follow or telling it that it is about to hear a certain episode.

Oiez quele aventure li a Edens envoie. SB: 577

■^When a formula is contained in a hemistich, only that hemistich will be quoted, unless the rest of the verse is thought to be of some particular interest. uo

Oiez quele avanture Jesu li a done. SB: 2369

Au jor de Pantecoste, dont vos m ’oez conter. SB: 47 Si les a apelez con vos oir porrez. SB: 838 Bueves parla premier, * con ja ofCr porrez. SB: 889 Malaquin le salue, con ja oxr porrez. SB: 24-38 Pour delivrer no gent si com ay aves. BC: 1058" Leur aimeares portent ainSi com vous oes. BC: 1091 Et que baptisie furent, si com avez oy. BC: 1471 Nerbonnois repairerent si com avez oy. BC: 1805

The e:xpression oiez quele aventure occurs in circumstances which are similar. Its first appearance is after Buevon and his sons have been thrown into prison and Buevon has prayed desperately to God for help.

Damedieu, sire pere, qui me fextes ne, Car me donez la mort einz que soit avespre, Ou tel home m*envoies par la teue bonte Que de ceanz me giet ennuit a sauvete. Oiez quele aventure li a Dieus envoie; Leanz ot un paien de grant nobilite, Clarion de Vaudune l*ont Sarrazin nonme* (SB: 573-79)

Clarion, disgruntled at the way the amirant has deprived him of half of the castle of Barbastre which is rightly his by inheritance, is at that precise moment lying in his bed thinking of how good and just the

French are and how they would never disinherit him if they were in control of the town. He decides to go to Buevon and his sons and free them. Later in the poem Malatrie is faced with similardif­ ficulties. Her father wants her to marry Libanor, who is obviously not as good a knight as Girart since the latter has defeated him in single combat. Nevertheless her father berates her for speaking to

Girart and encouraging him to pursue her after the encounter between the two rivals. She too decides that the only way for her to secure to.

her wishes is to go over to the French cause. She is in bed while

thinking all this,

Mes ele ne dormist por l*or d*une cite, Einz reclaims Jesu, lou roi de majeste, Tel home li amoint, par la serue bonte, Que ele puisse dire son cuer et son panse. Oiez quele avanture Jesu li a done. Leanz ot un paien de grant nobilite, Malaquin de Sulie, einsi l*o5 nomer. (SB: 2365-71)

Malaquin, faithful to his lady who has a short time before dubbed him

a knight (SB: 1600-10), then goes as a messenger into Barbastre

and asks Girart, in the name of Malatrie, to visit the tents of

the Saracen ladies with several of his friends. Again God has sent

the French help in the form of a Saracen convert. The formula in

these two cases seems to be a kind of tag or trigger that the poet

associates with the stock situation: the advent of help in the form

of a turncoat from the enemy camp.

Jongleurs are always ready to swear to the accuracy of the

tales of their chansons, and indeed it seems that their poems were

looked upon by the people who heard them as containing real his­ torical accounts.^ Adenet and the anonymous poet are no excep­

tions to this desire for credence, as these phrases show:

Ce vos di sanz fauser, ^. SB: 51h Eh la terre ai este, pour ce le vous tesmong. BC: 314

jFor borrowings by writers of chronicles from this popular his­ toriography, see Ramon Menendez Pidal, La Chanson de Roland et la Tra­ dition epique des Francs (Paris: A. et J. Picard et Cie., i960), pp. 445-47. 14 This claim to have been in Saracen lands may be true, as Adenet accompanied his patron Guillaume de Flandres on the crusade of Louis IX in 1270-1271. See Henry, Les OEuvres d*Adenet le Roi. Vol. I: Bio- graphie d»Adenet — La Tradition manuscrite. p. 22n. De ce soiies tous fis. BC; 960b, 1622b Pour voir le vous disons. BC: 1373b Pour voir le vous plevi. BC: 2158b

They are not always quite as sure of themselves, however:

Plus furent de cent mi]., par le mien esclent. SB: 145 Par le mien esclent, des le tans Moyses. SB: 784 Bt d*or fin esmere sont fet, mien escjent. SB: 986 Que nus n*en set le conte, par le mien esc£ent. SBi 1481 Par le mien eseient prise fust la cite, SB: 2218 K,eschaper ne peussent, par le mien esciant, BC: 1137 Tel gent en sont saisi, par le mien eseient, BC: 1308 Ja fust mien eseient Gerars a mort livres. BC: 1908 Ne n*a talent dou rendre, par le mien eseient, BC: 2229 S*il ne fust, noiies fusse, par le mien esciant. BC: 3163

Frere, ce a dit Guis, ainsi com je le croi. BC: 234 Pres furent’de dis mil, ainsi com je l*antent. BC; 272

Des melllors de la vile, ainsi com je 1*entent, BC: 1338 Poi sont plus de deus cens, ainsi com je 1*entent, BC; 2222

Bueves et si enfant, si com moi est avis. BC: 507

Such hemistichs can be interpolated at almost any point in the story

to fill a verse which contains factual material, but they also serve

to reinforce the feeling that the medieval man undoubtedly had that

what the jongleur was saying was true, and that if he were not certain

of the accuracy of what he was telling, he would inform his audience

of this. The tone of a chanson de geste is one of complete fervor

and absolute sincerity.

The poets of both works point out heroes or groups of characters

in the course of the narrative.

Es Hunaut de Bretaigne poignant de randonnee. BC: 462

Es vos Tin chevalier brochant a esperon. SBi 125 Es vos au chief del pont catorze Nubiant. SB: 196 Es prez desoz Nerbone es vos paien logiez. SB; 348 Es vos nostre Franpois vers les paiens alez. SB: 824 Es vos nostre Franpois enz el pales armez. SB: 922 Es vos devant Buevon les Franpois retomez. SB; IO69 Ez vous un chevalier de l'esploitier se paine. BC: 147 Es vous par la bataille roi Balistant d ’Octrente. BC: 298 Es vous par la bataille Espinart d'Avaloingne. BC; 328 Es vous par la bataille Renaut de Montarmier. BC: 361 Es vous par la bataille Buevon de Conmarchis. BC: 500 Ez vous poignant Sorb rant parmi le pre herbu. BC: 2001 Ez vous un Sarrazin moult richement monte. BC: 2067 Ez vous un Sarrazin qui ens el tre entra. BC: 23U Es vous Gerart poignant encoste une falise. BC: 3790 Es vous un Sarrazin, le frain abandonne. BC: 3841

Estes vos par la pree Guielin et Buevon. SB: 1336 Estes vos le mesage corant tout abrive. SB: 1524 A icele parole, estes vos un Esder. SB: 2650 Estes vous un mesage qui a pie descendi. BC: 2148

Es les vos tout un val a esperon brochant. SB: 167 Qant Sarrazin l»oirent, es les vos aprestez. SB: 803 Qant ce virent paien, es les vos effreez. SB; 829 Es les vos meintenant aprestez et gamiz. SB; 1254 Chascuns en prent sa part, es les vos arrees. BC: 1092

A tant es un mesage qui del aler se paine. SB: 91 A tant es acorant Berangier et Sanson. SB; 238 A tant es par les rans Otran de Baleguez. SB; 1352 A tant es un mesage a espreon brochant. SB; 1620 A tant es tin mesaje pansant d*esperonner. SB; 1727 A tant es libanor qui descendi au tres. SB; 1741 A tant es Emengart, la contesse au cors gent. BC: 554 A tant es Clarion, vers aus vint tout en mi. BC: 1017 A tant es l*amirant qui Buevon escria. BC: 1117 A tant es Sarrazins p a m i le pre herbu. BC: 3913

A tant es vos Girart, le fiz au due Buevon, SB: 107 A tant es vos Guillaumes d»0renge le guerrier. SB: 208 A tant es vos Buevon qui le corage or fier. SB: 384 Atant es vos Gaudras, qui s*est haut escrSez. SB: 1202 A tant es vos Girart et Guion par les prez. SB: 1449 A tant es vos pongnant le conte Guielin. SB; 2170 A tant es vous poignant Guillaume le guerrier. BC: 350 A tant esvous poingnant et Rainfroi et Gautier. BC: 367 A tant es vous Buevon, le preu et le sene. BC: 445 A tant es vous Franyois poignant de randonnee. BC: 1708 A tant es vous Corsolt poignant de randonnee. BC: 1718 A tant es vous venu le Tolousant Gautier. BC: 3275

A itant estes vos le pere et les deus fiz. SB: 2202

These epideictic phrases are extremely useful to the poets, as they provide fillers for hemistichs •which would otherwise be lacking from 44

one to six syllables; es (one syllable); es vos (2 ); estes vos. es

les vos and a tant es (3)» a tant es vos (4); a itant estes vos

(6 ). At the same time as they solve a technical problem of versi­

fication, they inject an element of excitement into the story by

their exclamatory character. Some of them are combined with other words to constitute formulas: es vos nostre Francois (2 examples in

SB), es vous par la hata^ne (4 in BC). es vous un Sarrazin (3 in

BC). a tant as ■urt mesage (3 in SB), a tant es vous pongpant (2 in

BC). a tant es vous plus a proper name (5 in SB, 3 in BC), and a

tant es plus a proper name (2 in BC). Of these formulas 14 are in

BC. 10 in SB.

Further evidence that the chansons de geste were presented before an audience is provided by the formula which opens La ve3!s-

siez . . . and goes on to enumerate what the listener would have

seen had he been present at the event which is being described. The

objects that would have been seen are usually prefaced by the adjec­

tives tant (in SB) or maint (in BC).

Ce jor i ve£ssiez tant lance froisier, Et noz gentis Franjois sor Sarrazins aidier, Ces chies et ces viaires ledir et depecier. SB: 217-19

La veissiez de lances si tres grant froissement Et d*espees sor hiaumes si tres grant chaplement Que sor terre en gisent maint Sarrazin sanglent. BC; 265-67

La veissiez moult tost mainte targe embraeie, Maint eseu et maint hiaume ou li ors reflambie Et maint paien qui ot d ,assai,llir grant envie. BC; 1518-20

Maint arc, mainte arbaleste i veissiez tendu, Maint escu, mainte targe et maint dar esmolu, Mainte pierre geter et maint cailleu cornu. BC: 1537-39

La veissiez tendu maint paveillon, maint tre 45

Et mainte riche tente a maint ponmel dore. BC: 2065-66

A l'assambler i ot et grant noise et grans cris, La veissiez de lances moult tres grant froissels Et maint escu percier blanc et vermeil et bis. BC: 2802-04

Par l ,estour velssies mainte anste en tronjons mise Et maint paien gisant sanglent souz la chemise. BC: 2823-24

Veissiez par les chans maint cheval estraier, Lor resnes traynant contreval le gravier, Maint escu, mainte targe gesir deseur l ,eibier. BC: 2866-68

A l'assambler i ot reverse maint penel Et maint foie pereie et coupe maint boel; Qui brisie ot sa lance le bran traist dou forrel. La veissiez moult tost conmencier tel maisel Dont de mainte armeure rompirent li clavel Et parmi les chans furent espandu maint cervel. BC: 3732-37

Par les chans velssies des lances mainte astele, Maint cheval estraier corant a vuide sele, De quoi li seignor gisent sanglent par la vaucele; Grant noise et grant huee ot parmi la praele. BC: 3743-46

The series of objects (usually pieces of armor) -which are enumerated in these passages constitute a review in miniature of the actions of the battle. The La veissiez . . . formula thus presents a panorama of the fighting and allows the iongleur to devote his full talents to the single combat scenes within the larger battle.

Enumerations with tant and maint often occur without the epi- deictic expression La veissiez . • . , and even though in this form they are not necessarily evidence of oral presentation, they will be listed here for the sake of convenience.

Einz en ert tant escu, tanta lance froee; Tant paien gite mort, envers, gole baeej SB; 39-40

K ’ades les guerroiierent sans pitie, sans manaie; Souvent les enchaucierent par bos, par mainte haie, S ,en orent sor les cors maint coup et mainte plaie. BC: 48-50

Ains en seront percie maint escu a lion 46

Et mainte pesant targe et percie maint blazon Et en avra on taint maint vermeil siglaton, S'en ©rent derrompu maint hauberc fremillon, Maint e entraille coupee, maint foie et maint pomon. BCs 191-95

Hui nous avez fausse mainte broingne doubliere, Maint vaillant Sarrazin gete mort en l'ourdiere De vo trenchant espee, cui la passions fiere, De nostre meillor gent avez fait mainte biere. BCs 422-25

Maint cembel i ot fait et mainte retomee. BC: 469

Mainte baniere i ot au vent desvolepee Et mainte riche ensaigne en son lance levee. BCs 47^-75

Maint tre i ot tendu et mainte tente mise, Mais ainjois que il l'aient reterne ne prise Sera sor mainte tente mainte fois nois remise. BC: 524-26

Ains en iert mains paiens sanglens sous sa chemise Et mains haubers percies desor pelice grise. BC: 533-34

Que par terre en verres maint Sarrazin gisant, S'en ierent derrompu maint hauberc jaserant, Mainte targe percie et maint pegnon sanglant. BC: 675-77

Mains des nos nous i ant navre et mal bailli, Et maint en ont gete a la terre estourdi Qui mais ne reverront ne parent ne ami, Maint hauberc nous ont hui rompu et dessarti. BC: 614-17

Ains i iert maint broigne percie et derronrpue, Mains hiaumes descercles, mainte targe fendue. BC: 921-22

Ains seroient coupe maint bras, maint poing, maint pis. BC: 942

Mainte fenestre i fu eft mains huis encharchies, Maint escu, mainte targe enbracier velssies, Maint martel et maint pic de fer bien afaities; A l*aprochier des murs i fu mains dars lancies, Mainte saiete traite, mains quarriaus descochies. BC: 1172-76

Par la porte s'en fuient maintre conraunaument, Maint en i laissent mort, maint navre, maint sanglent. BC: 1335-36

Car par toi i a mort mainte bele jouvente-; Parmi Barbastre en gisent maint a chiere dolente. BC: 1363-64

Par lor prouece furent puis maint escu percie Et maint elme enbarre, maint hauberc desmaillie, Maint Sarrazin navre, maint mort et maint blecie. BC: 1419-21 47

Maint paian i sont mort et maint s*en sont fui. BC: 1489

Maint Sarrazin i ot a ce poindre abatu, Mainte targe i ot frainte et percie maint escu, Maint hauberc desma1.l3.le, maint clavain derrompu, Maint cuvert Sarrazin parmi le cors fera. Panni l*ost des paiens font grant noise et grant hu, "Mahonmet, font il, sire, bien nous est avenu Puisque Nerbonnois sont dusk*a no gent venu." Tost furent vers no gent li plusour acoru; La ot le jour maint coup donne et receu, As brans d»acier i ot maint cervel espandu Aval la praerie. BC: 1555-65

As brans d'acier i ot mainte brongne rompue, Maint hiaume descercle, mainte targe f endue Et maint paien gete sanglent sor l*erbe drue. BC: 1581-83

A l'assambler i ot grant noise et grant huee, La ot maint escu fraint, mainte targe troee Et mainte grosse lance par pieces tronjonnee, Maint cheval estraier corant panni la pree, De quoi li seignor gisent sanglent gueule baee. BC: 1710-14

Mais ains verront des lor Sarrazin mainte biere Et verront despaner mainte brongne doubliere, Maint bouel trayner parmi la sablonniere Qu*il aient de Barbastre prise la tour premiere. BC: 1792-95

Mains Sarrazins en fu puis mors et detrenchies Et mains haubers rompus, mains hiaumes depecies; En celui jour meisme fu il si ressaiies Que mains paiens en fu laidement domagies. BC: 1970-73

Mnjois k*a l*amustant soit la citez rendue, Y ara mainte brongne percie et derompue, Maint hiaume depecie, mainte targe fendue, Maint coup feru de lance et d*espee esmolue Et maint paien gete sanglent sor l*erbe drue. BC: 2039-43

Mainte ensaigne i choisirent au vent desvolepee, Maint hiaume, maint escu, mainte targe doree, BC: 2107-08

Ainjois en sera morte mainte bele jouvente Et mains escus percies, mainte brongne sanglente Que je ne moustre en France en eest an mainte tente; Mainte dame i cult faire de cuer triste et dolente. BC: 2241-44

Mains Turs, mains Sarrazins la bele convoia. BC: 2331 Mains en noia le jour dedans cele riviere Et mains an gist sanglens parrai la sablonniere. BC: 2900-01

Mais de lor gent i a maint mort, maint afole. BC: 2919

La prendent tin conseil dont mains paiens verses Fu puis et mains haubers rompus et depanes Et mains escus percies, mains, hiaumes descercles. BC: 3203-05

Que, par icel Seignor cui baptesme tenons, De lances et d»espees ainpois tant i ferrons Que vuis en remanra mains destriers arragons Et en sera percies mains foies, mains ponmons, Coupes et detrenchiez mains escus a lyons Que la bele pucele o nous n'en aportons. BC: 3612-17

Maint escu i ot frait et mainte targe bise Et mains cors de vassal sanglent souz sa chemise. BC: 3783-8*+

Maint ruiste coup i ot departi et done Si que par terre en jurent maint mort, maint afole. BC: 38O6-O7

Dedens le pas se sont li plusor eribatu, Maint en y a noiie et maint s*en sont issu. Moult par i ot grant noise et grant cri et grant fu, La endroit ot maint coup donne et maint rendu. BC: 3921-2*f

Many of these passages begin with Ains (einz) or Aincois. and express the idea that before the desired object of the battle is won, many martial feats will have to be accomplished. It will be noted that only two of the enumerative passages are from SB. Adsnet le Roi has used the device to excess, and this is one of the main reasons, in my opinion, why his story is so lifeless. In almost every battle scene there is one (or several) of these banal lists of shattered helmets, broken shields and wounded pagans. One comes to expect them, but they add nothing to the narrative. This feature of Adenet*s style will be commented on again below,^

15P. 193. H9 \ Of the expressions treated in this chapter, 45 from SB and

60 from BC were repeated at least once within the same poem in

substantially the same form and can therefore be considered to be formulas. CHAPTER H I

EPITHETS

In this chapter •will be treated those nouns in apposition, adjectives and adjectival phrases and clauses which, occurring repeatedly with certain nouns, are usually called epithets. A further effort to define the epithet in the chansons de geste will be made at the end of the chapter when the relationship between noun and modifier is seen more clearly, that is after the examples found in the two poems have been seen in their entirety.

Epithets will be grouped according to the nouns with which they are found, even though it sometimes occurs that an epithet modifies nouns of different classes (for example auferrant nrisie.

SB: 2146b, and Hunaus le proisies. BC: 3315b). When this latter phenomenon occurs it will be noted. The reason for such a classi­ fication of epithets is that with it one can see just what epithets the poet had at his disposal for each different class of nouns.

The order of discussion will be: proper and common nouns denoting persons, proper and, common nouns of place, parts of the body, horses, amnor and articles of dress, buildings, and finally natural objects.

50 51 Proper and Common Nouns Denoting Persons

The most characteristic use of the epithet in Old French epic poetry is to ornament the names of heroes. It is only natural, then,

that in a genre which is dominated by the ideals of feudal warfare,

a large number of epithets should be devoted to the aggressive qualities of the warrior.

At times knights are described simply as •♦warlike.”

A tant es vos Guillaumes d»Orenge le guerrler. SB: 208 Llbanor le guerrler. SB: 1804b Fbrs seulement Guillaume au cort nes le guerrler. SB: 2331 Guielin le guerrler. SB: 2603b Guillaume le guerrler. SB: 350b Aymeri le guerrler. SB: 357b Et Hunaut le guerrler. BC: 3^50b

The word ber, used as an adjective, denotes skill and courage in

battle.

Ice dites Girart del Conmarchis le ber, SB: 2395 Que fetes, Hunaut ber? SB: 2602b Et Havari le ber. BC: <>3b, 3027b

The knight is sometimes called "hardened to war."

Con vassaus adnrez. SB: 14-32b Le vassal adure. SB: 1978b Com vassal adure. BC: 1201b Guillaume l*adure. BC: 1693b Com vassaus adures. BC: I89^b Et Joffroi 1*adure. BC: 2059b

In SB he is t e m e d hardi. and in BC this adjective is found in con­

junction with the substantive combatant (or combatere).

Li chevalier hardi. SB: 1806b

Le hardi combatant. BC: 679b, 2429b Le hardis cowbatans. BC: 2459b Et hardis combatere, BC: 3^90b 52

Likewise fier is found with a proper noun or with the noun vase- lage. but neither of these uses is present in BC.

Li fiz Buevon le fier. SB: 2598b Li cortois et le fier. SB: 2609b Avez vos pris Girart del Conmarchis le fier? SE: 2627b Ne fust li dus Girart del Conmarchis le fier. SB: 2641b, 2658b 0 le fier vaselage. SB: 110?b

The notion of the knight as a good vassal is frequently expressed in both works, but especially in BC.

Con vasaus esprovez. SB: 1406b, 1447b Et vassaus esprouves. BC: 708b

Con vassaus adurez. SB: 1432b Le vasal adure. SB: 1978b Com vasal adure. BC: 1201b Com vassaus adures. BC: 1894b

Conme vassaus prisies. BC: 1183b, 1993b

Com vassaus esleus. BC: 1916b Le vassal esleu. BC: 2019b Vassaus preus et eslis. BC: 2781b

Le vassal esmere. BC: 2062b Pour vassal esmere. BC: 2962b

Con vassaus preus et fier. BC: 2858b Com vassaus sans faintise. BC: 3796b

The sum of all the knightly virtues is proece:

Qni prouece a dnree. SB: 5b A cui groece agree. SB: 17b Ou proece s*alie. SB: 1545b, 1551b Qui proece governe. SB: 2067b Et de haute prouece. BC: 1195a De tres haute prouece. BC: 1631a, 3216a Qui proece ot entiere. BC: 3300b

Et li preus Clarions. BC: 1376b Li preus de Conmarchis. BC: 2271b Gerart le preu son frere. BC: 3087b

The adjective preu is sometimes combined with others to fora a mul­

tiple epithet. 53 Li preuz et li senez. SB; 94lb Le preu et le seme. BC: 445b Qui preus est et senes. BCs 3212b

Con son preu et vaillant. SB; 1505b

Qui est preuz et hardiz. SB: 2320a Qui preuz est et hardiz. SB: 2334a A preu et a hardi. BC: 2687b Et Renaus, conme preus et hardis, l»a loe. BC: 3484 Qui preus fu et hardis. BC: 3777b

A highly laudatory compliment for a knight is that he is feared by

his enemies.

Qui tant est redoutez. SB: 1401b Qui si est redoutez. SB: 1404b Qui tant par est doutez. SB: 1198b

The last of these phrases, strangely, is applied to the sea.

It is to be noted that in all of these examples of adjectival

expressions of bravery or prowess applied to various knights, only

three are alloted to orthodox Saracens: Libanor in SB: 1804 and BC:

708, and the amirant in BC: 2858. In all other cases these most

desirable of qualities are found in French knights or in Saracens who have converted or at least given signs of coming over to Chris­ tianity. The Christian warriors of the chansons de geste do not see

their enemy as worthy of praise, and even when they are defeated, it

is by "unfortunate cowards" as in BC: 1654b. Libanor is an exception to this rule, but one feels that he is presented in a favorable light

only so that more glory might accrue to Girart when he eventually

defeats him in single combat. Likewise Guielin and the amirant are together called vassaus preus et fier. but a few verses later Gerars

steals the Saracen leader's horse (BC: 2875-76). The contrast is 5* black and •white — only Christianity makes a man brave or good.

Let us now pass from the qualities 'which are related to the con­ duct of the knight in battle to those ■which are not exclusively con­ nected with war. In this categoxy may be included the second member of the double epithet which has been treated above,^ sane, since wisdom is valuable in any undertaking and not specifically limited to the battlefield.

Hermsnjart la senee. SB: 10b Ne Guibert le senez. SB; 710b Guielin li senez. SB: 2398b Dux Bueves li senes. BC: 1910b Emengart la senee. BC: 2996b

There is another, longer expression for this same idea:

Ou il n*ot k*ansaignier. BC: 378b, 3267b Eh cul n*ot k'ensaignier. BC: 23506 K'en lui n»a k'ensaignier. BCs 3440b

Senee is applied to Hermenjart and En cui n»ot k tensaignier is said of Malatrie. Thus although both these attributions of wisdom are made to men, the virtue is not a purely masculine one in the two poems.

Sachans, another word indicating acuteness of Judgment, is not found in SB.

Clarfons li saehans. BC: 1953b Sachent par verite. BC: 2063b Dist Gerars li saehans. BC: 2479b Navari le sachant. BC: 3380b

Among the more strictly courtly virtues are frankness, honesty and courtoisie.

1P. 53. 55 Dame Hermenjart la franch©. SB: Et Navaris li frans. BC:

Aimer li cortois. SB: Clarion li cortois. SB: Malaquin li cortois. SB: Li cortois et li fier. SB:

Franc chevalier oneste. SB: 2062b,

A knight’s renown is often as important as any intrinsic

xlities.

Qui sont de moult haut pris. BC: 504b,

Le chevalier meribre. BC:

Et Hunaus le proisies. BC:

Et Geufroi I’alose. SB: Et Hunaut l’alose. SB: 2443b, Ne Huon l'alose. BC: Le preu et l’alose. BC: Qui tant est aloses. BC: Qui moult est aloses. BC:

Applied to men and women we find examples in BC of the adjective m i 11 arit, meaning "of value" or "brave,"

Guillaumes li vaillans. BCs 1963b La pucele vail 1.ant. BC: 244-7b S'amie la vaillans. BC: 2455b

Gente. "of noble birth," is used exclusively of Malatrie, who is later to become the bride of Girart at the end of SB.

Malatrie la gente. SB: 1992a Malatrie la gente. BC: 715a, 2231b, 2308a, 2932a, 3134b, 3166a, 3649a, 3656a, 3681a

In personal epithets an adjectival clause is sometimes found instead of a simple adjective when there is need of filling a whole hemistich.

Qui n ’ot pensee avere. BC: 97b, 3085b Clarions, en cui n*ot nule pensee amere. BCs 1452

Et ses enfans en cui n*a pensee laniere. BCs 1737 Par samblant n*avoit pas la pensee laniere. BCs 2388 Et Joffrois en cui n*ot pas pensee laniere. BCs 3299 Cis consaus ne vint pas de pensee laniere, BCs 3436

Je croi bien k*anvers nous n'ait pensee renarde. BCs 3537

Three of these epithets, BCs 1452, 3299 a*id 2388, are not contained entirely in the second hemistich. This type of epithet "which crosses over the caesura is found only in BC. The poet of SB is more atten­ tive to the needs of the rhythm than Adenet and would not join the 2 two hemistichs in this manner.

At times adjectival phrases are used with nouns in apposition.

Con horn de grant vertu. SB: 1048b Com gent de grant vertu. BCs 3908b

Com hom bien ensaignies. BCs 3306b

Com gent de fiere guise. BCs 3800b

These expressions too take up an entire hemistich, although their

. ^There is a large set of relative adjectival clauses, found for the most part after verse 2696 of SB. which furnishes an excel.! ant example of the rigidity of the formulaic epithets in SB. These clauses' describe the power of God or, with a change of one word, the power of the king of France. Qui France a a garder. SB: 3132, 3229, 3808, Looys 3844, 3855, 3824 Qui France a a baillier. SB: 3595, 4481 Qui France a en baillie. SB: 3918, 5188, 5637 Qui tot a a garder, SB: 388I Qui tot a a baillier. SB* 2098, 4018, 4274 Jesus BCs 1781, 3 W f Deus Qui tot a en baillie. SBs 3295, 3329, 3339 Mahom Qui tout a a jugier. SB: 1891, 3589 BC: 345 Qui tout a a sauver. SB: 3121, 3259 BC: 1444 57 semantic function is essentially no more than that of a simple

adjective. In meaning there is little difference between the simple bon and the compound con horn de grant vertu.

Often the epithet consists of a noun in apposition followed

or preceded by a simple adjective.

Seignor, frans chevaliers. SB* 980a, 1023a, 1058a Li chevalier loial. SBs 1057b, 106lb Li chevalier hardi. SB; 1806b Franc chevalier loe. SB; 2439b Vo droiturier seignor. SB; 868b La dame droituriere. BC: 407b

This construction too performs essentially the same function as the

cimple adjectival epithet as far as the meaning goes.

The French knights and ladies are described physically as well

as morally by epithets. Young men are quick of movement.

Cil legier bachelor, SB: 59b, 155a Fouquere le legier. SB: 212b

Ft Jeufroi li adroiz. SB: 884a

The older heroes are distinguished by their beards.

Ne Garin d*Anseune, qui a le poil melle. SB: 709

Et Richart li floriz. SB: 314b Et l ,amustant de Cordres a la barbe florie. SB: 1270

Et Huon le Mansel qui a la barbe grise. BC: 983 Gautiers li Toulousans qui ot la barbe grise. BC: 3795

Ki veist Aymeri a la barbe chenue. BC: 1579 Gautiers li Toulousans a la barbe chenue. BC: 3937

No one of these physical epithets is applied to one person to the

exclusion of others. The only one that is the sole possession of

one man in these poems is the famous au cort nes of Guillaume:

Et Guillaume au cort nes. SB: 708b, 1980b A Guillaume au cort nes. SB: 1193b 58 Guillaumes au cort nes. BC: 1880b

This epithet had the form al curb nes. *»with the hooked nose," earlier in the tradition, as is shown by the fact that the scribe of the Chan­ son de Guillaume employs curb 38 times and curt only twice. Later in the twelfth century curt became the preferred form, and in the Charrol de Nlmes a poet, ignorant of the original curb, even tells how William had a piece of his nose cut off by the giant Corsolt in a battle before 3 the walls of Rome and became Guillaume au cort nes.

There is one epithet of physical description which is applied indiscriminately to men and women in SB.

Malatrie la bele qui a clere fayon. SB: 440 Et Navari le conte a la clere fayon. SB: 1338 La pucele se drece a la clere fajon. SB: 1597 M ’amie te donrai a la clere fajon. SB: 1991 Et Sauvari le conte a la clere fajon. — SB: 2467

Another form of this same idea is found only in connection with women.

La pucele se drece o le viaire cler. SB: 1581 La pucele descent o le viaire cler. SB: 1706 Et verrons Malatrie la bele au cler viaire. BC: 3578

Et nies dame Heimengart la contesse auvis cler. SB: 1981 Et encontre sa fille Malatrie au vis cler, SB: 2225 II demands sa fille, Malatrie au vis cler. SB: 2340 Dieus, si grant duel demaine Rainmondine aucler visI BC: 487 Mainte bele pucele i ara a cler vis. BC: 2283

Of these 13 epithets which refer to the face, only three are from BC.

The body too is used in epithets, usually referring to women but

in one case to a man.

Ne Helisent m*amie, vo mere o le cors gent. SB: 566 A tant es Eimengart, la contesse au cors gent. BC: 554'

^Ramon Menendez Pidal, La Chanson de Roland et la Tradition epique des Francs (Paris: A. et J. Picard et Cie,, i9 6 0 ), pp. 419-21. 59

Sist l'amustans de Cordres les sa fille au cors gent. BC: 2201 Lors a mande sa fille au gent cors esmere. BC: 2931 Cult je bien reveoir Malatrie au.cors gent. BC: 3116

Puis vint a la pucele au gent cors avenant. BC: 2421 Si verrons Malatrie au gent cors avenant. BC: 3393

Malatrie ma fille, 0 le cors avenant. SB: 1495

La s* a sist Malatrie, la bele au cors mole. SB: 1517

Par delez la pucele, qui a le cors legier. SB: 2600 Malatrie les garde, la belle au cors legier. SB: 2624

Se Dieus sauve Aymerl, mon pere au cors poissant. BC: 678

Les les dames s'assieent as gens cors afaities. BC: 3664

Again there seems to be no conscious effort on the part of either poet to associate any one epithet with one person.

Malatrie is often called la bele. and this may -be an epithet which the poet habitually associates with her.

Malatrie la bele. SB: 440a, 1599a , 1801a Malatrie la bele. BC: 2413a, 3246a, 3428a, 3802a

La bele Malatrie. SB: 1538b, 1633a, 1677b, 1688b La bele Malatrie. BC: 2165b, 2712b, 3603a

Que Malatrie vient, la bele, l*eschevie. SB: 1685

To these add SB: 1517 and 2624 and BC: 3578 where bele is combined with another epithet.^ In the section of SB which follows verse 2696

there are ten more occurrences of this formula and all but one refer c to Malatrie.-' An alternate epithet in the form of a relative clause

is found in BC:

Malatrie la gente, ou tant ot de biaute. BC: 2932

4 The last of these is found above, p. 5 8 . The others are above on this page.

5 SB: 2821, 3105, 3247, 3 6 5 6, 4723, 4795, ^36, 5 0 2 0 , 7160, 7 3 2 6 . 60

Quo ne conquerons cole ou ot tant de biaute, BC: 3502

This formula is not, however, unknown to the poet of SB as it occurs later in his work.^

There is only one simple adjectival epithet found with the name of God,

Damedieu reclama, le pere droiturier. SB: 619 Biaus fieus, secores moi, pour Dieu le droiturier. BC: 390

This word is also used, in one case, to refer to the Virgin.

La dame droituriere. BC: 407

Other epithets which accompany the name of God are in the form of an adjective plus complementary phrase,

Hois plains d'umilite. BC: 844b, 1659b or of a noun in apposition plus complementary phrase,

Le roi de majeste. SB: 717b, 1379b, 1935b, 2366b, 2585b Le roi de majeste. BC: l684b or of an adjectival relative clause,

Qui de la seinte Virge nasqui en Biauliant. SB: 171 Celui qui de la Virge en Bethleem nasqui, BC: 2685 Qui de la seinte Virge nasqui en Belleant. BC: 3707

Qui me fextes ne. SB: 573b, 592b

Qui penez fu en croiz, SB: 893b Qui en croiz fu panez. SB: 931b, 1093b, 1357b, 1362b Qui en crois fu penes. BC: 1086b, 1889b

Qui se lessa paner! SB: 1594b

Qui en crois fu drecies. .. BC: 1196b 7 Puis que Dieus fu en crois mis pour les mescreans. BC: 194-7

6SB: 34-36, 3956, 3984-, 4416, 4704, 5260, 7 0 0 0 ,

^This is not an epithet but is analogous to some of the epithets quoted here, and so I have included it in this list. 61

Qui fist parler l'image. SB: 1122b, 1124b

Pour Dieu qui ne menti. BC: 1015b Qui onques ne menti. BC: 1469b, 1823b, 2684b

Par Dieu qui tout forma. BC: 1111b Qui le mont est ora. BC: 3242b, 3640b Qui tout le mont forma. BC: 3412b

Of these 17 are from SB, 19 from BC.

• Up to this point only those epithets have been considered which have a favorable denotation. The subjects themselves have been, for the most part, French knights and ladies. One might say that in the chansons de geste. except for an occasional traitor or convert, the division between friend and enemy is clear-cut. Epithets are dis­ tributed accordingly, and depreciatory and insulting terms are gen­ erally saved for the Saracen adversary.

Here are some of the epithets of insult from the two poems.

With felon:

Des felons Sarrazins, qui Dieus doint pute estraine. SB: 99 Des felons Sarrazins, c?ui Dieus doint male estraine. BC: 151b i Li Sarrazin felon. SBs 250b

Desoz paiens felons. SB: 427b.

Les felons maleoiz. SB: 525b

Nul felon losengier. SB: 910b Li felon losangier. SB: 2124b, 2284b

Levant meint Turc felon. SB: 1603b

Li felon traitor. SB: 1632b

Li felon renoie. SB: 2093b

Si felon dosnoiier. BC: 360b 62

Culvert:

As cuiverz renoiez. SB: 372b Ces culvers renolles. BC: 1192b Les cuvers renolles. BC: 1980b

Les cuiverz maleolz. SB: 502b De cuiverz maleiz. SB: 1243b 11 culvert malei. SB: 1294b, 1305b

LI culvert mescreant. SB: 989b Li culvert mescreant. BC: 1033b Li culvert mescreu. BC: 1215b, 1540bj 3914b Des cuvers mescreus. BC: 1918b Maint cuvert mescreu. BC: 2006b

Li cuvert desfae. BC: 1201b, 1654b, 3857b Les cuvers desfaes. BC: 186?b

Li culvert soudniant. BC: 1154b Les cuvers soudulans. BC: 1949b

Fel culvert parjurez. BC: 1354b

Lui culvert pautonnler. BC: 1836b

Traitor:

Sarrazin traitor. SB: 338b De traitor puant. SB: 965b Con treitres provez. SB: 1356b Et traitres clames.” BC: 1180a

Gloton:

Li gloton souduiant. SB: 1468b, 1508b, l6l6b

Fel gloton losengier. SB: 2240b, 2285b Dui gloton losengier. SB: 2242b Les glotons losangiers. SB: 2693b

Li glouton losengier. BC: 399b

Li glouton pautonnler. BC: 347b

Pautonnler:

Lui culvert pautonnler. SB: 1836b

®Add to these the examples above under felon, p. 61. 63 Lecheor pautonnler. SB: 2300b II glouton pautonnler. BC: 3^7b Cele gens pautonnlere. BC: 403b, 1798b

Losengier:

Deus garjons losengier. SB: 631b

Nul felon losengier. SB: 910b Li felon losengier. SB: 2125b, 2284b

Sarrazin losengier. SB: 2110b

Fel gloton losengier. SB: 2240b, 2285b IXii gloton losengier. SB: 224>2b Les glotons losangiers. SB: 2693b Li glouton losengier. BC: 399b

These tenas of opprobrium, all referring to the enemy (except in the laisse which tells of the argument between Girart and Buevon after the f o m e r returned to Barbastre without Malatrie, and which includes verses SB: 2240, 2285 and 2300 cited in the above list) appear to be varied, and it is, I believe, of interest here to txy to determine what differences there are among them.

Culvert, felon and trajCtor present no semantic problem for the modern reader since thej, _._ve the same meaning as the contemporary

French words couard. felon and traitre.

Gloton (nominative case gloz) is from the Latin gluttonem. In the Middle Ages this word had the meaning of the mo d e m English Mglut­ ton, " but in these poems it carries the more general sense of ’•miser­ able, despised person," which is widely attested, in the Roland for example.^

Losengier is related to the Old Provencal word lauzengier. which

^Walther Von Wartburg, Franzosisches Etvmologisches Worterbuch. Vol. IV (Basel: Helbing und Lichtenhahn, 1952), pp. 172-73* 64

designated, in the terminology of courtly love, a person "who spies upon lovers and reveals their secrets to the jealous husband of the

lady.^ The basic meaning of the word is "flatterer, but in the phrases with which we are dealing neither "detractor" nor "flatterer"

seem to fit. The word is always found in reference to the Saracens, whereas both of the meanings given above imply a certain degree of

social proximity in the courtly milieu: one can neither flatter nor

detract unless one is fairly close to the person who is informed or

cajoled. The word seems to have become less specialized in the con­

text with which we are concerned, that is in epic poetry, although

none of the Old French dictionaries register this extension of sense.

The only reference that I was able to find is in Bail Levy's Proven-

zalisches Supplement-Worterbuch where these verses of the fragmen­

tary Provencal epic At Maurin are cited:

Sonnent li graile e mesdent 11 caplers, No*l poc sufrir coars ni lauzengers.

Levy remarks: "Der Sinn verlangt etwas wie • Schwa chling.' Aber

kaim 1. die Bedeutung haben?"^ Apparently Levy had found no other

instances in Provenjal of this meaning, but it is obvious that it

would accord very well with the other words in our list from SB and

BC as a general term of opprobrium having nothing to do with the

10As, for example, in Bernard de Ventadour, Carl Appel, ed., Ber­ nard von Ventadom. seine Ljeder (Halle: Max Niemeyer Verlag, 1915), 31: 35 and 37* 42. ■^"Frederic Godefroy. Dtctionnaire de l'anciemne Langue franpaise. V (Paris: Vieweg, 1888), 3^«

^ B n i l Levy, ProvenzaHsches Supplement-WoVterbuch. (Leipzig: O.R. Reisland, 1894-1924), IV, 345. "flattery" of courtly life. The word is employed as an adjective in all the oases we have cited, so that we may adopt the meaning

"cowardly" here.

Pautonnler. from GQLd Low German *palta. "rag, "^3 simply has the meaning "boy, valet" in the twelfth century, but later takes on the more pejorative connotations which we see in SB and BC. 14 Gode- froy gives the adjectival meaning "dur, mechant, insolent, drole, lache."^

Thus it appears that none of these terns of opprobrium is very specific. Culvert and losengier entail cowardice on the part of the people to whom they are applied; tractor and felon Imply some betrayal of trust; gloton and pautonnler are simply general terms meaning

"worthless creature" or something of the sort. But Saracens are called traitor and felon when there is no particular question of betrayal (one can scarcely call a sworn enemy of the Christian faith, who has never been baptized, a traitor to it) and culvert and losen­ gier even when they are acting bravely (as in SB: 2124 where they are called felon losangier Just before one of them, Corsout de Taba- rle, challenges Girart to single combat), so one must conclude that all of these words used to refer insultingly to the pagan forces have

^^Von 'Wartburg, Franzosisches . . . , XVI, 6l6 .

^Wendelin Foerster, "Etymologien," Zeitschrift fur romanische Fhilologie. X XXH (1908), 429ff.

1 5 Vol. VI, p. h8 . 66 been lumped together into one vague category and that none of them has an exact meaning. This is perhaps no more than should be expected, for epithets are, after all, cliches, and the constant use of such fixed phrases as ve have here cannot but weaken the denotative value of the individual words, especially where, as in this case, there exist several words each of which (except for the rime) can be sub­ stituted for any of the others.

There is a set of locutions and circumlocutions with which the poets of SB and BC vary their references to the French and their

enemies. Instead of merely referring to them as Saracens or Chris­ tians, they make abundant use of the nouns gent and geste combined with adjectives which either designate the nationality or are suf­ ficiently pejorative as to leave no doubt to whom they apply. Those

indicating mainly nationality are:

Geste:

De la geste Aymari. SB: 5& De la geste Aymeri. BC: iSa, 119a A la geste Aymeri. BC: 3576a

De la geste Mahon. SB: 23^b Vers la geste Pepin. SB: 1533b De la geste grlfaine. BC: 150b De la geste Noiron. , BC: 171b De la geste pullente. BC: 299b Moult sont de franche geste. BC: 822p

Gent sajrrar.inei

De sarrazine gent. SB: 951b De la gent sarrazine. SB: 209Op Por la gent sarrazine. SB: 258?p Vez la gent sarrazine. SB: 2699a Qui sor gent sarrazine. BC: 47a

^ S e e also BC: 583b, 3138b, gent pull ante 67 K*a la gent sarrazine. BC: 131a Contre gent sarrazine, BC: 4§3p, 3926p La gent sarrazinoise. BC: 1362a Sor la gent sarrazine. BC: l451p, 36l9p, 3740p De la gent sarrazine. BC: 1700p, 191%), 3529a Cele gens sarrazine. BC: 180%>, 3781a Car la gent sarrazine. BC: 2036a Cele gent sarrazine. BC: 2111a Quant la gens sarrazine. BC: 2321a A la gent sarrazine. BC: 308%) Car a gent sarrazine. BC: 3562a Qu'il sor gent sarrazine. BC: 3567a Vers la gent sarrazine. BC: 3871p

Gent -paienne (paiennie):

Cele paiene gent. SB: 419b Sitre paienne gent. SB: 1971a

Cele gent paiennie. SB: 691b, 1145b, 1682b Cele gent paiennie. • BC: 1513b Mais de gent paiennie. BC: 1039b Vez ci gent paiennie. BC: 1863b

Ehtre la gent paienne. SB: 822a Et sor la gent paienne. SB: 827a Gent paienne acorue. BC: 1585b

Gent plus name of ancestor or god:

0 la gent Faraon. SB: 242b De la gent Apolin. SB: 472b, 1532b

De la gent Tervagant. BC: 1136b, 3154b, 3710b

Gent crestfenne:

La gens crestlenne est . . . BC: 1224a Nostre gent crest£enne. BC: 1536a

Gent francoise:

Nostre frangoise gent. BC: 553b, 761b, 4368b

D'aidier no gent franjoise. BC: 824a Et com la gens frangoise. BC: 2322a Fu nostre gent frangoise. BC: 3873a 68

Gent de Nerborme:

Que la gent de Nerbonne. BCs 701a De la gent de Nerbonne, BC: 1531a

Gent Coranarchd.se:

Et la gent Conmar6hi.se, BC: 984b Cele gens Conmarchise. BC: 2292b, 2821b, 3785b

The alternate forms gent paienne and gent paiennie are to be explained in terms of the rime, Gent paiennie and paienne gent appear at the rime, but gent paienne is only found in the first hemistich position, as in SB? 822 and 827, This is probably because of the difficulty the jongleur would encounter composing a laisse in -enne (there are, in fact, none in either poem).

All the circumlocutions in this group occur more frequently in BC except the gent paienne type.

The terms of opprobrium fall into four principal groups.

Gent mescreant (mescreue):

De la gent mescreant, SB: l6lb A la gent mescreant. SB: 1463d, 1474b Sor la gent mescreant, BC: 3718b

De la gent mescreue, BC: 919b, 1570b, 1587b, 3928b Cele gent mescreue, BC: 1282b, 2032b

Gent averse (aversiere):

A cele gent averse, SB: 2064b Sor cele gent averse, 2595p

Cele gent aversiere, BC: 408b, 1784b, 2905b, 3471b

De la gent l*aversier. BC: 344b

Gent haie:

La pute gent ha£e. SB: 1676b Vint cele gent hale. SB: 1693b La gent de Dieu haye. BC: 274-3b 69 Gent desfaee:

Ains est gent desfaee. SB: 1162b Cele gent desfaee. BCs 1732b, 2105b, 3001b Vers la gent desfaee, BCs 3880b

In two of the cases from SB It Is the Saracens themselves who use

the pejorative terns to refer to their people. In one instance,

it is true, the speaker, Malatrie, has already given ample evidence

that she is in love with Girart and will eventually became a Chris­

tian, but the juxtaposition of the possessive noz and the unfavor­

able epithet is a strange ones La dedenz a noz tres a cele gent

averse. SBs 2064-, Si the other case (SB: 161) the amirant of

Spain tells his lieutenants to take vint mile de la gent mescreant

and capture Aymeri, There is no logical reason for him to call his

own men unbelievers since they all believe in God and the prophet.

Both these curious uses of pejorative terns seem to be the result

of a tendency on the part of oral poets to employ formulas in places where they are technically correct according to the needs of

versification, but where they make no sense,^ The force of habit

and the mechanism of formulaic composition lead the .jongleur into

an inappropriate combination.

There are other epithets found with gent which do not occur

twice and therefore cannot be considered formulas. Among them are pute. convertie. desreee. malostrue. esperdue and mal aprise. ^

1 7 Jean Itychner, La Chanson de Geste (Geneva: Droz, 1955), pp. 137-38.

18See SB: 531, 830, 881 and 1566, and BC: 403, 1640, 17*4-3, 2048, 2109, 283^7 2970, 3**72, 3897, 3908, 3930 , 3931.’ 70

Although more formulaic epithets of this pejorative type occur in BC (17) than In SB (8), the non-formulaic occurrences In the former poem also outnumber those In the latter by a count of 12 to h. Counting & n epithets used with gent, both those denoting

"pagan" or "Christian" and those which have a purely pejorative

Intent, 65 occur In BC (approximately 16 per 1000 verses) and 23 in SB (about 8 per 1000). Epithets of this type can be said, then, to occur twice as frequently in the later poem than in the earlier one.

Proper and Common Nouns of Place

Names of places are accompanied by adjectival epithets in much the same way as proper names of persons.

Names of countries:

Vers France la vaillant. SB: lh65b D*Espangne la loee. SB: 1576b £h France la loee. BC: 1237b

Names of cities

De Cordres la garaie. SB: 1689b Sarragouce la bele. SB: 2078b De Toulouse la lee. BC: 3516b

The common noun cite:

Sa grant cite louee. SB: 7b La grant cite loee. SB: 35b

Ta bone cite large. SB: 1108b

La fort cite mirable. SB: 1118b

La fort cite entie. SB* 1128b, 15^1b, 1553b, 1683b Cele cite entie. SB: 1283b La fort cite entie. BC* 629b 71 Par la cite garnie. SB: 1135b

&1 la cite vaillant. SB: 1456b La fort cite vaillant. SB: 1503b

La cite seignorie. BC: 1505b, 1840b

La cite bien seans. BC: 2473b

These epithets, like the ones accompanying proper names of p e r s o n s , ^ do not seem to be fixed in such a way that a particular adjective is always used with the same noun.2® Garnie. vaillant and loee can be attached to the name of a city or merely to the word cite.

Since cite only consists of two syllables the jongleur must insert another word, usually an adjective (bone, fort, grant) before it in the verse to fill the necessary number of syllables.

Parts of the Body

Parts of the body are accompanied by epithets, and are used themselves in an attributive way with the preposition a. to describe

characters.

Chiere:

A la chiere hardie. SB: 689b, 1147b, 1259b BC: 622b, 1037b, 1838b, 2711b A la hardie chiere. BC: 1786b

A la chiere meribree. SB: 9b, 16b, 1567b, 1571b BC: 457b, 2110b

^ S e e above, pp. 51-61.

2®Cf. France la garnie. SB: 3 9 H and 4232; Espangne la garnie. SB: 5172; Leride la garnie. SB: 7118; Barbastre la lee. SB: 7212; Bar- bastre la large. SB: 527^7 Nerbonne la large. SB: 5529? Sarragcuce la large. SB: 6177 and 7331. 72

Ch events:

Qui a les ch evens blons. BC: 715b, 2564b, 3603b

Cuer j

Qui ot le cuer mari. BC: 598b, 1473b Eh ont les cuers maris. BC: 2793b

Qui n*a pas cuer d*enfant. BC: 684b, 2443b

Hon fill au cuer sene. BC: 1660b Gerart au cuer sene. BC: 2058b

Bueves au cuer hardi. BC: 2126b

Qui n*ot pas cuer lanier. BC: 2875b Trop aroit cuer lanier. BC: 3432b

Le cuer triste et ire. BC: 2935b De cuer triste et iree. BC: 2973b

S'en ai le cuer dolent. BC: 3 H 3 b De cuer triste et dolante. BC: 2 2 ® , 3135b

There are other adjectives found with cuer. but they do not appear to be formulaic.^ Corage (in Old French synonymous with cuer) is accompanied by fier (Mfierce«»):

Hernaut au fier oorage. SB; 1120b 0 le corage fier. SB: 2655b Gerart au fier corage. BC: 3544b

The word cuer is not absent from SB. but it does not occur in connection with a formulaic epithet in the section before verse

2696.

Horses

Horses' too have their epithets.

2L5C2£i, BC: 30; sachant. BC: 683; esmeu. BC: 1267; hardl. BC: 2126; tres hardi et ose. BC: 2943. 73 Chgyal;

Le bon cheval corsier. SB: 2279b

Destrier:

Au bon corant destrier. SB: 95a De lor coranz destriers. SB: 2122b Seur un corant destrier. BC: 1758b, 2837a St sor corant destrier. BC* 2358b Le bon corant destrier. BC: 2845b

Tant bon destrier corant. SB: 190b Sor un destrier corant. SB: 1611b Sor les destriers corans. BC: 89a Et sist sor un destrier corant conme levriere. BC: 2383 Son bon destrier corant. BC: 2420a Sor un destrier corant. BC: 2907a Lor bons destriers corans. BC: 3663a

Ses bons destriers crerxuz. SB: 297b LI bon destrier crern. BC: 1551b Sor les destriers creams. BC: 193** Sor un destrier cream, BC: 2007b

De l ’auferrant destrier. SB: 2031b tJn auferrant destrier. SB: 2257b

Sor tm destrier armez. SB: 1389b Sor son destrier armez. S3: 1939b

Vees com ses destriers par est tres abrlves. SB: 2 5 H

Desor le vair destrier. SB: 209b

Lore prist le bon destrier, BC: 2707a

Del destrier arrabl. SB: 2200b Le destrier arrabl. BC: 1477b

Ce destrier arragon. SB: 2004b Des destriers arragons. SB: 2505b Mains destriers arragons. BC: 3614b

Del destrier de Sulie. SB: 1696b Un destrier de Surie. BC: 2168b

Le destrier de Castele. SB: 2050b BC: 2653b Chascuns ot bon destrier, fort et isnel et sain* BC: 114 Chascuns ot bon destrier, qui ne va mie lent* BC: 259 Et sist sor un destrier qui vint de Cateloingne. BC: 329

Auferrant; ^

Son auferrant gascon. OT: 127b De 1'auferrant prisie.^ SB: 2146b Son auferrant corsier* OT: 2272b

Mule:

Et le mul sullen. SB: 2085a Sor le nrur afeutre. OT: 2226b Sor le mur sullen. OT: 2280b Sor la mule coursiere. BC: 2391b

Of these epithets for horses 23 are from SB. 20 from BC.

The epithets found with mentions of horses in the chansons de geste from the Roland to Buevon de Conmarchis have been studied by

M* Jean Frappier in a paper read at the Colloque de liege: "Les

Destriers et leurs Epithetes. He draws several conclusions from his research: (1) all the epithets are adjectives of praise and ornament contributing to the aura of idealization that one finds throughout the Old French epic; (2) they are part of the style of the oral epic; (3) the number of epithets grows as the dates of the works approach us, with the rimed poems generally containing more than the Roland, the faiiUamne or Gormont et Isembart. In regard to this last point, Frappier distinguishes three periods in

^ I n addition to the examples given under destrier above, p. 73*

^Ehis adjective is also used in reference to men above, p. 55*

Technique litteraire des Chansons de Geste: Actes du Col- loque~de liege (Septembre 1957) YParis: Societe d*Editions "Les BellesLettres," 1959), pp. 85-104. 75 the evolution of the epithets for horses) the first is comprised of the three chansons mentioned above; the second includes Cou- ronnement de Louis. Guillaume U (sometimes called the Chanson de

Rainouard). Charroi de K$mes. Moniage Guillaume I and XI, Prise de Cordrea et de Seh-ile. Prise d*Orange and other epics of the early and middle twelfth century; the third period begins with 25 Aliscans (usually dated around 1165) and goes on up to Buevon de Conmarchis. which is chronologically the last poem treated in the article. This last period would include Le Siege de Barbastre. although this poem does not figure in Frappier*s findings. He does mention, however, several characteristics which are similar to those of the epithets on our list, such as the alternation of destrier auferrant and auferrant destrier, and of corant destrier and destrier corant according to the needs of rime. In general his findings are in accord with ours: there are lh different terms in the epithets from SB. 12 in those from BC. This difference is not great enough to be conclusive, but it does bear out Frappier* s hypothesis in that it shows that the relative frequency of formulas in the poems of the last period is at about the same level. Fur­ ther remarks about the formulaic nature of these epithets will be made at the end of this chapter.

Armor and Articles of Dress

The knight* s annor is habitually described or qualified with

^See Raphael Levy, Chronologie apprr»riniative de la Litterature francaise du moven age (Tubingen: Niemeyer, 1957), P« 1^. 76 epithets* The following are the epithets of articles of dress and armor worn by the knight in battle:

Escu:

En l'escu de son ool. SB: 269a, 1209a Sor l'escu de son col* SB: 2024a Que l'escu de son col. SB: 2148a

Les escuz d'azur bis* SB: 321b Devant en l'escu bis* SB: 2193b

Et les escuz dorez. SB: 733b Sor les escuz dorez* SB: 1429b

Cest escu qu'est bendez* SB: 754b Sor l'escu d'or bendes. BC: 1899b

De sor l'escu vostiz* SB: 1315b Le for escu vosti* SB: 1318b

Un escu de cartier* SB: 1848b En l'escu de cartier. SB: 2018b Sor l'escu de cartier. SB: 2273b, 2277b

Les escuz a lion. SB: 2504b Maint escu a lion. BC: 191b, 3616b

Et un escu liste. SB: 748b A son col son escu qui d'or estoit listez. SB: 1916 Et dens escus listes. BC: 1077b Et mon escu liste. BC: 2702a Sor son escu liste. BC: 2929b

Sor les escus boudes. BC: 1896b

Es escus vemissies. BC: 2599b

En l'escu qui flambie. BC: 2750b

Et mon escu pesant. BC: 2500a

Ses fors escus voltis. - BC: 2775b Lor fors escus voltis. BC: 3769b

Targe:

Sor la targe florie. SB: 126lb, 1276b, 1288b 77 Et mainte pesant targe. BC: 192a

Et hiaume et targe bonne. BC: 869a Hiaumes no targe bise. BC: 991b Et mainte targe bise. BC: 3783b

Sor la targe listee. BC: 1722b

Mainte targe doree. BC:2108b II • •

Li verz hiaumes jumez. SB: 273b Et les hiaumes a or. SB: 733a L'iaume lacie el chief, qui est a or jume. SB: 94-3 Ehtor l'iaume jeme. SB: 944b A cel hiaume geme. SB: 2540b Et deus hiaumes gemes. BC: 1075b Et 11 hiaumes geme s. BC: 1876b Ses vers hiaumes geme s. BC: 2514b

Et oest hiaume dorez. SB: 755b Et 11 hiaumes dorez. SB: 1949b Corsolt feri sor l'iaume a la cercle doree. BC: 1727 Et de hiaume dore a pierres de vertu,I BC: 2005

S'a veu l'iaume cler. SB: 2527b

Et hiaumes de Pavie, BC: 1044b

Et bon elme luisant. BC: 1156b

Maint escu et maint hiaume ou li ors reflambie. BC: 1519

Amont deseur son hiaume qui a or fu brunis. BC: 1606 A vers hiaumes brunis. BC: 2277b Mon vert elme bumi. BC: 2701b Onques ses hiaumes bruns. BC: 2756a

Hauberc:

Son hauberc fremillon. SB: 252b Un hauberc fremillon. SB: 1605b Le hauberc fremillon. SB: 2503b Maint hauberc fremillon. BC: 194b Mains haubers fremillons. BC: 3591b

Li blans haubers safrez. SB: 274b Ces blans haubers safrez. SB; 732b Un blanc hauberc safre. SB: 74-7b 78

Les blans haubers vestuz. SB: 307b Un blanc hauberc doblier. SB: 1846b A son blanc hauberc trere. SB: 2059b 11 blans haubers vastus. BC: 1932b

L'auberc vosrtu el dos, qui fu menu ovre. SB: 942b

Desoz l'auberc tresliz. SB: 1296b

De l'auberc jazerant. SB: 1458b Maint hauberc jaserant. BC: 676b

Brolngne:

Mainte broingne doubliere. BC: 422b, 1793b La forte broigne endossee. BC: 868b

Bliaut:

EL bliaut de cartier. SB: 385b Un bliaut de cartier. SB: 1808b

Un bliaut d'Inde chier. SB: 1807b

Pieces of weaponry include:

Hanste (wooden part of the lance):

En son poing tint la hanste dont li fers fu dores, BC: 1877 Et a brandie l'anste dont li fers fu quarres. BC: 1895 De vostre anste aceree. BC: 2993b

Bspiel (spear):

D'un fort trenchant espie. SB: 2651b Tint l'espiel noele. BC: 43** Un viel espiel i trueve, qui mal estoit fourbis. BC: 948 Et un espiel qui ert trenchans et aceres. BC: IO76 El poing l'espiel plane. BC: 1695b Et d*espiel esmolu. BC: 2004b Et a brand! l'espiel au fer trenchant agu. BC: 2012 Escru ot et espiel a fer qui fu trenchans. BC: 2474 Puis saisi son espiel dont li fers fu quarres. BC: 2531 De lor espiels burais. BC: 2774 De deus espiels trenchans dont chascuns fu massis. BC: 2778

Branc:

11 a trete l'espee dont li branz fu letrez. SB: 256 Si a trete l'espee dont li brans fu letrez. SB: 1371 79 Et tenise en mon pong tin branc d'acier letrez. SB: 719 Li quens tint en son poing le bran d'acier lettre. BC* 439 Je les convoierai au bran d'acier letre. BC: 1685

Meint branc d*acier foxbi. SB: 734a Au branc forbi d'acier. SB: 907b ‘ Un branc foxbi d'acier. SB: 1847b Del branc foxbi d'acier. SB: al|8 8 b Es poinz les branz foxbiz. SB: 1246b Le bran foxbi d'acier. BC: 362b Et tint le branc foxbi. SB: 1317b

Ses brans ert mal fouxbis. BC: 513b Guillaumes tint le bran, qui estoit mal fouxbis. BC: 1592 As brans fouxbis se vont asprement acointier. BC: 2859

Si trest le branc d'acier. SB: 2157b Dou bran d'acier le cuide . . . BC: 413a As brans d'acier i ot . . , BC: 1565a, 1581a Ou poing le bran d'acier. BC: 3791a Ou en son poing le bran d'acier trenchant tout nu. BC: 3906

0 les brans acerez. SB: 939b A noz brans acex*ez. SB: 1419b

Des bons brans acerins. SB: 887a

0 les brans v£ennois. SB: 900b

Guillaume tint l'espee, dont li brans fu moluz. SB: 293

gainst chascun le bran cler. BC: 71b

Pendoit li riches brans. BC: 1962b

Espee:

As espees tranehanz. SB: 1224a Et espee trenchant. BC: 1157b Et a saisi l'espee qui fu bonne et tranchans. BC: 1965 Et d* espee txrenchant. BC: 2004b As espees trenchans. BC: 2575a

A la trenchant espee. SB: 1575b De vo trenchant espee. BC: 424a

A l'espee forbie. SB: 1544b A l'espee fouxbi. BC: 638b Mainte espee fouxbi. BC: 1045b 80

Hist la mein a 1*espee, dont li poinz fu dorez. SB: 1933 A l'espee d'acier. SB* 2320b

Et espee aoeree. BC: 473b

Ja sara se m*espee fu de maistre forgie. BC* 2740

To these we add a passage from BC In vhich Gerars describes his sword.

Pensons de l'esploitier, car moult foment me tarde Que sane de Sarrazins pam i les chans esparde A m* espee qui pas de petit cops n'escarde. Donna le moi Guillaumes, faite fu a Bobarde. Bel puet tenir haubers ne c*uns poules c'on larde Tant est de bonne forge. (BC* 3538-43)

Trappings and accessories for riding into battle:

Esperont

D'or sonb li esperon. SB: 1604b Des esporons d'or mier. SB* 2015b Des esperons trenchans. BC* 212a Des esperons a or. BC* 436a Des trenchans esperons. BC* 2586b

Sele:

Qui ot sele doree. SB* 1178b 0 la sele orfresie. BC* 2173b La sele erb de fin or. BC* 2392a

Range:

Par la range a or mier. SB* 1867b Par la range d'or fin. SB* 2195b

Resne:

Par la resne doree. SB* 1170b Par la resne d'or mier. SB* 2046b Par la resne a or mier. BC* 2876b Les resnes furent totes ovrees a or mier. SB: 2265

Par les resnes noez. 1334b, 1445b, 1451b

Par la resne doubliere. BC: 4llb 81

Confanon:

Le confanon de sole fet au vant baloier, SB: 1872 Le confanon de sole fit au vant baloier. SB: 1883 Le confanon de sole v± au vant deploier. SB: 2267

Son vermeil confanon. SB: 117b Au vermeil confanon. SB: 1609b, 1995b

A un gonfanon d ’or. BC: 2373® Gonfanon 1 ot riche d'or qui fu flamboians. BC: 2475 Gonfanon 1 ot riche a cinc clos d'or fennes, BC: 2532

Siglaton:

Maint vermeil siglaton. BC: 193b Son vermeil siglaton. BC: 3340b

Ensalgne:

L'ensaigne au dragon vlrent ou 11 ors reflambie. BC: 1862 S'ensalgne flambolant. BC: 2423b Et l'ensaigne luisant. BC: 2436b

Of the epithets for armor, weapons, clothing and accessories for battle, 78 a**e from SB and 88 from B C .

BiH lriings

The following epithets accompany nouns connected with build­ ings of any kind: palaces, castles, tents; and also parts of these edifices and their contents.

Pales:

En ce pales plenier. SB: 630b Par ce pales plenier, SB: 902b De ce pales plenier. SB: 2112b Lez un pales plenier. SB: 2473b

El palais seignori. BC: 1812b 82

Chastel:

Le chastel seignori. BC: l^+68b, 2156b Le chastel bien seant. BC: 3695b

Tre:

Fu. en son tre de sole. BC: 3167a Dedens le tre de sole. BC: 3661a

T or:

Et en la tor antle. SB: 702a Dedens la tour antle. BC: 103*ft>

Desor la tor car re e. SB: 1163b

Ne n*i a tour nesoit faite de pierre bise. BC: 2296

Sale:

En la sale pavee. BC: 2985b

Cordes (of a tent) and peson (weights);

Les cordes sont de sole et 11 peson d'araine. SB: 72 Les cordes sont de sole, et 11 peson d'alier. SB: 1795 Et les cordes de sole. SB: 1463a

Fenestras:

As fenestres de marbre s*est alez apoier. SB: 1876 As fenestres de mart>re m’en alai acouter. BC: 3015 Fu a vine fenestre faite de pierre bise. BC: 537

Faudestuel: *.

Sor un faudestuel d #or, que ainz ne vi meillor. SB: 1006 Sor un faudestuel d*or, einz tel ne vit nus hon, SB: 1013

Twelve of chese are from SB. 9 from BC. Plenier is used here with pales, but elsewhere with estor (SB: 233*0 and pais (SB: 1835)*

Ob.jects in Nature

The qualities of natural objects are intensified by epithets 83 Elbe:

Gerars descent sor l’erbe qui estoit verdoians. BC: 196^ Doucanent la descent sor l*erbe verdoiant. BC: 2422

Pre:

Par mi le pre flori. SB: 1298b Par mi ce pre flori, BC: 2698b Farmi les pres floris, BC: 3772b

EJsni les pres herbus. BC: 1931b Par mi le pre herbu. BC: 2001b, 3913b

Ehs el pre verdoiant, BC: 3395b, 3713b

Olivier:

Soz 1*olivier rame. SB: 2526b Souz 1*olivier rame, BC* 2952b

Forest:

Les la forest antie. BC: 1859b

Under this category will also be included building materials, whether they are used to denote the natural state or the finished product,

Roche:

Siet sor roche nayue. BC: 2025b

Qui siet seur roche bise, BC: 1647b Les une roche bise, BC: 2835b

Pierre:

Faite de pierre bise, BC: 537b, 2296b

Marbre:

Sor le marbre liste, BC: 1669b Deseur le marbre bis, BC: 971b As murs de marbre bis. BC; 1619b 84

Areinne (bronze):

Des raurs d'areime bis, SB: 323b, 1247b

Thus bis (gray) is found modifying roche. -pierre. marbre. escu and aajur.26 Verdoiant can modify erbe or pre. It seems, then, that these adjectives are not linked with their nouns in fixed phrases, but rather that the .jongleur, knowing that bis and verdoiant are convenient words to keep in mind for laisses which rime in -is and -ant, also keeps a store of nouns with which they can be readily combined.

Others

There are two sets of epithets of high frequency which may be treated together since they are practically identical in meaning, and which are applied indiscriminately to persons and to objects,

Prisier:

Con fetes a prisier! SB: 376b

Qui firent a prisier. SB: l66lb

Font foment a prisier. SB: 1797b

Qui molt font a prisier. SB: 1799b Qui moult font a prisier. BC: 2335b

Qui tant fist a proisier. SB: 1850b, 2095b Qui tant fet a prisier. SB: 1893b, 2479b Qui tant fait a prisier. BC: 3428b

Qui si fet a prisier. SB: 1882b

Qui tant i soit prisiez. SB: 2298b

'or the last two, see above, p. 76. Cui chascuns doit prisier. BC: 375b

Qui sont de moult haut pris. BC:

Qui ont pris fineoient. BC: 562b

Qu*on deust prisier. BC: 1756b

Loer:

Qui molt font a loer. SB: 1595b Qui molt fet a loer. SB: 1710b Qui moult font a loer, BC: 60b

Con par fet a loer! SB: 1738b

Qui tant fist a loer. SB: 2219b Qui tant font a loer. SB: 2557b

Molt fetes a loer. SB: 25^5b

The first of these two groups contains rimes in -ier. -iez, -is

and -ent. the second in -er. Since -er and -ier are among the

most frequent rimes in the two poems, these phrases were extremely

useful to the poet.

But in addition to their usefulness from a technical point of

•view, it will be noted that they are, in another sense, almost

Unnecessary. The audience of these chansons de geste knows, for

it has been told many times from the very beginning of each poem,

that the French knights and ladies perform actions which are worthy

of praise. Certainly such phrases as qui tant fet a prisier and qui

tant fist a loer do not add much to the audience*s knowledge of

the characters. In the passage

Voiz con riche presant t’est ci abandons, Con de filles a roi qui tant font a loer. (SB: 2556-57)

the last hemistich of the second verse could be left out with but little consequence. It is not the needs of meaning, of fur­ therance of the plot, that the .jongleur is satisfying with this

epithet, but only the technical requirements of the alexandrine verse and the riming laisse.

In judging ■which of the two formulas, qui tant fet a prisier

or qui tant font a loer. he should insert, the only criterion the

.jongleur uses is that of rime. If the laisse had rimed in -e he

could just as well have used, qui ont tant de biaute. if in -ent. qui ont pris finement might well have been the choice. For epi­ thets in SB and BC are, in general, employed without regard to their precise meaning. No one epithet (with the exceptions of

On i n aiime au cort nes and Malatrie la bele) is attached exclusively to one person, place or object. The very interchangeableness of these expressions is what makes them valuable to a jongleur com­ posing oral poetzy.

Let us take the opening verses of SB as an example.

Plest vos oir chanson bien faite et conpasee? Toute est de vielle estoire estraite et porpansee. Molt fait bien a o£r, pieja ne fu contee; Tote est de la lingniee que Deus ot tant amee, De la geste Aymeri qui prouece a duree. Ce fu a Pantecoste, cele feste ennoree: Li cuens fu a Nerbone, sa grant cite louee. (SB: 1-7)

The last three b hemistichs are purely ornamental: that is they

are not necessary for the development of the plot of the poem.

What they do contain is not the meat of the tale, not the story-line, but rather elements of the tone that the jongleur is creating around his plot. As Frappier has mentioned, it is an aura of idealization 87 and of the perfection of society (Christian society, that is) from the noblest of kings to the inanimate water of fountains qui moult font a prisier (BC: 2335).27

Since the epithet-hemistichs do not carry any important elements of meaning, they are all the more capable of being interchanged one for the other. One indication of this can be found in the variants to Perrier1 s edition of the Siege de Bar­ ba stre (pp, 233-268) •which takes ms, C as its base. For verse 5» ms, D has the variant a la chiere membrea. a phrase which has nothing to do with the nrouece of ms, C but which means roughly the same thing as an epithet of praise, What possible reason could a scribe have for replacing the one reading with the other? The answer does not lie with the scribe, but rather with the method of oral composition. The two epithets are actually two different means, chosen by two singers whose poems are reflected in these manuscripts, of solving the same technical problems, those of producing verses at rapidly succeeding intervals and of slowing down the narrative in order to have more time to dig into the stock resources for the following verses, "Which particular epithet each poet chooses is a matter of little import for him, as long as he has one with the right rime at his disposal,

X further look at the table of variants from the Perrier edition of SB supports this conclusion. The following list contains

^See above, pp, 7^-75. 88 epithets from the C version and their variant forms from the other manuscripts.

c Por amor Deru le grants m. Por Bleu le realment. v. 199 c D'Qrenge le guerrier. £ Au cort nes, au vis fier. V. 208 c Qui meint en temite. AB£ Le roi de majeste. V. 596

c Le pere droiturier. D Le verai jostisier. V. 619

C Deus garjons losengier. D Deus garyons paltonier. V. 631

C En la sale votie. D Sor les murs de Persie. V. 697

C Le rois de majeste®. DAB Qui an croiz fu penez. V. 717

C Et les escuz dorez. DABP Et les escuz listez. V. 733

C Cest escu qu’est bendez. D Cost fort escu liste. V. 75^

C Et cest hiaume dorez. m. Et cest hiaume jane. V. 755 c Nul felon losengier. D Nul quivert pautonier. V. 910

C Ehtor l'iaume jerae. D Dont li hiame ert ovrez. V. 944

C Li culvert mescreant. D 11 culvert souduiant. V. 989

C Qui fist parler l'imaje. D Qui m'a fet a s'imaje. ABP Qui nous fist a s'imaje. V. 1122

C Par la resne doree. DABP Par la resne doblee. V. 1170 89 c Qui tant par est doutez. • D Qui si est redoutez. v. 1198

C Sor la targe florle. I De la lance brunie. v. 1276 c Cele cite entle. £ La fort cite garnie. v. 1283 c Le felon Maladi. D 11 quivers Maladis. v. 1295

C Sor les escus dorez* D Sor les escuz listez. v. 136 6 c Dont U brans fu letrez. £ Qui 11 pant au coutez. v. 1371 c Sor les escuz dorez. D Sor les escuz listez. v. 1429 c A la gent mescreant. D De France la vaillant. v. 1466

C Malatrie la grant. D A1 gent cors avenant. v. 1501

C La fort cite vaillant. I La fort cite manent. v. 1503 c A la trenohairt espee. m . A la chiere manbree. v. 1575 c Qui se lessa paner. D Qui an croiz fu penez. v. 1594

C Cele gent hale. D La gent de Persie. V. 1693

C Le verai jotissier. D Qui tot a a jugier. v. 1818

C Et le pals plenier. DABP Et le pales plenier. v. 1835

C Qui tout a a Jugier. DABP Qui tout a a baillier. v. 1891

C Franc chevalier oneste. DABP Chevalier de bon ere. v. 2062 90 c Qui proece goveme. 1 0 proece converse. v. 2067

c Franc chevalier oneste. m Frans horn de bone jeste. v. 2069 c Sarrazin losengier. 1 e Li paien aversier. v. 2110

c Sarrazin losengier. D Li gloton aversier. v. 2119

C De 1 * auf errant prisie. D De l*auferrant corsier. v. 2146

C Del destrier arrabi. DE Do bon destrier de pris. v. 2200

C li felon losangier. DP li quivert losangier. v. 2284

C Malatrie au vis cler. DEABP Au gent cors anore. v. 2340

C Et Geufroi l*alose. D Et Hunaut le meribre. v. 2399

C A cel hiaume geme. DE Qui si vient abrivez. v. 2540

C Ii cortois et le fier. D Qui bien sot le santier. v. 2609

It is evident that the epithot-heniistichs form, as it were, lacunae in the texture of the epic poem that the .jongleur can fill in according to the stock phrases which belong to his individual repertory. Thus the singer of D sometimes inserted ouivert where

C has felon, lists where C has dorez. And the above list of variants is, according to Perrier, only a selection.

In conclusion one can say that an epithet in the chansons de geste is an adjective, noun in apposition, prepositional phrase or relative clause, used to meet the needs of rhythm and rime, but relatively unimportant as a vehicle of plot and only loosely linked in meaning with the noun it accompanies.

Four hundred and twenty-one of the epithets treated in this chapter can be said to be fomrulaic since they are found more than once in substantially the same form in one cf the two poems. Of this number 2.9k are from SB (approximately 71 per 1000 verses) and 227 from BC (approximately 58 por 1000). Thus as far as epi­ thets are concerned SB is more formulaic than BC. CHAPTER IV

PROPER NAMES

The technique of formulaic composition may be looked upon as a tray of telling stories to audiences, without a written text, by filling up verses with phrases of the proper length. The poetic form of the chansons de geste requires the .jongleur to make all his syntactic structures adaptable to the hemistich or the whole verse. None may exceed this length or fall short of it, or his art fails in this respect. This rule makes it neces­ sary that even the names of people, places and nationalities or tribes be poured into the metrical mold.

It is for this reason that the names of many of the characters of Le Siege de Barbastre and Buevon de Conmarchis fit so neatly into the six-syllable hemistich. The following pages deal with the forms of proper names which occupy a hemistich either by themselves or in conjunction with other names, titles or desig­ nations of relationship.

The French and Saracen heroes have names which are composed of a forename plus the name of their place of birth or of a town which they have conquered, and the combination of these words fills a hemistich of alexandrine verse.

92 93.

French:

Buevon de Conmarchis. SB: 10a, 174a, 209a, 310a, 312b, 388a, 422a, 515a, 570a, 6l8a 833a, 1107a, 1219a, 1735a, 1896a, 2427a BC: 500b, 928b

Entdaume d*Avdgnon. SB: 108b

Render de Montarmer. SB: 110a, 176a, 316a Render de Montaraider. SB: 2482b? 2521a, 2601b Renaut de Montaxmder. BC: 66a, 361b, 488a, 1664a, 2273a, 2872b 3438b, 3482a, 3508a, 3635a, 3941a

Navards de Toulouse. SB: 112a, 177a, 214a, 314a, 690a,' 705a, 884a BC: 369a, 1662a, 2061a, 3214a, 3332a, 3364a, 3638a

Gilbert de Terrascon, SB: 112b, 210a

Aymeri de Nerbone. SB: 205a, 258a, 708a BC: 220a, 276a, 452a, 457a, 818b, 1586a, 1627a, 1772a, 3367b, 3418a

Broon de Bargelune, SB: 213a

Amaurris de Sassodngne. SB: 324b

Gdrart del Conmarchis. S3: 1723a, 1874a, 1898a, 2128a, 2453a, i' 2461a, 2689a, 2698a BC: 115a, 164a, 2875a, 3200a, 3308a, 3417a, 3485a, 3624a, 3657a

Sauvard de Toulouse. SB: 2602a

Emengars de Pavde. BC: 140a, 319a, 597a, 630b, 1530b

Amaurrd de Senlds. BC: 292a

Bernart de Hondddder. BC: 366b

Saracens:

Corsout de Tabarde. SB: 160a, 287a, 393a, 433a, 453a, 468a, 511a, 796a, 1104a, 1552a, 1697b, 2125a, 2147a, 2583a, 2618a, 2626a Corsolt de Tabarde, BC: 290a, 620b, 730a, 738a, 1166a, 1512b, 1832b, 2179b, 2725b Clarion de Vaudune. 579a, 651a, 701a, 837a, 1370a, 1450a, 2168a Clarion de Valdonne, BC: 812b Clarion de Valdune. BC: 891a, 997a 1247a, 1382a, 1402a, 1414a, 1424a, 1485a, 3493a

Otran de Baleguez. SB: 1352b

Malaquin de Susane. SB: 1521a Malaquin de Sulie. SB: 2371a Malaquin de Tudele. BC: 3322a, 3586a, 3622a, 3642a, 3747b

Fabur de derlion. SB: 1606b

Salatre d,Allsandre. SB: 1623a

Coruleas de Noble. SB: 1624a

Llbanor de Tumie. SB: 1986a, 2349a, 2673a Ldmbanor de Tudele. BC: 2167a, 3035a

Espinart d'Avaloingne, BC: 328b

Brunars d'Orcanie. BC: 2180b

The Saracen forenames are largely imaginative and, at least In the cases of Brunars ("Browny"), Espinart ("Spinach") and Salatre

("Salty"), humorous. The toponyms reflect the general Ignorance of the jongleurs concerning the geography of most of idxat lay on the other side of the Pyrenees, although Tudele (Tudela), Baleguez

(Balaguer) and Alisandre (Alexandria) are real cities. Clerllon is evidently meant to have an exotic ring about it, and Avaloingne may be the enchanted land of Avalon to which King Arthur was pur­ ported to have gone to be cured of his wounds. Note that Adenet has changed the toponym "Tumie" to "Tudele,"-and has attached it to the names of both Malaquin and ldnbanor. Malaquin de Susane is the same person as Malaquin de Sulie ("Syria"); this is a dis­ crepancy which a literary poet would no doubt have corrected upon 95 re-reading his text, “Valdune" is written "Valdonne" by Adenet the one time it appears at the rime, no doubt because it would have been impossible to find adequate rimes in -Tine.

Another way of making a name fill a hemistich is to have the plaoe of origin indicated by an adjective,

Gautier li Toulousanz. SB: 179®, 1925a, 2095® Li Toulousanz Gautier, SB: 210b Gautier li Toulousanz, BC: 2119a, 2477b, 3715®, 3795®, 3937® Joffrois li Angevins. BC: 459®, 2450a Li Toulousanz Gautier, BC: 3275b

Gautier li Toulousanz is a full hemistich form of the name Gautier de Toulouse, which appears in BC but not in SB.

Whan the hero's name is not long enough to fill a hemistich even with the addition of the name of his place of origin, the verse is "stuffed" with another word, usually a co-ordinating con­ junction, but sometimes a preposition, a subordinating conjunction, or even a verb,

Et Guillaumes d'Orenge. SB: 17®, 173®, 292a, 325®, 1545a BC: 175®, 552®, 672a, 721a De Guillaume d'Orenge, BC: 648a Quant Guillaumes d'Orenge. BC: 306a

Et Geufroi l'Ehgevin. SB: 111a, 178a, 212a, 315b, 1337®, 2399®, 2446a, 2522a Et Joffrois 1'Angevin. BC: 165a, 2792b, 3213®, 3331®, 3409a, 3588a, 3636a Ne Joffroi 1'Angevin. BC: 1663a

Et Hunaut le Breton. SB: 111b, 180a, 315a, 1337b, 2466b Es Hunaut de Bretaigne. BC: 462a Et Hunaut de Bretaigne. BC: 1819®, 1955®, 2060a, 2272a, 2451a, 3028a, 3300a, 3637®, 3799® Ne Hunaut de Bretaigne, BC: 3215®

Ne Garin d'Anseune. SB; 709® Et Garins d'Anseune. BC: 683a Ne Bernart de Brabant, SB: 7Ha Et Bemars do Brabant. BC: 682b

Hit Salot de Baudas. SB: 2193a Fiort Salot de Baudas, BC: 437a

Et Sanson de Boorges. BC: 368a, I8l6a

Et Emaus de Biaulande, BC: 684a

Et Huon le Hansel. BC: 983a

De Gui de Conmarchis. BC: 1924a Quant Guls de Corsnar-chis. BC: 2010a, 3085a, 3477a Et Gul de Conraarahis. BC: 2059a Guls de Conmarchis broche . . . BC: 2845a A Gui de Conmarchis, BC: 3716a Guis de Conmarchis tint • * , BC: 3934a

Quant Gautiers de Tolouse. _ BC: 2125a Vint Gautiers de Toulouse, BC: 2785s- Dist Gautiers de Toulouse, BC: 3915s-

Thus if there is some need to add a conjunction to the hemistich

■which contains the name of Joffrois 11 Angevins, this name is written

Joffrois 1'Angevins, the elision of the article freeing one syllable for the inclusion of the extra word. These alternate forms for names which occur so frequently are very useful to the Jongleur

since they make metrical compensation a simple matter. They may also simplify the task of riming: the name Guillaume d*0renge

occurs only at the caesura, and the form found in the second- hemistich position is Guillaume au cort nes, which is much’ easier to rime than the word Orenge.

Short proper names may be mentioned two to a hemistich. They may be compound subjects or objects, and are joined by et:

Berangier et Sanson. SB: 109b, 238b 97 Galdon at Berangier. SB: 211b Gaidon et Berangier. BC: 371b

Savarl et Richier. SB: 214b

Girart et Guielin. SB: 232a, 313b, 385a, 423a, 663a, 883a, 1049a, 1120a, 1551a, 1572a Gerart et Guielin. BC: 60a, 64a, 113a, 258* 277a, 446a, 3587a

Golias et Gaudin. SB; 471b, 1499a

Navari et Hunaut. SB: 780a

Guielin et Buevon. SB: 1336b

Truaquin et Truqant. SB: 1498b

Guielin et Girarz. S3: 1673b

Girart et Malatrie. SB: 2566a Gerars et Malatrie. BC: 3693a

Aymeris et Guillaumes. BC: 30a, 631a, 1550a, 1752a, 1797a

Sanson et Berangier. BC: 394b

Sabinete et ELandrine. BC: 583a

Joffroi et Navari* BC: 1027b, 1817b, 2272b

Renaus et Navaris. BC: 2449b, 3589a

Hunaus et Navaris. BC: 2792b

Bueves et Navaris. BC: 2874a

Sormarinde et Sarete. BC: 3218a

Rustin et Malquidant. BC: 37Hb, 3822a

If the two names add up to only four syllables, ©t . . . et is the conjunction.

Et Poin§on et Beraut. SB: 211a Et Poinces et Berauz. SB: 317a Et Pon9on et Berart. BC: 370a

Et Gaidon et Garin. SB: 317b I

98

Et Gaunor et Gaudin, SB: 452a

Et Antiaume et Beraut. ■SB: 539b

Et Guillaume et Buevon. BC: 197b

Et Rainfroi et Gautier. BC: 367b

Et Joffroi et Hunaut. BC: 984a

Et Gerart et Guion. BC: 1529a

Rarely, in BC. three names are grouped together.

Bueves, Gerars et Guis. BC: 883a Buevon, Gerart ne Gui. BC: 1005b

Renaus, Joffrois et Guis. BC: 3798a

In three cases the names which share the hemistich are not syntac­ tically parallel, but rather one is the subject of the verb and the other the object.

Bueves del Conmarchis lor a ocis Jonas, Et Guielin Mauprin, et Geufroi Golias. (SB: 1219-20)

Aymeris Ermengart a perdre moult ressoingne. BC: 3Z3&

Many of the names and pairs of names -which occupy entire hemistichs occur in groups of several verses, such as the ones shown in the following examples.

A tant es vos Guillaumes d*0renge le guerrier, Buevon del Conmarchis desor le vair destrier, Gilbert de Terrasconne, li Toulosanz Gautier, Et Poinjon et Beraut, Gaidon et Berangier, Et Geufroi l,Ehgevin, Fouquere le legier, Droon de Bargelune et le conte Eugelier, Navari de Toulouse, Savari et Richier. Bien furent trois cenz Franc a l*estor conmencier, Icil corent ensenble por paien domagier. (SB: 208-16)

Malaquins de Tudele fiert ens a esperons, Gerars et Guielins le sivent as talons Et Joffrois 1*Angevins et Hunaus le Bretons, Renaus et Navaris, chascuns fu moult preudons. (BC: 3586-89) 99 These passages constitute short catalogues of the French heroes and are usually found at the beginning of the battle scenes, as a sort of dramatis personae.

These names or combinations of names ■which take up an entire hemistich are by no means the only designations 'which occur in the poems, as many hemistichs contain both the proper noun subject and the verb, and sometimes even an adverb. But only in rare cases ~vri.ll a proper name composed of a forename and a place-name be partitioned into two hemistichs. The examples below are the only ones 'which occur in the two poems.

A tant es vos Guillaumes d'Orenge le guerrler. SB; 208

Avez vos pris Girart del Conmarchis le fier? SB: 2627 Parlor au due Girart del Conmarchis le ber. SB: 2385 Ice dites Girart del Conmarchis le ber. SB; 2395

Vous m * avez Li.mbanor de Tudele donne. B C : 29^8

A ce conseil fu Guis de Conmarchis huchies. BC: 331^

Lors apela Joffroi 1*Angevin tout errant. BC: 3383

Apres Joffroi Hunaut de Bretaigne apela. BC: 31*08

K'avoec iroit Gautiers de Tolouse la lee. BC: 3516

La porte on a Fouchier de Chartres conmandee. BC; 3519

The verses in this list from SB are all according to the same pattern

(place-name plus epithet) in the second hemistich, whereas Adenet*s

are of varying construction. Thus even this exception to the normal way of doing things is more strictly systematized in the anonymous

poem. ( 100

In addition to the names of heroes, the Saracen gods appear

tiro to a hemistich,

Maham et Tervag&nt. SB: 148b, 591a, 981b, 1475b, 2358a BC: 688b, 757a, 1446a, 2166a, 2409b Tervagan et Mahon. BC: 185b, 806b, 1384a

Mahom et Apolin. S3: 438a

Apolin et Cahu, SB: 1476a

Mahomet et Cahu. BC: 899b, 1254b

Apparently Mohammed was considered to be a god rather than merely

a prophet, as his pairing with Apolin, Tervagant and Cahu would

seem to show.

The names of tribes and nations are similarly grouped.

Paien et Sarrazin. SB: 344a, 2203b BC: 2196a

Ehgevins et Bretons. SB: 443b

Tant Turc, tant Sarrazin. SB: 857a

Sarrazin et Persant. SB: 144b, 987b, 1459b BC: 2881a, 3164b, 3709b Sarrazins ne Persans. BC: 2481b Persant et Sarrazin. SB: 1531b

Et Turc et A$oparz. SB: 1670b Et Turc et Achopart. BC: 1823a

Persant et Esclavon. SB: 1024b BC: 180b, 716b, 799b, 3359b

Sarrazin et Escler. SB: 1707b, 1742b, 2582b, 2611b

De Francis et de Turs. SB: 2597a

Ne Franjois ne Bourgoing. BC: 317b

Persant et Arrabi.' BC: 601b, 505b, 1822b, 2132b, 2686b

Franjois et Alemant. BC: 681b

De Turs ne d ’Arrabis. BC: 941b 1 0 1 00 H fO

Et Conmain et Luti. BC: V

Et Persant et Luti. BC: 2152b, 2810b

De Turs et de Persans. BC: 2073a De Frans et d* Arrabis. BC: 2286b

Mains Turs, mains Sarrazins. BC: 2331a

Persant et A 5 opart. BC: 2827a

Pour Turc ne pour Escler. BC: 3059b

K*ainc Turs ne Sarrazins. Bg* 3256a

It does not seem to matter by ■what name the tribes and nationalities

are called, as long as the pair of words fits into a hemistich and,

in the cases where it is a question of the second-hemistich posi­

tion, the rime. Conmain, Luti and Achopart are probably imaginary 1 de signations.

A proper name is sometimes expanded by the inclusion of a

feudal, or in two cases a religious, title, although this does

not always make it take up an entire hemistich.

■Cuens (object case oonte);

Cuens Aymeris no sire. SB; llha Mes li cuens Aymeris. SB; 279a.

•Ha 11 lam Wistar Comfort, in "The Literary Bole of the Saracens in the French Hfc>ic," FMIA. LV (19^0), 63I, speaking of these and other fantastic names for the pagans of the chansons de geste. says; "It is futile to ask who these people were, for the author himself would have been embarrassed to reply. It made no difference to him, or to his audience, who they were. It was sufficient that all these peoples were tarred with the same stick of heresy* The desired Im­ pression was that of an indefinite Hinterland of Islam, from which thousands and mill ions of unbelievers swarmed upon call to combat the hosts of Christendom. In conveying to us this impression, the mediaeval poets have fully succeeded." 102

Fu 11 quens Aymeris, BC: 56a Et Aymerls le quens. BC: 71a St le conte Aymerl. BC: 2086a

Et le conte Engeller. SB: 213b BC: 368b

Et Navari 11 conte. SB: Et Navari le conte. -BC:

Et Savari le conte. SB: 2400a, 2447a, 2467a,

Contesse:

Ermengars la contesse. BC: ?4a

Apostles:

Et 1*apostles sains Piere. BC: 4l7b

Bnperere:

Looys l'ernperore. BC: 1458b

Ghapelein:

Renier, un chapelein. SB: 846a, 877a

Arcevescue:

Que KLchiers 1 ‘arcevesques. BC: 1273a Et Richier l*arcevesque. BC: 1298a, 1817a, 1956a Richiers 11 arcevesques. BC: 1410a Dist Richiers l»arcevesques. BC: 1440a

Rois:

Et au rol Looys. SB: 1195a Et 11 rol Cucament. SB: 1623b Rol Balistant d'Octrente. BC: 298b Et U rois Limbanors. BC: 2330a Mais 11 rois Limbanors. BC: 2604a Fu 11 rois Limbanors. BC: 2624a Dont U rois Limbanors. BC: 2708a

The title plus a place-name may suffice.

Sire:

Le sire de Hulin. SB: 469b 103

LI sires d'Aumarle. SB: 1279b Qui sire est de Persle* SB: 1540b Qul sire est de Tudele. SB: 2056b Sires fu de Baxbastre* BC: 801a

Amirant:

Quo 1*amirant d'Espengne. S B : 30a, 6la, 86a, 159a Mes 1* amirant d ,Espengno. £g: 594a gC: 1736a LI aniranz d'Espemgne. SB: ll4la, 1482a, 1528a, 2211a, 2223a, 2234a, 2337a BC: 2329a C'est 1» amirant d ’Espengne, SB: 1185a Et l'amirant d'Espengne* SB: 1237p, 1540a BC: 2202a A l*amirant de Perse* SB; 2063b, 2132a L'amirans de Baxbastre* BC: 754a, 2156a

Amustant;

Et 1*amustant de Cordres. S B : 62a, 306a, 340a, 470a, 1270a, 2224a, 2235a, 2338a BC: 2930a Qant 1*amustant de Cordres* S B : 409a BC: 1503a, 1566a Sire amustant de Cordres. S B : 1232a LI amustant de Cordres* SB: 1489a BC: 743a, 2l65a, 2253a Que 1'amustant de Cordres* SB: 1996a BC: 179a L*amustant d'Avalonne. BC: 21^, 3374b K'a 1»amustant de Cordes. BC: 1305a A l'amustant de Cordes* BC: 1368a, 1374a, 2093a, 3825a, 3844a Vers l*amustant de Cordres* BC: 1603a Sist l'amustans de Cordres* BC: 2201a Ne l*amustans de Cordres* BC: 3347a

Rois:

Qul rois estolt de Cordres* BC: 282a Qui rois fu de Persle* BC: 2194b

The seme procedures are employed to expand or contract these desig­

nations as were found in the examples of proper names* io4

Indications of blood relationship fulfill the same function.

Pere:

Son pere Maupriant. SB: l60a

Neyeu:

Et son neveu Gaudras. SB: 452b

Fille:

Malatrie ma fills. SB: 1495a

Frere:

Et son frere Prdmant. SB: 1499b Et son frere Sanson. BC: l66b Et son frere Ranier. BC: 369b Et Guillaume mon frere. B C : 679a Et Aimers mes frere s. BC: 682a Et ses freres Normans. BC: 1956b Et Guielins mes freres. B C : 2452p Et son frere Guimer. B C : 3028b

Cousin:

Escorfaus ses cousins. BC: 2180a

Expressions such as Escorfaus ses cousins, et son frere Primant

and son frere Rn-tawnr are pure fillers for the second hemistich,

since these characters appear in the verses cited and then never

again in the course of the two works. They are obviously merely-

thought up for this particular occasion by the .jongleur.

Like the names of persons, place-names in the two chansons de

geste are often combined two into one hemistich. This is done by

means of the conjunction et:

Normandie et le Meinne. SB: 443a Normendie et Henaut. SB: 1746a Normandie et Bijou. SB: 1999a Aufaleme et Luis erne. SB: 2074a Placite et Paleme. SB: 2075b 105 Boourges et Berri. BC: 1028b or et • • • et:

Et Lohereinne et Flandres. SB: W4a, 2000a Et Vigome et Tudele. SB: 20? Et Burs et Carion. SB: 2076a or ne . « • ne:

Ne France ne Bourgoingne. BC: 339b or else they are simply Juxtaposed by ellipse of the conjunction:

Onquepuie, Candle. SB: 2075a Tolete, Panpelune. SB: 2078a

Some of these cities correspond to geographical realities (Palerne,

Tudele, Burs, Carion, Tolete, Panpelune and Candle are Palermo,

Tudela, Burgos, Carrion, Toledo, Pamplona and Gandia) but others are taken from the realm of fantasy like much of the non-French toponymy of the other chansons de geste. Vaudune, Aufaleme, Lui-

seme, Placite, Vigome, Onquepuie, these are meant to lend an air of exoticism to the Saracen hosts.

Simple though the mechanism of these proper names which take up entire hemistichs may be, it is nevertheless as much a part of the iongleur>s art as playing the vielle. Of the h-55 hemistichs "which are occupied by proper names either standing alone or in combination with titles and place-names, 205 are from SB (approximately 76 per

1000 verses) and 250 from BC (approximately 63 per 1000 verses). CHAPTER V

DISCOURSE

A great part of any chanson de geste consists of the words of the individual characters speaking to one another in normal social intercourse or shouting threats and deprecations at their enemies. It is in no way strange, then, that among the various formulas ■which convey the story-line in the epic those ■which serve to begin direct quotations should appear frequently. So often do these particular formulas occur in the text of the two works that I have, for purposes of coherence of presentation, grouped them together with other phrases which are found principally in the speech of the characters to form a classification which will be called **forraulas of discourse,** This heading will be limited to formulas and expressions which are part of the mechanism of speech in the two poems.

This chapter will first treat of expressions which serve to open the discourse of a character, whether in reply to what has been said by another, or as the first speaker in a conversation.

The problem of indirect discourse in the chansons de geste will be discussed, and then there will be an examination of spoken expressions which do not open discourse, such as oaths, excla­ mations and battle cries. 106 107 Expressions Opening Discourse

A common pattern for verses -which introduce direct discourse is that in -which the first hemistich contains the verb of speaking with its subject, the person who is about to speak,'and one other word, generally a monosyllable, according to the needs of the versification. The following examples are from a pattern of this kind in which the verb is respondre.

Et li paien responent. SB: 844a Et respont li (his Bueves. SB: 1081a Et Bueves li respont. SB: 1400a, l4o£a, l4-21a Et l ’amirant respont. SB: 1403a Et respont l*amirant. SB: 1335a, 1408a Li amiranz respont. SB: 14-97a Libanor li respont. SB: 2668a

These formulas provide a good example of what I call Metrical

compensation," where one or two words are added to or subtracted

from a hemistich to make up for a loss or addition of syllables

due to the rearrangement of the words of the formula. In the third example given above, the verb and the subject Bueves are both bisyllabic, so that two "filler" words are needed to bring the number of syllables to the required six. In the second example, however, Bueves is found at the caesura so that its e mute is not

counted in the syllabification: the word dus is therefore added to make up for this loss. The same phenomenon can be seen in the

last two cases. Libanor, a proper name, cannot have a definite

article before it, so the missing syllable is supplied in the form

of the indirect article li. Notice that the verb and the subject

are the only necessary words in hemistichs of this kind, but they 108 are usually accompanied by other words, articles, titles, indi­ rect objects, conjunctions, 'which can easily be altered or elim­ inated to bring about the necessary adjustment.

The verb respondre provides another example of this process of metrical compensation. When the verb occurs in the past tense it becomes trisyllabic, so that one of the other words in the hemistich must be altered. Here it is the noun which disappears, giving place to a pronoun subject.

Et ill ont 'respondu. SB? 166a, 1506a, 2185a, 2213a Et oil ont respondu. SB: 245a, 1865a, 2120a Et oil li respondirent. SB? 1253a, 2502a, 2606a Et cil li respondi. BC: 2323a

The striking thing about such systems of metrical compensation is their economy. The jongleur only uses certain combinations, omit­ ting others entirely. Thus there is no case where the poet of SB uses the phrases Et cil ot respondu or Et cil li resnondi: the singular number of respondre is found only in the present tense so that the number of combinations which present themselves to the poet in the course of composition is reduced by half. The poet does not have to choose from a n the possible combinations of respondre and a subject, for this would be too difficult a task for him to per- foim and still maintain the steady pace of oral composition. But the respondre system, limited to the necessary syntactic forms, is ingrained in his memory in such a way that he can instantly produce 109 ■what Marcel Jousse has called "ce de clench era etnt mi-automatique, mi-volontaire de gestes propositiormels cliches."^

The word respondre. used to introduce a direct quotation, is found in SB only in the combinations which have been described here. Consequently the efficiency of the respondre speech system is such that there are only five patterns in its (l) et plus pro­ noun subject plus ont respondu: (2 ) et plus noun subject plus respont or responent;^ (3 ) ©t plus pronoun subject plus indirect object plus respondirent; (4-) and (5 ), the two phrases which are found only once each, 11 ami rang respont and Libanor li respont. neither of which appears -unusual in the light of the metrical compensation mechanism.

In BC, on the other hand, the pattern described above does not appear. There are only two instances of the use of the word respondre. The one which is listed with the SB formulas above,

Et oil li respond!, does not occur in SB, but it too could con­ ceivably be a product of metrical compensation and would fit into the pattern of the SB formulas quite easily, were it not for the fact that it does not introduce a direct quotation but is used for indirect discourse. The full verse is:

Et cil li respond! que ainsi le fera. BC: 2323

This pattern does not exist in SB. for there is no indirect discourse

^Marcel Jousse, Etudes de Psychologic linguistique: le Style oral rrthmique et mnemotechnique cfaez les Verbo-moteurs. "Archives de Phi­ losophic.w Vol. U (Paris: Beauchesne. 1925). cahier 4, p. 157*

^Responent at the caesura is bisyllabic like respont. 110 in SB. Every tine a persomage speaks the poet presents his words

In direct quotation. This practice of using only direct discourse is, I believe, proper only to the chansons de geste among the nar­ rative genres. In message scenes the messenger is told what he should say to the person who is to receive the communication, then he in turn gives the message in full when he delivers it, often

In the very same words in which it was given to him, and finally the person receiving it may repeat it, again word for word, to his counsellors. Adenet does not hold to this scheme, and we shall discuss more examples of his use of indirect discourse later in this study.

The other instance where Adenet uses the word respondre is in

BC: 902.

Que a grant peine a il un soul mot respondu.

This verse obviously has no relation to the highly stylized pat­ terns of SB. On the contrary it is an indication that, at least for the verb respondre. Adenet did not use a system, for of the two examples out of 39h? verses of BC one is obviously not linked to a system of any kind whereas all 21 uses of the word in SB are systematic.

For the verbs crier and escrier a similar set of patterns can be described. Usually the verb is found at the caesura, as with the instances of respondre in SB. and whenever the verb is not in this position it is at the rime.

Et la gentil contesse si s’ala escrlant. SB: 198 Mes Aymerls s*escr£e, ne se tint mie muz. SB: 298 in

Id. amiranz escrle: Car prenez oez glotonsj SB: 428 L'amorolves escrle: Seignor, estez on pals. SB: 526 Et Bueves lor esorles Seignor, trop demorez. SB: 923 Les aquet a l'espee, sovent s'est escrlez. 1377 EL chief del rano s'est trez si s'estolt escrlez. M s 1394 Guillaume ont escrle: N*i garlres, trechiere. BC: 420 A nostre gent escrle: Ne sollez alentis. BC: 498 Vint poignant vers nos gens, si les a esorles. BC: 1884 No gent a esoriee si que bien l'ont oft. BC: 2008 A nos Francis escrle: Foie gent et despise. BC: 2838

In this case the two poems do not seem to differ, although the use of escrler here is not as rigidly formulaic as respondre was. Es- crler. and also crier, are used for battle cries.

Aymerls s'escria: Monjoie la Charlon. SB: 228 Puis escrle: Monjoie, aidiez, seinte Marie. SB; 1291 11 escrle: MonjoieJ, qant celui vit feni. SB: 1328 XL escrla: Monjoie I dus Bueves, car m'aidiez! SB: 2160 Bueves crie: MonjoieJ et Ranier: Monpancier! SB: 2165 XL escrle: Monjoie! chevalier, car m'aidiez. SB: 2617 Qnr>+. 41 r\Ki M r m M o A nftw. SB: 2662 . BC: 449 Valdune lor escrle, de bien faire haities. BC: 1185 A l'abaissier des lances ont Nezbonne escriee. BC: 1709 Tabarie escrla au fuer de sa contree. BC: 1725 Francois crlent Monjoie, l'ensaigne saint Denis. BC: 2800 Guis escrle: Monjoie! Feres, franc damoisel. BC: 3738 Joffrois escrle Anglers et Gerars Conmarchis. BC: 3775

In six cases in SB the first hemistich contains the subject, the verb of speaking or hearing, and the battle cry; in BC this pattern is present five times. This is not a significant difference in treatment, although SB does seem to present greater rigidity in its patterns.

Escrier and crier are often found in combination with an adverb which reinforces the verb without really adding anything 112 to the idea it contains. Host often this adverbial element has to do with the force of the character's voice.

Aymerls de Nexbone s'est en haut escrlez. SB: 258 En haut par mi Barbatre est li feus escrlez. SB: 798 Atant es vos Gaudras, qul s'est haut e series. SB: 1202 Qant li mes fu a cort, hautement li esorle. SB: 1686 Venu sont en la vile, haut se vpnt escrlant. BC: 1146 Clarions des oretiaus lor a haut escrle. BC: 2727

Two forms of this adverbial complement are highly formulaic. The first is used when for some reason it is convenient for the verb and the adverbial phrase to appear in the same hemistich.

A haute voiz crla: "Retomez, chevalier!" SB: 206 A haute voiz s*escrle del mur enciennor. SB: 331 A haute voiz escrle: "Ferez, mi fil, ades.11 SB: 778 A haute voiz escrle: "Bonier de Montermier." SB: 2601 A haute vois s'escrie: »*Mahon, je suis trays." BC: 964

The one occurrence in BC is not formulaic, since by definition a formula is a phrase repeated in the same poem, but Adenet may be simply borrowing the expression from SB.

A similar system of formulas, expressing in a whole verse

•what the one just discussed condensed into one hemistich, is made up of a six-syllable adverbial phrase and a six-syllable predicate or a subject predicate combination.

A sa voiz clere et haute apela Aymeri. SB: 324 A sa voiz qu'il ot clere duremant s'escria. SB: 677 A sa voiz qu'il ot clere les avoit escrlez. SB: 765 A sa voiz qu'il ot clere fierement li escrle. SB: 1281 A sa voiz qu'il ot haute, conmenpa a crier. SB: 1729 A sa voiz qu'il ot clere conmenpa a crier. SB: 1966 A sa voiz qu'il ot clere conmenpa a huchier. SB: 2127 A sa voiz qu'il ot clere conmenpa a huchier. SB: 2634 A sa voiz qu'ilot clere conmenpa a huchier. S3: 2652

There are no occurrences of this formula in BC. It is to be noted that the second hemistich of this pattern is not stable; nevertheless 113 it does contain some qua si- fo rmulaic elements. The second, third and fourth cases all contain escrler and the last five all begin with conmenca. which is followed either by a crier or a huchier.

There are, then, vaiying degrees of stability for formulas, and the pattern which made the anonymous poet use the phrase conmenpa a huchier in the second hemistichs of these verses (he uses conmancent a huchier. which is substantially the same, in SB: 4879) is not so well established in his mind that he does not at times choose to replace it with other hemistichs. The first hemistich, on the other hand, is fairly stable. At least the poet had the habit of joining a sa voiz with either clere or haute (see SB: 5925, A sa voiz haute clere) in these circumstances.

The verb apeler (in the two senses "to call to" and "to name") is generally used in a non-formulaic manner in both SB and BC.

There are, however, two fixed patterns. One consists in placing the object of the verb (that is the person being called) at the beginning of a hemistich and the verb at the caesura or the rime.

Barely the subject may begin the verse.

Ses chevaliers apele. SB: 334a Un Sarrazin apele. SB: 451a Corsout en apela. SB: 471a L ,amiraus apele. SB: 512a Le chartraniar apele. SB: 528a Beelas apelerent. SB: 548a A Girart apele. SB: 570b Si les a apelez. SB: 809b, 838a Clarion les apele. SB: 979a, 1022a Li cuens Girart apele. SB: 1056a L* amustant apela. SB: 1111a Et Buevon m •apelerent. SB: 1402a Li amustant apele. SB: 1498a Le Turc en apela. SB: 1560a I

114

Li amiranz apele, SB: 1697s Le marechal apele, SB: 1905a Guielin apele,- SB: 2445b Ses puceles apele. SB: 2530a Aimeri en apele, BC: 578a Buevon en apela, BC: 622a Clarion l*apeloient. BC: 799a Nos Francois apela,. BC: 1127b Buevon a apeler. BC: 1429b Malaquin apela, BC: 3232b

The second pattern reverses the order and places the verb first, and the object at the caesura or the rime.

Apela Madiant. SB: 159b Apela Aymeri. SB: 324b S»apela Fouqueres. SB: 779b En apela Guion. BC: l6*Jb Lors apela Joffroi, BC: 3383s

The verbs aresnier. clamer. conmender and saluer are used both to introduce direct quotations and as verbs describing actions outside of actual dialogue,

Aresnier:

Aymeris l*aresone. SB: 140a Ses a aresonnez. SB: 1071b L'en a aresonne. SB: 2535b, 2543b Tous seuls enaraisonne. BC: 817b A Buevon araisnie. BC: 1402b De Girart araisnies. BC: 2590b Gerars bel araisnies. BC: 3670b Gerars, moult l*araisonne. BC: 3368b conmender:

L'amirant conmende. SB: 353s Que l ,amirant conmende. SB: 469s Se li dus le -conmende. SB: 2429s clamer:

Et amirant clamez. SB: 1409b Et traytres clames. BC: 1180a Et recreans clamer^ BC: 3074b 115 saluer:

La pucele salue. SB! 1535a Et Girart la salue, SB: 2053a De Dieu les salua, BC* 1018a L*amustant salua. BC* 1483a, 2149a Limbanors la salue, BC: 2189a,» 2396a Et Gerars la salue. BC: 2661a

The locution entendre (escouter) sa reson is also found.

Escoutez ma reson. SBi 113b, 2493b Entendez ma reson, SB! 240b Bien entanz ma reson. SB: 2468b

The most frequent of the verbs of speaking, however, is dire. It occurs more times than the combination of all the others which have been discussed.

Usually dire is found in conjunction with the name of the person who is speaking (the subject) and the designation of the person who is being addressed,^ so that it does not fall into the same patterns as do the other verbs. This is simply because of the physical size of the word. Hire is always either monosyllabic or bisyllabic and thus leaves enough room in the hemistich for other important elements to be introduced. When the subject and the designation of the per­ son who is being spoken to add up to less than five syllables, the neutral object ce can accompany dit for metrical compensation.

The designation of the person who is being addressed sometimes takes the form of a proper name,

Clarion, ce dit Bueves. SB: 823a, 890a Clarion, dist Gerars. BC: 939a

^Ebcceptions to this are Si lor dist en riant (SB! 979b), Si li dist en riant (SB: 1621b), Or ne set an a dire (SB: 118a), He le sau- rait a dire (SB! 976a) and One geru sache redire TSB: 2063a). 116

Clarions, dist dux Bueves. BC: 976a, 1315a Corsolt, dist l*amirans. BC; 1492a Girart, dit Malatrie. SB: 2550a Girart, dist la pucele, SB: 2567a Gerart, dist Limbanors. BC: 2553a Libanor, dit Girart. SB: 2001a Hmbanor(s), dist li rois. BC: 2135a, 2255a Limbanor, dist la bele. BC: 2266a Malaquin, dit la bele. SB* 2377a, 2393a, 2536a Malaquin, dit Girart. SB: 2414a, 2468a, 2493a Malaquin, ce dist Guis. BC: 3594a Malaquin, dist Joffrois, BC; 3598a Renaut, ce dist Gerars. BC: 3504a

The person*s rank in feudal society is sometimes used as the term of address.

Amirans, dist li rois. BC: 735a Baron, ce dist li quens. BC: 354a Baron, dist li dux Bueves. BC: 1863a, 1983a Baron, dist Limbanors. BC: 237*Ja Dame, dit Aymeri s. SB: 1150a Dame, ce dit Gaudin. SB: 1542a Dame, dit Malaquin. SB: 2379a,2386a, 2405a, 2537a Dame, ce dit Girart. SB: 2545a, 2570a Dame, ce dist Gerars. BC; 80a Dame, dist Aimeris, BC: 556a Dame, ce dist Guillaumes. BC: 570a Dame, ce dist ELandrine. BC: 3189a, 3206a Damoisele, dist il. BC: 2204a Ce dit Girart, pucele. SB: 2080b Seignor, ce dit Girart. SB: 113a Seignor, dit Clarion. SB: 735a, 810a, 839a, 1250a Seignor, ce dit due Bueves. SB: 1189a Seignor, dit Libanor. SB: 1857a Seignor, ce dist Gerars. BC: 167a Seignor, dist Aymeris. BC: 223a Seignor, dist Clarions. BC: 1040a, 1202a, 1 3 U a Seignor, dist Malaquins. BC: 3602a Seignor, dist il a aus. BC: 3625a Sire, dist l*amustant. SB: 430a, 1115a Sire, dist l*amirant. SB: 447a Sire, dit Clarion. SB: 636a, 646a, 667a, 692a, 896a, 1073a Sire, dlent Francois, SB: 938a Sire, ce dit Corsout. SB: 1129a Sire, dit la pucele. SB; 1713a, 1721a, 1750a, 1756a, 1765a, 1772a, 2141a, 2348a Sire, dlent si home. SB: 1856a Sire, dlent li home. SB: 1919a 117 Sire, ce dit Gautier. SB: 1928a Sire, dit Malatrle. SB: 1959a Sire, dit Malaquin. SB: 2443a, 2495a, 2517a Sire, ce a dit Bueves. BC: 226a Sire, dient paien. BC: 248a Sire, ce dist Guillaumes. BC: 356a, 455a, 1674a Sire, dist Emengars. BC: 558a Sire, dist l ’amirans. BC: 710a, 712a Sire, dist Clarions. BC: 956a, 987a, 1001a, 1445a Sire, ce dist Gerars. BC: 1061a, 3051a Sire, dist Limbanors. BC: 2138a, 2263a Sire, dist Malatrie. BC: 2217a, 2526a Sire, ce dist Gautiers. BC: 2482a, 2493a Sire, dist la pucele. BC: 2741a, 2946a Sire, dist Navaris. BC: 3369a Sire, ce a dit Guis. BC: 3584a Sire, dist Malaquins. BC: 3595a Sire, dient Franyois. BC: 3611a, 3628a Vassal, dit Aymeris.. SB: 150a Vassal, dit Navaris. SB: 7l3a, 740a Vassal, dit Clarion. SB: 722a Vassal, dist Limbanors. BC: 2436a Vassal, ce dist la bele. BC: 2665b Vassaus, dit la pucele. SB: 2068b Vassal, dist la pucele. BC: 2670b

The word Sarrazin is also used in this way.

Sarrazins, dist li dus. SB: 398a Sarrazin, ce dit Bueves. SB: 411a, 633a, 640a, 68'Sa, 769a Sarrazins, dist Gerars. BC: 2541a, 2551a, 2570a, 2619a, 3321a, 3353a Sarrazin, dit Girart. SB: 1972a, 1948a Sarrazin, ce dist Bueves. BC: 635a, 673a Sarrazins, ce dist Guis. BC: 885a Sarrazins, dist Gautiers. BC: 3281a

Words indicating the speaker’s acquaintance with or blood relation­

ship to the person addressed also fit into this category.

Amis, dist Navari. SB: 728a Amis, di!ent Franjois. SB: 730a Amis, ce dit Gautier. SB: 2426a Amis, ce dit Girart. SB: 2441a, 2450a Amis, dit Malatrie. SB: 2544a Amis, dist Aymeri s. BC: 190a Amis, ce dist Gerars. BC: 903a, 3309a, 3606a Amis, ce dist dux Bueves. BC: 915a Amis, dist li dux Bueves. BC: 1003a, 1194a 118

Amis, ce dist Joffrois. BC: 1037a Amis, dlent Franyois. BC: 1130a, Amis, dist la pucele. BC: 2314a, Bele, dit l'amirant. SB: Bele, ce dit lou rbi. SB: 1745a, Bele, dit Libanor. 1758a, 1778a, Bele, dit 11 baron. SB: Bele, ce dit Girart. SB: 2145a, Bele, dit Malaquin. SB: Bele, dist Limbanors. 2415a, 2522a, Bele, ce dist Gerars. 2747a, 3676a, Fille, dist 1 *amustant. 1725a, 1764a, Fille, dist l*amustans. BC: Frere, ce dit 11 dus. SB: Frere, ce dit Girart. SB* 2179a, Frere, ce dit 11 cuens,>• SB: Frere, ce dist Gerars,>• BC : llSa, Frere, ce a dit Guis. 3088a, 3H9a, Pere, ce dit Girart. __ 2249a, 2283a, Peres, ce dit Guillaumes. BC: 442a Biaus amis, ce dist Guis. BC: 934b Biau fis, ce dit 11 dus. SB: 1171a Biaus fils, ce dist 11 quens. BC: 453a, 1687a

In all of these cases it is evident that BC is as foranilaic as SB.

One of the most frequent locutions of the chansons de geste is the invocation of God and the saints. Mien the poet presents the

Saracens as calling upon their God, it is Mohammed himself 'who is mentioned, although, as any informed Christian even during this period might know, the founder of Islam never claimed to be divine.

The formulas which convey this misinformation probably came into use before the period of the crusades and are an example of the archaiz­ ing nature of the oral tradition.

At times the invocation of God or Mohammed or the saints takes on the form of direct discourse, analogous to the ordinary formulas of speech between mortal men which are listed above in this chapter.

Et Dieus! ce dit Girart. SB: 648a, 1884a Ha, Dieus, ce dist Gerars. BC: 844a He, Dieus, ce dist chascuns. BC: 1032a Hal Dieus, dist li dux Bueves. BC: 1457a Ha, Dieus, dist Ermengars. BC: 1659a Mahom, dit l'amirant. SB: 2645a, 2665a Mere Dieu, dist dus Bueves. — BC: 2114a

Each of these formulaic expressions concludes •with the character calling upon his respective divinity, or the Virgin, to help him in combat.

This type of address formula also serves the purpose of allow­ ing knights to insult each other on the field of battle with the same set patterns.

Gloz, ce dit Clarion. SB: 1359a A| glouz, dit l'amirant. SB: 1425a Et glozl ce dit li dus, SB: 2240a, 2300a HaJ glous, dist li dux Bueves. BC: 3053a Aj leres, dist Corsols. BC: 1223a

The terms of opprobrium are analogous, in the way they are handled, to the terms of address.

These same categories (terms of address, including name, rank,

"nationality,” titles of friendship, names of divinities and expres­ sions of opprobrium) are used in conjunction with the verb faire in much the same fashion, since this verb shares with dire the advan­ tage of limiting itself to one syllable in the present and simple past tenses. The following list of the occurrences of faire in SB and BC in hemistichs analogous to those listed above for dire will need no explanatory notes,

Amirans, fait il, sire. BC: 1484a Amustans, fait il, sire. BC: 695a Baron, fait soi li dus. SB: 2117a Damedieus, fait il, pere. SB: 2098a Damedieus, fait il, sire. BC: 432a Douce amie, fait il. BC: 770a 120

Fait ele, sire fis. BC: 1630b Fait il, ma douce amie. BC: 2737b Fill a putain, fait il. BC: 1118a, 1885a, 2009a Flandrine, fait la bele. BC: 3186a Foie garce, fait il. SB: 2343a BC: 2941a Frere, fait il a lui. BC: 3101a Gerars, fait ele, sire. BC: 2680a Hal Clarion, fait-il. BC: 1359a Raj lere glous, fait il. BC: 2989a HaJ sire Dieus, fait ele. BC : 102a, 2736a Mahomet, font il, sire. BC: 1560a Mi neveu, fait li cuens. BC: 214a Seignor baron, fait il. BC: 3307a Seignor, fait il a aus. BC: 1019a, 1385a Seignor Francois, fait il. BC: 1186a Sire, fait il a lui. BC: 1403a, 2127a Sire frere, fait il. BC: 3548a

It is noteworthy that only three out of these 28 occurrences of the

verb faire in this type of hemistich are from SB. This-is the first

of the systems examined here which is found to occur predominantly

in the later poem, and it is perhaps because the use of faire to mean dire was not yet firmly established at the time of the composi­

tion of SB. Leo Jordan explained this new usage of the word fait by the fact that the £ of the simple past dist began to be omitted « in pronunciation so that the present, dit. was indistinguishable

from the past. The present tense -il faitT "he says," began to re­ place il dit toward the end of the twelfth century, that is at about the same period as the silencing of the s_. The past, "he said," however, continued to be expressed by il di(s)t.^ Thus fait il and dit il are not originally two different formulas; the former is a development from the latter necessitated by homonymy due to an Old

French phonetic change.

\rchivum romanicum. IX (1925), 292. The abundance of these hemistichs in both SB and BC can be easily accounted for. If the mo d e m reader of a written work happens to lose track of which character in a dialogue is speak­ ing, he can always turn back to the previous page and trace the dialogue forward to the point at which he became confused. But in an oral work it is up to the poet to be so clear that his audi­ ence will have no doubts about which of the characters is speaking.

The easiest way for him to do this is to introduce as many quota­ tions as possible with a hemistich which includes the name of the speaker and that of the person being addressed. The poet must be careful not to repeat the same phrase to the point of tedium, but, from what one can gather in reading oral poetry, the audiences must have been far less demanding in this regard than m o dem readers of written works. In laisse 68 of SB. for example, in 7^ verses there a m no less than 16 hemistichs of this kind, containing the name of the person being addressed, the verb of speaking and the name of the speaker. Such a set of formulas is indispensable to the craft of oral poetry, especially if there is a great deal of dialogue, which there certainly is in both our poems.The tech­ nique of the mo d e m novel where, after the first phrase spoken by each character in a dialogue, there is often no indication at all of who is speaking, would be impossible to put into effect in an

^Erich Roll, Untersuchungen uber das Verhaltnis des "Siege de Baitaastre" gum ,rBueves deCommarchis" von Adenet le Roi und die Shall-png der Prosafassung (Greifswald: Hans Adler. 1909), p» ^5» states that 133^ verses of the 3860 of BC (not counting the intro­ duction), and around 250 of the first 1000 of SB consist of direct discourse. 122

oral work. Discounting hemistichs in •which one or two of the three

elements (addressee, verb of saying, speaker) were absent, there

are in SB 103 hemistichs of this type, or approximately 38 for every

1000 verses, and in BC 130, or approximately 33 P©** 1000.

Replies

The verbs of discourse "which have been treated so far often

introduce short formulaic phrases which give an immediate positive, negative or indifferent reply to the question or command which im­ mediately precedes them in the narrative. These replies are almost always found in the second hemistich of the verse.

The two most common replies to commands in SB are Biau sire. volentiers and Si con vos conmendez.

Biau sire, volentiers. SB: 1865b, 2120b, 2606b, 2668b Si con vos conmendez. SB: 8h4b, 938b, 2213b, 2560b, 2570b

l'leither of these egressions, which are among the definitely estab­

lished formulas of SB, are present in BC.

The other replies to commands are, for the most part, non-

formulaic in both works.

Et il ont respondu: ”Tot a vostre talant." SB: 1506 Par mon chief, ce dit Bueves, ce fet a ostroier. SB: 913 Et ill ont respondu: lfSire, a vostre conmant." SB: 166 Bele, dist Limbanors, tout a vostre conmant. BC: 2*4-15 Amis, d£ent Franpois, a vostre volente. BC: 1206

Adenet not only does not employ the same formulas for replies

to commands, but he does not substitute anything in their place. In

situations where one character gives his assent to what the other 123 has said, Adenet simply does not use a set of formulaic expressions

of any kind.

For positive replies to previous proposals, SB has the for­ mulaic expression voire plus verb of saying plus subject, always

in the first hemistich position.

Voire, dit Guielin. SB: 64?a Voire, ce a dit Bueves. SB: 758a, 1186a Voire, ce dit Girart, SB: 1088a

In BC the word oil is used in a similar type of expression, of which

there are two examples.

Oil, dist l*amustans. BC: 2213a 0x1, dame, fait il. BC: 3651a

For a negative reply, SB has:

Non voir, ce dit Girart, SB: 1182a Nanin, dit Clarion. SB: 729a

and also a more polite expression:

Se vos plest, non feron. SB: 430b Se vos plest, non ferez. SB* 1928b Se vos plest, non fere. SB: 2h26b

The expression Ce fet a mercier. an elaborate way of saying "thank

you," is found six times in the whole of SB. but only twice in the

section before verse 2696.

Ce fet a merci'er. SB: 1715b, 1750b

Another phrase which means roughly the same thing is Molt est riches

li dons:

Molt est riches li dons. SB: 447b, 1751a 124

These formulas of thanks do not appear in BC, but there are anal, ogous phrases.

Ci a riche present. BC: 2217b Ci a don seignori! BC: 605b Cis dons est seignoris. BC: 2263b

BC has not one instance where a reply to a previous proposition is formulaic.

Terms of Address

There are other patterns for hemistichs which open a direct quotation with the name (or some other designation) of the addressee, but they are less common and not nearly as rigid in their styli- sation. An entire hemistich is sometimes used to carry the name of the person who is being spoken to.

AhiJ Mahonmet sire. BC: 26l6a, 2626a Amirant, riches, rois. SB: 1687a, 2653s- Amirant, riches sires. SB: 1730s- Apollin, riche dieu. SB: 592a Daraedieu, sire pere. SB: 301a, 573s , 706a EJ amirant, biau sire. SB: 678a Et Guillaume d*Orange. SB: 325a Franc baron esleu, BC: 980a, 1023a Gerars de Conmarchis. BC: 3^17®, 3**Q5a Guillaume, sire fil. SB: 200a HaJ Bueves, sire pares. BC: 1926a BaJ Guielin, biau frere. SB: 2l6la Hahi, Buevon, biau fiz. SB: Libanor,9 biau doz fiz. SB: 2666a Renaut de Montarmer. BC: h88a Seignor, franc chevalier. SB: 980a, 1023a Sire, frans chevaliers, SB: 1058a Seinte Marie, dame. SB: 261a Chet if maleure. SB: 600b Fill a putain, Francois. SB: 766a, 1203a, 1282a Fill a putain, traitres. SB: 957a,9 1 3 5 ^ Fill a putain, larron. SB: 1378a 125 These hemistichs are usually found in places where there is no doubt -who is speaking. Often the verse which precedes them con­ tains the speaker* s name. The process by which they are formed is a simple one. Usually they are composed of the name of the person addressed (or his title, or a term of opprobrium) plus a noun in apposition. In some cases they are simply hexasyllabic names, such as Renaut de Montarmer. and in one case a noun of two syllables is modified by an adjective of four, chetif maleure.

In the following examples the same problem (avoiding the verb of speaking because it has been supplied in the previous verse) is solved by opening the direct quotation in the same hemi­ stich in which the addressee* s name is mentioned.

Amis, lieve tost sus. SB: 673a Amis, de mon cheval . . . SB: 1906a Amis, vos an irez . . . SB: 2384a Amis, ovre la porte. SB:' 2424a Biaus amis, or te garde . . . BC: 3599a Aymeri, secor moi. SB: 199a Barons, secoron le. SB: 783a, 2604a Ber, car secor ton fil. SB 2109a Barons, alons nos en. SB 2210a Deus vos saut toz, barons. SB 2439a Baron, or dou haster. BC 1930a Baron, or dou ferir. BC 3718a Bele, fait il, ja fust. BC 2291a Bele, j*ai non Gerars. BC 2691a Bueves, moult honnories. BC Chevalier, ore as annes. SB 914a Corsolt, fait il, plains estes . . . BC 1231a Clarions, je vous pri . . . BC 998b Cousin, venez pa fors. SB 1926a Tornez, cuens Aymeris. SB 332a Dame, que il n*i soit . . . SB 262a Dame, or oiez noveles. SB 1526a Dame, par ce Hahon. SB 1564a Dieus, tante riche ensangne . . . SB H 7 7 a DieusJ con baloient la . . , SB 226a Dieusi fait il, bonne estrine . . . BC 264a 126

Dieus, si grant duel demaine , , BC: 487a Vrais Dieus, et car fust ore . . . BC: 1068a Qui est ce, belle fille? SB: 1722a Biau fiz, desfandez vos. £S3: 563a Desfandez vos, biau fiz. SB: 567a Biaus fieus, secores moi. BC: 390a Frere, tenez cestui. SB: 2198a Frere, paries a moi. BC: 229b Frere, fait il, com estes . . . BC: 3131a, 3570a Certes, Gerart, moult estes . . . BC: 2556a Gerart, fait il, j»ai si . . . BC: 2644a Gerars, or verrons nous. BC: 2728a Gerars, en vous ai raise . . . BC: 3348a Pour les sains Bleu, Gerart. BC: 3387a Gerars, quant tel jpucele. BC 3495a l'lahom, je sui trays. BC 964b Mahoml con il est biaus. BC 1948a Biaus nies, ouvres la porte. BC 2479a Pere, vestez l ’auberc, SB 753a Portiers, ouvres la porte. BC 3274a Re£ne, se geu pert . . . SB 263a Seignor, adoubez nos. SB 170a Seignor, estes en pais. SB 526b Seignor, trop demorez. SB 923b Seignor, de voz noveles. SB 1072a Je ne puis certes, sire. SB 142a Sire, a vostre conmant. SB 166b Sire, autre chose i a. SB 1082a Sire, or poez o£r. SB 1112a Sire, moult vous devroie . . . BC •105a Sire, prenons en gre . . . BC 773a Sire, par cel seignor. BC 1243a Sire, se ma raisons . . . BC 1275a Sire, font il, merci. BC: 1339a,, 1389a Sire, a Nerbonne aloient. BC 2075a Kouvons, sire, fait ele. BC 2413a Ales, sire, fait il. BC 2498a Sire, fait il, drois. est. BC 3528a Sire, fait il, tout fussent . . . BC 3827a A ceste foiz, biau sire. SB 513a Biau sire, avant venes. BC 1080b Ne vous devez, biau sire. BC 3012a Or en soiez, douz sire . . . BC 107a Sire oncle, car montez. SB 276b 1'ial sanies, sire fiz. SB 571a Sire dus, je vos di. SB 902a Sire rois, vostre fille. BC 2151a Sire rois, esgardes. BC 2508b Vasal, por menacier. SB 628b Qui estes vos, vassal? SB 1967a 127 Vassal, fait il, fols estes. BC; 652a

These terms of address are for the most part the same as in the combinations "which were treated earlier. Obviously there is no clearly delineated hemistich system in this case. One can say only that the term of address is usually found in the first hemistich, and generally at the beginning of the verse, and that it is of one, two or three syllables but no more. The rest of the hemistich is made up of the beginning of the direct quotation, and possibly a verb of saying, fait il.

In order to give a complete picture of the mechanism for opening discourse in the two poems, one would also have to examine those verses "which do not fall into well-defined patterns, but this would be a tedious process without much in the way of a positive conclusion. Suffice it to say that both poems have verses which do not fit into, some one of the patterns treated here, but that they are few in number.

There is one procedure worth mentioning here in connection with the alternatives which the poet has to using a formulaic ex­ pression, and that is the problem of indirect discourse. It is one of the characteristics of the chansons de geste that the characters always speak in direct quotations, and that the poet never merely says that they said something, but rather gives their exact words. BC does not share this characteristic with the other chansons de geste. as is seen in the passages where we are told that someone is speaking, and perhaps even what he is talking about, 128 but not the exact words he used: BC: 1022-25, 1217-18, 2070, 2323,

3240-42. This is, I believe, a serious departure from the conven­ tions of the genre which are otherwise universal in this respect.

Oaths and Prayers

There are other formulas of discourse which do not necessarily open direct quotations but which are commonly found in the mouths of the characters. Most frequent among these are the oaths which the knights swear at the slightest pretext.

Par ice seint apostre q»an a Ronnie requiert. SB: 2291 Par ice seint apostre que quierent chevalier. SB: 2296 Que par icel apostre que quierent chevalier. SB: 2487 Que, par ce seint apostre c*on quiert en Noiron pre. SB: 2561 Mais, par cel saint apostre c*on quiert en Noiron pre. BC: 848 Frere, ce dist Gerars, par 1*apostre saint Piere. BC: 3^62 Damedieus les confonde et 1'apostles saint PierreI BC: 1785 De trayson me rette mais, par sainte Marie. BC: 2739

Or me rsndez les contes, por Deu le fiz Marie. SB: 688 On m ’apele Buevon, par Dieu le fill Marie, BC: 624 Se nous fussiens arme, par Dieu le fill Marie. BC: 1038

Mes par ice Mahom que ge doi aorer. SB: 2231, 2345 Car par cel saint Mahom a cui sui otroiies. BC: 661 Mais par cel saint Mahom cui je aour et pri. BC: 1494 Par ce saint Mahonmet par cui il pluet et vente. BC: 2240 Par cel saint Mahomet a cui je rent servise. BC: 2840

These phrases constitute a formulaic system in spite of many varia­

tions.

It appears, from evidence assembled by M. Oner Jodogne that

Par cel apostre followed by some form of the verb querir in the

second hemistich is a formula which occurs extensively in the 129 twelfth-century chanson de geste.^ It is interesting that both

SB and BC have an analogous egression for Mohammed ■which may be placed in the mouths of Saracen characters. There are many shorter phrases which there is no need to list here in complete verses, such as Par (uor.oour) Dieu (BC: 201, 2 3 6 , 939, 946, 1111, H30,

2576), Par (xaor) Mahom (SB: 1954; BC: 610, 618, 704, 1780) and Par

Mahonmet (BC: 60 5 , 1764, 2947).

A very common pattern, ■which is definitely formulaic, is the oath upon the body of a certain saint.

An cors seint Semion. SB 119b Et le cors seint Gervais. SB 536b Par le cors saint Germain. BC 120b par le cors saint Symon. BC 167b Le cors saint Lazaron. BC 178b Par le cors saint ELoi. BC 231b Par le cors saint Climent. BC 556b Par le cors saint Denise, BC 976b Le cors sainte Marie. BC 1525b Et le cors saint Denise. BC 1639b Pour le cors saint Rani. BC 2127b Et le cors saint Denise. BC 2834b Par le cors saint Ctaer. BC 3051b Par le cors saint Climent. BC 3101b Pour le cors saint Hylaire. BC 3572b Et li cors saint Denise. BC 3786b

The name of the saint is always found at the rime. This is a clue to the way the poet manipulated this system: the words of the formulas can remain basically the same except for the saint* s name, which is changed according to the rime. Thus for laisses in -ise Adenet always

"Sur 1 * Originalite de Raoul de Cambrai." La Technique litteraire des Chansons de Geste: Actes du Collogue He T.iage (Septembre 1957) (Paris: Society d*Edition "Les Belles Lettres," 1959;, p. 54. M. Jodogne fails to cite, among his examples of the formula Par cel apos­ tre. SB: 2291, 2296, 2487. 130 employs saint Denise, and in the two laisses in -ent he uses saint

Climent. Lazaron. Semion and Svmon are all found in laisses in -on but not -under the same syntactic conditions, the last two being conqolements of prepositions. In BC: 178 there is no need for the word oarr which is usually found before le cors in these phrases, because the locution .iurer le cors is used there, and so the name of the saint must be three syllables long and 5vmon is excluded.

This set of formulas may be said, then, to show perfect economy since under the same conditions of meter and rime the same formula is used. Note that this formula is found 14 times in BC, but only twice in SB.

The short expression par foi is found only in SB. where, in the entire poem, there are eleven occurrences. Only one of these, however, is in the part which is imitated by Adenet, SB: 1834.

A longer version of the same idea,Foi cue ie doi . . . , is found frequently in BC and once in the first part of SB.

Mes, par la foi que doi. SB: 119a Par la foi que vous doi. BC: 226b Car par la foi que doi, BC: 357a Foi que doi Dieu porter. BC: 1440b Mais foi que doi Jhesu. BC: 1684b Mais foi que doi Mahon, BC: 2265a, 2302a Foi que je doi, fait il. BC: 2442a Foi que doi Dieu, fait il. BC: 2487a Foi que je doi ma mere, BC: 299*6a, 3088b Car foy que doi Guillaume. BC: 3081a Mais foi que doi mon oncle. BC: 3H4a Foi que doi s§int Symon. BC: 3354b

This set of formulas is similar to the system Par le cors . , , , but it is basically a first-hemistich system whereas the latter is always found in the second hemistich. 131 A whole-verso group found exclusively in BC consists in an oath in the first hemistich and an epithet in the second.

Par cel seignor, fait il, qui tout puet gouvemer. BC: 3064 Car, par cel saint Seignor qui tout a a baillier. BC: 3^ * Car, par cel saint Seignor qui nous fista s*ymage. BC: 3558 Par cel seignor, font il, qui fist ciel et rousee. BC: 3890

In one oath the knight swears upon his own head rather than by any saint or god.

Par mon chief, ce dit Bueves. SB: 913s Bien le sai, par mon chief. SB: 2249b Par mon chief, dit li dus, SB: 2281a Par mon chief, dit Girart, SB: 2295a, 2498a, 2519a Par mon chief, dit la bele. SB: 2380a Par mon chief, dist li dux. BC: 3049a

This formula is predominantly from SB, and the one occurrence in BC

is an imitation of SB: 2281a, both verses being found at the same O point in the two narratives.

l-'any of the cries uttered by knights on the field of battle

are invocations of aid from God or requests for the damnation of the

enemy. These two classes of expressions tend to have parallel forms.

Que Jesu beneie.9 SB: 1260b Cui Jhesus benei'e. BC: 636b

Qui Damedierus maudie. SB: 639p, 99Qp Jhesu te maleie. SB: 685p Que Kahomet maudie. SB: 1130b Qui Damedieus maldie. SB: 1148b Qui Jesu maleie.9 SB: 1269b Damedieus les maudie19 SB: 159^p Li cors Dieus les maudiel° SB: l630p

^See the treatment of similar epithets above, p. 5 6 .

^Cited by Roll, Untersuchungen . . . , p. 78.

^These phrases are spoken by the poet rather than by one of the characters. 132

Damedieus les maudiej10 - BC: 28%), 1521b Maudite soit tel bestej BC: 79%) Cui 11 cors Dieu maudie.10 BC: 1828b Cui Hahonmes maudie. BC: 2733b

And to these might be added:

A Dieu benexjon. SB: 245b, 2502b

Benediction can thus be quickly converted to malediction, and Christ to Mohammed, for the two creeds and various situations.

A set of similar parallel wishes places Damedieus and Hahonmes in corresponding positions.

Mahomet le confonde. SB: 1557p Hahonmes vous confondeJ BC: 426p Hahonmes le confonde, BC: l6l4a, 1781a Damedieus les confonde.10 BC: 1785a, 1745p Damedieus me confonde. BC: 3377a Damedieus les conduise. BC: l697a, 3520a Damedieus les consaut.10 BC: 3901a

The nnn^iisa and consaut formulas nay have been introduced because

of the similarity- of the first syllable of these words with confonde.

At any rate this series of formulas has the same configuration as the malexe-benexe set.

The remaining optative agressions are divided between two forms. The first of these is analogous to the two sets just dis­

cussed, but employs a variety of verbs, such asdoner:

Qui Dieus doint pute estraine.10 SB: 99b Dieus te doint raarementj SB: 411b Dieus te doint encombrierj SB: 633b Dieus te doint bone viej SB: 1983p Qui Dieus doint encombrier. SB: 2691b Dieus nous doinst si errer, BC: 80b Cui Dieus doint male estraine. BC: 153b

10These phrases are spoken by the poet. 133 Dieus • • • doinst k'encor vous revoie, BC: 546-47 Cui Dieus doinst mol dahe.ll BC: 1699b Cui Dieus doinst encombrier.il BC: 2881b

Damedieus les guerisejH- SB: l454p Cil sires le garisse. BC: 132a Dieus garisse Aymeri. BC: 153a Dieus les garb d ,encombrierjll BC: 373b, 1753b Damedieus vous garisse, BC: 1196a Que Dieus vous soit garans. BC: 2498b crayenter:

Cui li cors Dieu cravent. 11 BC: 271b Damedieus les craventJ BC: 563b Damedieus la craventejH BC.: 792b l’iahonmes te craventej BC: 1363b Kahonme s me craventej BC: 2237b Dieus les puist craventer! BC: 3052b Damedieus me cravente. BC: 3120b, 31^2b aidier:

Or lor soit Dieus aidisjH BCs 952b Or soit Dieus et sa mere crestiians en ayue.H BC: 204-7 Que Dieus lor soit aidis.H BC: 2818b Dieus lor soit en ayde.ll BC: 3868a Cui Dieus soit an ayue.H BC: 3929b Dieus lor soit an aye jl l BC: 394-Tp panser:

Or pant Dieus d'Aimeri. 11 SB: 100a Or pant Jesu des contes^H SB: 536a Dieus pant de Guielin."21 SB: 1273a Or pant Dieus de Girart.H SB: 2088a Dieus penst dou repairier.H BC: 360b

The second optative pattern is couched in the form of a conditional

clause, but the thought is "If only God help rae,M so that it can

•^•These phrases are spoken by the poet. 134 properly be classified as a -wish. Some of the same verbs as occur in the pure optative are also found in this f o m :

Se Damedieus ce done, SB* 1891a Se Dieus li donne vie,I2 BC* 1266a Si me doinst Dieus sante. BC: 3504b . . . Si me gart Mahons. BC* 255** Se Dieus nos veut aidier. SB* 908b S'aidier nous veut Mahons. BC* 2245a

Other verbs which are found only in this conditional form are* fairo*

12 Si cil sire nel fet. 12 SB: 280a, 484a, 1122a Se Damedieus nel fet. SB: 564a, 2585® 12 Se Gesu Crit nel fet. SB* 2511a panser*

Se Damedieus n'en pense.-1-2 SB: 112%, 1705p, 2513P Que se Jhesus n*en pense. BC* 3^5a Se Damedieus n'en pense. BC: 417a, 1919a Mais, se Jhesus n'en pense. 12 BC: 481a, 3786a Se n'en pense cil sires.^ BC* 520a Ara, se Dieus n'en pense^12 BC* 1054a Se cil sires n'en pense. BC* 3640a Mais s'or n'en pense Dieus.12 BC* 3706a plaire;

Se Dieu plaist et ses sains. 12 BC* 6a Se Dieu plaist, dist Gerars. BC* 2496a Se Dieu plaist, aigranent. BC: 3123b and, finally, sauver*

Se Dieus sauve mon cors. BC* 197a, 358a Se Dieus sauve Aymeri.. BC* 678a Se Dieus sauve Buevon.*12 BC* 1796a / IP Mais, se Damedieus sauve Gerart au cuer sene. BC* 2058 Si me puist Dieus sauver. BC* 3057b Si me soit Dieus sauvere. BC; 3091b Si me consaut Mahons. BC* 3604b

^These phrases are spoken by the poet 135 There are other optative expressions of which only one example can be found, and these have been omitted from this analysis.

When the request for help has been granted by God the charac­ ters do not fail to express their gratitude,

Damedieu en mercie. SB: 1054a Mahonmet en mercie. BC: 606b Dieu en ont graciie. BC: 1409b Damedieu en gracie. BC: 1639a Mahonmet en graci. BC: 2138b Damedieu en gracient. BC: 2834a Dont Damedieu gracient. BC: 29l4p

Other Miscellaneous Expressions of Discourse

In addition to the verbs of speaking discussed up to this point, some others which do not open discourse are used formulaically.

Conmander:

Que ge le vos conmant. SB: 1500b Que ge le li comment. SB: 1504b Que li ot conmande. BC: 2126a contralier:

Si l*a contralie. SB: 2239b, 2625b Avez contraloie. SB: 2330b Et me contraliies. BC: 653b deviser:

Tres bien li devisa. BC: 743b, 3234b amentevoir (to recommend):

Lor a amenteue, BC: 913b Li ot amenteue. BC: 1272b menacer:

Moult manace Buevon. BC: 1650a Moult menacent Buevon. BC: 1786a 136 tesmoignier:

Le peust tesmoignier. BC: 384b, 2871b Bien le puls tesmoignier. BC: 2340b

One speech expression found only in SB and unrelated to the others is truer (mar) fusiez onaues nez.

Buer fusiez onques nez. SB: 730b, 740b, 823b Mar futes onques nez! SB: 839b Mar fust ele onques nee| SB: 1186b Que mau fusent il ne! SB: 2699b

Both expressions are judgments pronounced by one character on another, one for being a good fighter (buer < bona hora), the other a bad one (mar <* mala hora).

When a character in SB is dismayed at seeing his companions perform an action he does not understand, he cries:

Que faites, chevalier. SB: 259a Que faites vos, biau fiz. SB: 325b Que fetes vos, Franjois. SB: 713s Que fetes vos, paiens. SB: 799a

One set of speech formulas is found in a rather special setting.

When hostile knights meet on the battlefield each one asks the other for his name, presumably not justto know idiom he is fighting, but to be able to report his own (expected) victory with all the details.

Conment as tu a non? SB: 130b, 1397a, 2450b Di va, conm’as tu non? SB: 389a Conment tu as a non. . SB: 1985b Conment avez vos non? SB: 2062a Conment as tu a non? BC: 623a, 3321b Conment avez a non? BC: 934a, 2670a

The answer to this is:

Fiz sui au due Buevon. SB: 1977a, 2066a Je suis fiz Aimeri. SB: 1401a Fiz sui a l'amirant. SB: 1987a Horn sui a la plus bele. SB: 2452a 137

Hies sui conte Aymeri. BC: 2547a Kies sui dame Emengart. BC: 2549a Fils 1*amirant d* Espaigne sui. BC: 2554a Fieus sui le due Buevon. BC: 2673a Fille sui l'amustant. BC: 2682a Horn sui a Malatrie. BC: 3323a or, with the verb apeler:

L»an m*apeie Buevon. SB: 390a L*en m'apele Girart. SB: 197^a, 2065a On m'apele Buevon. BC: 624a

This set of formulas is always connected to the motif of single

combat. 13 Usually it is followed by an attempted conversion on the part of both knights, but battle is always engaged and, inevitably, the Christian wins.

Another system always connected to the battle motif (but here usually the general melee) is that of battle cries. Host of these consist of the names of castles of French and Saracen leaders.

Aymeris s'escri’a: "Monjoie la Charlon." SB: 228 NerboneJ Dieus aide? SB: 299a Puis escrie: **Mon;ioie, aidiez, seinte Marie, Douce virge pucelej" SB: 1291-92 H escrie: "Monjoie!" aant celui vit feni. SB: 1328 Aidiez, sainte Mariel SB: 1330p II escrffa: "Monjoie! dus Bueves, car m ,aidiez!M SB: 2160 Bueves cria: "Monjoie!" et Renier: "Monpancier2 " SB: 2165 Et Clarion: "Vaudune!" SB: 2l66p II escrie: "I-lonjoie! chevalier, car m»aidiez." SB: 2617 Qant il 0 1 : "Monjoie." SB: 2662a "Nerbonne" a Aymeri oyssies haut huchier, BC: 374- Et "Monjoie" a Guillaume, cui chascuns doit prisier, BC: 375 "Conmarchis" a Buevon, a Renaut '•Montarmier," BC: 376 Senses crie "Borgoigne" et Ponces "Konpellier" BC: 377 Et Navaris 'Toulouse," ou il n ’ot k»ensaignier. BC: 378 Crier oant "Kerbone," cele part sont tome. BC: 449 "Valdune" lor escrie, de bien faire haities. BC: 1185

*^Gne exception is Hon sui au roi Yon, spoken by the messenger in SB: l4-3a. 138

A l'abaissier des lances ont "Nerbonne11 escrlee. BC: 1709 "Tabarie" escria au fuer de sa contree. BC: 1725 Francois crient •flfonjoie," l*ensaigne saint Denis, BC: 2800 Et Clarions "Valdune," li nouviaus convertis. BC: 2801 Eknni les Sarrazins et "Valdune" huchier. BC: 2870 Quis escrie "MonjoieJ Feres, franc damoisel." BC: 3738 Joffrois escrie "Angiers" et Gerars "Conmarchis," BC: 3775 Et Hunaus "Saint Malot," le cri de son pays, BC: 3776 Et Renaus "Montaimier," qui preus fu et hardis, BC: 3777 Et Navaris "Toulouse." BC: 3778

The SB battle cries occur in isolated verses, but Adenet tends to group several together, perhaps to give the impression of the gathering impetus of the Christian attack. All of the cries are from the Christian side except for "Tabarie," the cry of Corsolt,

"Valdune" is a pagan cry too, but by the time Clarion gets into battle he has been converted to Christianity. The patterns for these battle cries are rather loose, -except that the cry itself is usually given at the caesura and that the verb escrier is often employed.

Three hundred and sixty formulas of discourse from Le Siege de

Barbastre (approximately 133 per 1000 verses) and 3^7 from Buevon de Comarchis (approximately 87 per 1000 verses) are found in this chapter. These expressions all occur more than once in one or both poems in substantially the same form. Hemistichs which are alike except for the difference of a proper name or the designation of a person have been counted as formulaic. CHAPTER VI

BATTLE SCENES

Since the principal subject natter of almost all the chan­ sons de geste is warfare, it is not at all strange that many of the formulaic expressions denoting the actions of the characters should come under the heading of ’'expressions describing battles."

As with the formulas of discourse, an entire chapter will be con­ secrated to the exposition of the means the jongleurs had at their disposal for narrating the events of combat. The rest of the expressions of action will be taken up in the next chapter.

Of all the motifs- of SB and BC the one most commonly found is the single combat between two knights. The battle scenes of the Old French epic are seldom panoramic: when the poet sings of war his optic tends to focus on the details of the fighting rather than give the entire sweep of vision. In order to find out just how the individual hemistichs and verses, many of which are for­ mulaic, fit into a motif, the first part of this chapter will examine the various ways the jongleur has of presenting the dif­ ferent elements of the single combat motif. The rest of the chap­ ter will deal with the other formulas which occur in connection with battle, for the most part in general battle or messenger motifs.

139 140

In the first 2696 verses of Le Siege de Barbastre there are

19 passages in -which a French knight is seen fighting with a single

enemy warrior,"^ When these passages are examined together there

emerges a fairly well-defined stylistic configuration. P a r W y because of the traditional nature of the jongleur’s technique and partly on account of the limitations inherent in feudal warfare, the order of the elements in this motif is always according to a

set scheme. If the knights came upon each other during a lull in the fighting, their encounter begins when one insults the other in

some way and the angered opponent replies by charging with his lance lowered. If it is in the rush of battle that they meet this first step of provocation is not needed, but in either case they

spur their horses and strike each other with their lances. The

shield and armor of one of the parties (usually the Saracen’s) are damaged and if he is not killed at once he at least falls from his mount and the victorious knight dispatches him by sword on the ground, and then leads his horse away as booty. It sometimes

occurs that the vanquished combatant is a person of high rank in

the enemy camp, and when this is the case he is usually just wounded or, in any case, escapes death in order to continue play­

ing his important role.

Usually one or several of these elements will be missing,

-Hftr. 267-75, 771-75, 1202-14, 1217-2 0 , 1258-6 6 , 1274-78, 1279- 9 2 , 1299-1304, 1324-27, 1345-49, 1350-72, 1382-84, 1394-1443, 2007- 47, 2146-52, 2191-95, 2272-8 0 , 2591-95, 2612-1 6 . HBL but In only two cases out of the 19 in SB do the elements come In a different order. There are also two cases in which, something is added to the stock series of actions, but they will be treated below. ^

Each of the elements described has assigned to it, in the poet's repertory, certain foxrmlaic expressions. In order to emphasize the rigidity of these expressions, I will list them here together with the non-fonnulaic expressions, under the elements with which they are found in the motif of single combat.

INSULT

Fill a putain, Franyois, ja mes n'i guerirez. SB: 1203 Fill a putain, Franyois, vos n'i gueriroiz miej SB: 1282 Fill a putein, traitres, fel cuivert parjurez. SB: 135^+

ANGER

Qant Girart l'entendi, a pou n'est forsenez. SB: 1206 Qant Guielin l'oi, si n'a talant qu'il rie. SB: 1285 Et qant Girart le voit, sin a le sans raarri. SB: 1299 Et qant le Turs l'oi a pou n'est forsenez. SB: 1363 Girart l'a entendu, cele part est alez. SB: 1382 De mautalant rougi, n'i ot que corrocier. SB: 201^ Qant Girart l'entandi, n'i ot que corrocier. SB: 2136

SPURRING

Et Girart esperonne, si se feri el tas. SB; 1217 II broche le destrier, des esperons l'aigrie. SB: 1280 Le destrier a brochie, s'a la lance brandie. SB: 1287 II broche le destrierpor aidier son ami. SB: 1300 Et broche le destrier par endeus les costez. SB: 136^ H broche le destrier par endeus les costez. SB: 1^27 Et Bueves esperonne,qu'il n'i est arestez. SB: lk2Q II broche le destrier, destort le confanon. SB: 2008 142

XL broche le destrier^ des esperons d'or mier. SB: 2015 Einz ne fina de poindre juque a l'estor fomi. SB: 2192 Einz ne fina de poindre, si vint a Guielin. SB: 2196 Einz ne fina de poindre, si vint as Sarrazins. SB: 2201 Apnes brocha chascun son auferrant destrier. SB: 2272

BLOW OF LANCES

Vait ferir Aquilant un roi de Balesguez. SB: 268 Grant cop li done el chief, si con vint a esles. SB: 774 Et vet ferir Goudras qui venoit a£rez. SB: 1208 Si la premiere pointe vet ferir Barnabas. SB: 1218 S'ala ferir un Turc sor la targe florie. SB: 1261 Merveilleus cop li done sor la targe florie. SB: 1276 Vet ferir le gloton sor la targe florie. SB: 1288 Un paien vait ferir de son espie foxbi. SB: 1301 Far desus son escu fiert un Amoravi. SB: 1325 Devant sor son escu feri un Esclavon, SB: 1343 Merveilleus cous se donent sor les escuz dorez. SB: 1366 S'ala ferir un Turc que ill abat es prez. SB: 1383 Merveilleus cous se donent sor les escuz dorez. SB: 1429 Et Bueves fiert bien, con vassaus adurez. ' SB: 1432 Lors se donent granz cous enbedui li baron. SB: 2010 Molt se donent granz cous endui li chevalier, SB: 201? Iibanor fiert Girart en l'escu de cartier. SB: 2018 Et Girart feri lui a loi d'onme prisie. SB: 2023 Corsout de Tabarie ala ferir premier. SB: 214? Fiert Salot de Baudras devant an l'escu bis. SB: 2192 Et ge li refer! sor l'escu de cartier. SB: 227? Sor l'escu de son col vait ferir Samuel. SB: 2592 Sor l'escu de son col vait ferir Galier. SB: 2613

PIERCING OF AHMOR

Eh l'escu de son col que il li est troez. . SB: 269 Eh l'escu de son col qu'il est frez et troez. SB: 1209 Desoz la boucle a or li a frete et percie. SB: 1262 Que l'escu li perpa et l'auberc li esmie. SB: 1289 Si que desoz la boucle les ont frez et troez. SB: 1367 Molt sont fort li hauberc con il ne sont fausse. SB: 1368 Si que desoz les boucles les ont fraiz et troez. SB: 1430 Que les cengles desronpant, li poitraus est cassez. SB: 1433 Les auves estandirent, li freins est descirez. SB: 1434 De l'un chief juque en l'autre li fet fandre et percier, SB: 2019 Forz fu li jazeranz, que maille n'en ronpie. SB: 2020 Sor l'escu de son col, lez la boucle a or mier. SB: 2024 Les cengles sont ronpues de 1»auferrant destrier. SB: 2031

^Two additional examples of this foimula occur in general battle scenes. See below, p. 156. 143

Les enaraes e standout, s*alongnent 11 estrier. SB: 2032 Que l'escu de son col 11 a fret et froissie. SB: 2148 Fort fu 11 jazeranz, qui maille n'en ronpie. SB: 2149 De l'un chief juque en 1*autre tot le me porfendie. SB: 2274 Forz fu 11 jazeranz qu'il ne pot enplrier. SB: 2275 De l'un chief juque en 1 'autre 11 a fret etcasse, SB: 2614 Le jazerant del dos ronpu et detnaillle. SB: 2615

FALL FROM HORSE

Tant con hante 11 dure l'a abatu es prez. SB: 271 Tant con hante 11 dure del destrier l'abati. SB: 1302 S'ala ferir un Turc que ill abat es prez. SB: 1383 Tant con hante 11 dure l'abati mort sovin. SB: 2194 Que par desus la croupe le fis jus trebuchier. SB: 2278

DRAWING OF SWORD

Puls a trete l'espee bone, tranchant, foible, SB: 1265 Si a trete l'espee^ dont li brans fu letrez. SB: 1371

DEATH OF ONE OF THE COMBATANTS

La cervele en abat, mort chei desconfes. SB: 775 Et 11 paiens cha£, qui a mort fu navrez. SB: 1211 Que mort ne le trebuche en la lande neimie. SB: 1264 Que mort le trebucha, voiant la conpangnie. SB: 1277 Si soef l'abat mort qu'il ne bret ne ne cr£e, SB: 1290 Tel cop 11 a done que 11 mort l'abati. SB: 1327 El cors 11 mist la lance o tot le panoncel. SB-: 2593 Le cuer qu'il ot el vantre 11 a per mi tranchie, SB: 2616

TAKING THE DEFEATED KNIGHT'S HORSE

m a pris le cheval, si s'en est retomez. SB: 275 II sessi le destrier, s'a sa voie accoillle. SB: 1278 Et sessi le destrier par les resnes noez. SB: 1384 Einz sessi le cheval par la resne d'or mier. SB: 2046 Et Girart a sessi _le destrier de Castele. SB: 2050 Et sessi le cheval^ par la renge d'or fin, SB: 2195 Si en menai o moi le bon cheval corsier. SB: 2279

It is obvious from these examples that the .jongleur of Le Siege de

\ h i s formula also occurs in SB: 256, ‘where Guillaume is ready­ ing himself to receive the first Saracen attack.

^See also SB: 237, St sessi la contesse au pan del siglation. and SB: 1867, Si sessi la -pucele -par la renge d'or mier. Barbastre had a stock of phrases at his disposal so as not to find himself at a loss for words -when he was called upon to describe any of the standard single combats which are found in all chansons de gester and that an entire combat scene could be described with­ out the poet’s having to go beyond his repertory of formulas.

There are, however, two examples of this motif which deserve

special attention as they contain traits which do go beyond the

stock elements. In most of the single battles the pagan adversary is either left unnamed and simply designated by the words paxen,

Turc. Sarrazin or Esdavon. or else he is identified as a minor

character such as Aquilant de Balesguez (SB: 268), the amoraive

of Barbastre (SB: 772), li sires d ’Aumarie (SB: 1279), Otran de

Baleguez (SB: 1352), Corsout de Tabarie (SB: 21*4-7), Salot de Bau-

dras (SB: 2193) or Gaudras (SB: 1208). But the two major duels

of this section of the poem involve the two principal personages

of each camp: Bueves against the amirant of Sjpain, leader of the

Saracen a my (SB: 1389-1*44-3) and Girart against Libanor, his counter­ part in the pagan array and his rival for the hand of Malatrie (SB:

2013-4?). In the first of these encounters the amirant is introduced

by a description of him and his horse:

A tant vint 1* amirant sor un destrier armez; Ses chevaus fu toz noirs et si fu estelez Et couvert d'un diaspre menuement ovre. L* amirant sist desus rich eon ent acesrae; Einz plus riches adors n*ot hom de mere nez. (SB: 13^9-93)

There follows a series of questions and replies. The amirant • 1^5 threatens Bueves and asks him his name and place of origin. Bueves tells him and asks the same information in return. The Saracen then proposes that they become allies and conquer France together.

Bueves in turn requests the amirant to convert to Christianity, and this angers the latter into attacking. The usual formulaic elements come into play at this points the charge, the blows of the lances, the penetration of the amirant1 s armor and his fall from the horse. This latter element is given a comic turn however.

Li destriers chancela, del cop est encombrez; A genouz est venu, 1*amirant est versez, Li chies juque as espaules est el sablon entrez. Li amiranz fu si bleciez et estonnez, Einjois peusiez estre une huchiee alez, Que l*amirant seust ou endroit fust plantez, Et si home i acorent environ de toz lez, A force l»en tra3nent, n*i est mie remes. Tres par mi le sablon un arpant me sure, Haugre Buevon en fu li amiranz menez. (SB; 1435-^)

The structure of this scene has the traditional formulaic elements as its base, but it is prefaced by the exchange of names (a motif in itself which is found in other parts of the poem^) instead of the simple insult element, and is followed by the strange sight of the amirant sticking head first into the sand.

The duel between Girart and Libanor is in some ways parallel to the one just discussed. Malatrie has led her fiance Libanor before the walls of Barbastre, ostensibly to enjoy the sunshine, but ac­ tually to tempt Girart, with whom she is in love, to leave the castle and challenge Libanor. This takes place exactly as she has

^See above, pp. 136-37* 146 planned and the exchange of names, attempts at conversion, formulaic

charge and blows follow in their expected order. Libanor strikes first and smashes Girart * s lance, but the halberc is strong enough to hold and Girart has a chance to return the blow. The effect of this stroke is described in detail.

Et Girart feri lui a loi d ,onme prisies Sor l»escu de son col, lez la boucle a or mier, Entre les catre clous li asist son espie, A la lance qu*iert grosse et roide de pomier, Et au fer et au branc et au cors del destrier, L ’anpaint Girart a force a loi de chevalier; L'escu a tot le braz li fist au cors plesier, Soz le piz del cheval fist le poitrail trenchier, Les cengles sont ronpues de l»auferrant destrier, Les enarroes estandent, s»alongnent li estrier. (SB: 2023-32 )

The enemy hero is thrown off his horse, with much the same result as in the battle between Bueves and the amirant.

Et Girart 1•enpaint outre par la croupe derrier, La sele entre les cruises fu el chanp envoiez, Si que le coinz de l*iaume fu en terre fichiez, Les geribes envoia contreanont vers le ciel. Einsi pres de la rive dant Girart l»abatie. Au relever que fist, que il se dut drecier, One ne sot mot li rois, en l*eve est trebuchiez, Si que l*eve li clot desus li chief plein pie. Si ne fust une estache, Libanor fust noiez, Que Sarrazins i mistrent por poisons engingnier, Ou au meins s»est repris et contreanont dreciez; A force s'est levez et as meins et au piez. (SB: 2033-44)

This comic scene of a boastful knight being sent legs flying onto the sand and then, stunned, groping his way into the fishpond and almost drowning under the eyes of his disdainful fiancee is cer­ tainly too out of the ordinary to be a part of the formulaic reper­ tory. It is an embellishment of slapstick humor that would not be found in the earlier epics such as Raoul de Cambrai or Gormont

et Isembart. These two passages show how the poet of Le Siege de Barbastre could add his own craftsmanship to the traditional treatment when he felt this was called for by the importance of the subject* The traditional poet can be just as inventive an artist as his literary counterpart.

As has been mentioned Buevon de Conmarchis recounts in its

3947 verses the events contained in approximately the first 2696 verses of Le Siege de Barbastre. In BC the single combat between 7 knights occurs 20 times, but these scenes do not all present the same neat order as the single combats of SB.

In the latter poem, it will be recalled, many of the encounters begin when one knight insults another and angers him into attacking.

Sometimes the insult is given by both combatants. There are certain formulas attached to both of these elements: for the insult, Fill, a puta in. and for anger A nou n»est forsenez or an equivalent expres­ sion M'i ot que corrocler. In BC the formula Fill a putain occurs three times in connection with single combat motifs (BC: 1118, 1885,

2009). In none of these passages, however, is it followed by the expression of anger which one could expect from the SB pattern, and this is true in spite of the fact that one of the BC passages is based on a scene from SB (BC: 1885, SB* 1203) which contains the two formulas in their ordinary relationship. In SB, on the other hand, the formula F^ll a putain occurs three times in the part

7Vv. 257-60, 261-62, 290-93, 306-18, 328-36, 337, 338-42, 434- 39, 1117-22, 1603-10, 1718-25, 1726-31, 1867-1906, 1998-2015, 2586- 2652, 2725-66, 2843-59, 3725-31, 3934-35, 3936. 148

imitated by Adenet, and each time (SB: 1203-06, 1282-85, 1354-63)

it is followed by a formula of anger. The step-by-step development

of the battle scenes of the earlier poem, which this configuration

of action and reaction illustrates, is abandoned in Adenet*s poem,

and the result is not only a less rigid style, but greater subtlety.

It would seem that Adenet *s audience did not require that each

logical step of the narrative be pointed out in an obvious manner.

Further investigation into the formulas which express provo­

cation and anger in both single combat and general battle scenes

shows the following: the formula A pou n'est forsemez is found

four times in the first section of SB, but does not occur at all

in BC; an alternate formula N*i ot que corrocier occurs six times

in SB. but not once in BC; Fill a putain is used six times in SB.

three in BC. Thus while Adenet uses the last formula sparingly

(in three verses out of 3947 as compared to six out of 2696 in SB), he has decided not to employ the first two formulas at all. He has

chosen to revise his model (Pour ce qu*est mal rimee. la rime amen-

derai. BC: 21), and his omission of certain phrases must have resulted

from his principles of remaniement. His practice in the use of

stock expressions in single combat scenes is much freer than that

of the author of SB. and we can say for certain, at least in this

case, that he did not feel obliged to retain the traditional for­ mulas. 'Whenever the anonymous poet wishes to present one of his

characters as being angry, he uses formulaic expressions, without a

single exception, while Adenet employs either formulaic or 149 non-formulaic phrases or oven has. the reader understand that his

character is angxy -without saying so. Not only are the formulas

"which we have examined little used by Adenet: there is no other

equivalent system to replace them.

To return to the comparison between single combat scenes in the two poems, another fairly constant element is the formula for

spurring a horse. The following verses are the ones which occur in

BC to signify this action.®

Corsols de Tabarie vint poignant une sente. BC: 290 Des esperons a or a le cheval hurte. BC: 436 A tant es vous Corsolt poignant de randonnee. BC 1718 Galdras broche Ferrant par andeus les costes. BC 1882 Vint poignant vers nos gens, si les a escries, BC 1884 Lors broche le destrier com vassaus adures. BC 1894 Lors broche le cheval par andeus les costes. BC 2533 Le destrier esperonne des trenchans esperons. BC 2586 Les hanstes ont brandies et les chovaus brochies. BC 2595 Guis de Conmarchis broche le bon corant destrier. BC 2845 Qui les ve£st andeus des esperons brochier. BC: 2850

It is evident that the action of spurring a horse is not expressed

in traditional formulaic language by Adenet, even though it is in

the poem which he is imitating. In the anonymous poem eight of the

expressions for spurring a horse contain the formula Broche le des­

trier. whereas Adenet only uses it in two verses. Further inves­

tigation reveals that Adenet employed this formula only three times

in all of BC, the other instance being in v. 3646: Malaquins de

Tudele le bon destrier brocha.

The impression that one gets from this comparison of the two

styles is that Adenet is consciously trying to vary his means of

Q See also the general battle and message scenes, pp. 156-60 150 expression. He seems to be careful not to say the seme thing twice, and in fact only two of the 22 hemistichs quoted above for the idea of spurring are identical. The author of SB. on the contrary, is not only unconcerned with the problem of varying his style, but ha.trftna.Tiv uses identical expressions for identical ideas.

Corroboration for this view can be found in the following list of verses from BC, arranged in groups which correspond to the vari­ ous elements of the single combat scenes of SB.^ The phrases for the elements of INSULT and SPURRING have already been treated.

BLOW OF LANCES

Puis refiert si un autre que sor l'arjon l'adente. BC 293 Tel coup donne un paien que del bras le fait moing. BC 311 Si en refiert un autre qui fu nes de Garsoing. BC 312 Si le fiert que la teste li fent desques el groing. BC 315 Tel coup donne Espinart en haut deseur la broingne. BC 333 Gerars fiert si un autre la teste li rooingne. BC 337 Fiert Salot de Baudras, si bien l*a assene. BC ^37 Si en fiert l'amirant que mort le trebucha. BC 1122 Et Guillaumes le fiert, com aigres d'ire espris. BC 1605 Si fiert un Nerbonnois sor la targe listee. BC 1722 Corsolt feri sor 1 1iaume a la cercle doree. BC 1727 II et Galdras se fierent sor les escus boucles. BC 1896 Si fu ferus Galdras sor l'escu d*or bendes. BC 1899 Sorb rant ala ferir, si bien l'a conseii. BC 2013 Si grans cops s'entredonnent es escus vemissies. BC 2599 Et li paiens le fiert en l'escu qui flambie. BC 2750 Si grans coups s'entredonnent, msntir ne vous en quier.. BC: 2852 Un Sarrazin feri c'on nommoit Manuel, BC : 3727 Si fiert un Sarrazin k'a celui coup le tue. BC : 3935 Et Gerars fiert un autre si qu'a terre le rue. BC : 3936

PIERCING OF ARMOR

Que il li a fendue et la brongne fausee. BC: 1723 Ii espiels Galdras est par piece tronconnes. BC: 1897 Kais li espiels Gerart n'est ne frains ne quasses. BC: 1898 Qu'il est frains et brisies et le haubers fausses. BC: 1900

^See above, pp. lJH-^-3* 151 Que l'escu a percie, l'auberc li a rompu. BC 2014 Que chascuns des escus est fendus et percies. BC 2600 Fort furent li hauberc, nus n*en est desmaillies, BC 2601 Mais chascuns des espies est dusk'as poins froissies. BC 2602 Li poitraus dou cheval fu rous et depecies, BC 2606 Et les cengles rompues, qui n'estoient pas vies. BC 2607 Si qu'il li a percie, mais n'a pas desmai.ll.ie BC 2751 La brongne, ains se tint bien, et la lance est brisie, BC 2752 Si que par maint tronjon fu dusqu'as poins froissie. BC 2753 Moult fu fors li haubers quant il n'est dessartis. BC 2776 Que les targes percierent li fer trenchant d'acier. BC 2853 Fort furent li hauberc, ne porent desmaillier, BC 2854 Mais ambesdeus des lances font dusqu'as poins froissier. BC: 2855

FALL FROM HORSE 'I Que desus la verdure en fist dou cors present. BC: 262 Amaurri de Senlis a terre nous cravente. BC: 292 A ses piez l'abat mort deseur le marbre bis. BC: 971 Que dou cheval a terre chay tons esbahis. BC: 1610 Tout abat en un mont seur l'erbe defoulee. BC: 1731 Si que mors a la terre est cheus enverses. BC: 1902 Si que raort a la terre 1 • abat tout estendu. BC: 2015 Trestout droit les la rive che£ moult corroucies. BC: 2611 Que tous pasmes chei sor l'erbe qui verdie. BC: 2760 Cil chey mors a terre desouz un arbrissel. BC: 3731 Et Gerars fiert un autre si qu'a terre le rue. BC: 3936

DRAWING OF SWORDS

L'amirans traist l'espee, mais Bueves s'avanpa. BC: 1120

DEATH OF ONE OF THE COMBATANTS

A ses piez l'abat raort deseur le marbre bis. BC: 971 Si en fiert l'amirant que raort le trebucha. BC: 1122 Que maint paien en a l'ame dou cors sevree. BC: 1716 Si k'a cel coup li est l'ame dou cors sevree. BC: l?2^i

TAKING THE DEFEATED KNIGHT'S HORSE

Un cheval a saisi qu'il trouva estraier. BC: 392 Un cheval a saisi par la resne doubliere. BC: 411 II saisi le cheval corant et abrieve. BC: 439 Par les resnes saisi le bon cheval Morel. BC: 2640

There are some similarities among certain of these verses: Tel coup donne plus name, Si grans coups s'entredonnent. the verb ferir 152 or referir with various complements in the first group of verses; dusqu»as poins froissier in the group PIERCING OF ARMOR; si que mors a la terre in the FALL FROM HORSE verses; l»ame dou cors sevree in the descriptions of DEATH OF ONE OF THE COMBATANTS; and finally un cheval a saisi in the last group. But these resemblances are neither as frequent nor as close as the formulas of 23 such as escu de son col which occurs six times in the single combat scenes of

SB but which is not to be found in BC; vet ferir plus name, five times in SB but absent from BC; targe florie. three times in SB but absent from BC. merveilleus cous se donent. three times in SB; les ont frez et troez. twice in SB; tant con hante li dure, three times in SB; fort fu li .jazeranz. three times in SB; de l*un chief .luque en 1*autre, three times in SB.

Although Adenet *s poem does contain some formulas in these scenes, they are not basic to the construction of his tale and he can get along without them as is shown by the variety of ways he has of saying the same thing.

As has been seen, most of the basic actions of the single combat motif in SB are also present in BC. But in addition to these there are some new elements which Adenet usually inserts at the beginning of an encounter between two knights. He often introduces one of the combatants with a phrase which shows him noticing a pre­ vious action by his adversary. A formulaic pattern, however loose, is constituted.^®

^®See above, pp. 3^35 and below, pp. I6I-63. 153 Quant Guillaumes dJOrenge vit si grief le besoing. BC: 306 Quant Bueves ot veus les cuvers desfaes. BCj I867 Quant Gerars vit Ferrant sel reconnut asses. BC: 1887 Quant la gens 1 *amirant ont Buevon perceu. BC; 1997 Quant Limbanors l'entent, plus iries ne fu hons. BC: 2580 Quant Gerars l'entendi, s'a sa resne sachie. BC: 2735

Another way Adenet has of presenting a knight is to show him coming into the enemy's proximity.

Gerars et Guielins vinrent premierement. BC: 258 Corsols de Tabarie vint poignant une sente. BC: 290 Vers l'anrustant de Cordres est Guillaumes guenchis. BC: 1603 Contre 1 'amirant vient com eil qui desirrier . . . BC: 2846 Devant les autres vint Guis, l'escu en chantel. BC: 3725 These verses do not seem to present any formulaic elements, but rather are quite varied.

There is a tendency in BC to mention the combatants' > and weapons before the duel; this is not usually the practice of the anonymous poet of SB.

Tint Joiouse s'espee ens en son destre poing. BC: 309 Guillaumes tint Joiouse, moult fierement l'enpoingne. §C 331 Lors embrace l'escu, tint l'espiel noale. BC 434 La targe ot enbracie et tint l'anste levee, BC 1720 A un fer moult trenchant dont l'alemele ert lee. BC 1721 Bien li sist li escus et li hiaumes games, BC 1876 Eh son poing tint la hanste dont li fers fu dores. BC 187? Et a brandie l'anste dont li fers fu quarres. BC 1895 Gsntement fu a m e s et d'auberc et d'escu BC 2003 Et d'espee trenchant et d'espial esmolu BC 2004 Et de hiaume dore a pierres de vertu. BC 2005 L'escu prist quant ses hiaumes fu a point refermes, BC 2530 Puis saisi son espiel dont li fers fu quarres. BC 2531 Les hanstes ont brandies et les chevaus brochies. BC 2595 L'escu tint enbracie-*-*- et la lance enpoignie. BC 2726 L'escu joint vers le pis et a brandi l'espier. BC 2848 Et la lance sor fautre a loi de preu danzel. BC 3726 Guis de Conraarchis tint el poing 1'espee nue. BC: 393** But again there is little formulaic material present in this element,

11See also BC: 2387a, Et l'escu embracier. 154

©von though it is found in ten of the twenty single combat scenes of BC.

The sparseness of formulaic diction in Adenet*s battle scenes

can perhaps be best illustrated by juxtaposing parallel passages from the two poems. The two texts presented below are single com­ bat scenes in which Girart and Gaudras are the protagonists. Each phrase which is to be found again in substantially the same form

elsewhere In the same poem will be underlined. Differences of

spelling, substitution of ••utility" words and changes in proper names will be permitted. Of course the more verses one has to com­ pare, the greater the chance of finding similar phrasing; so in order not to prejudice the comparison in favor of the earlier poem, which is same 7000 verses long, I have underlined only those SB hemistichs which occur again in the first 3947 verses of that poem.

Thus the same number of verses is used for each wo lie.

SB; 1202-11 BC; 1884-1902

Atant es vos Gaudras. qui Vint poignant vers nos gens, si s,est haut escrffez. 1202 les a escries: 1884 "Fill a putein. Francois, .ia "Fill a putain. fait il. certes mes n*i guerirez. 1203 n*i garires; 1885 A tart seroiz hui mes dedans Se vous estes vassal, les visa- les fexmetez. 1204 ges tomes." 1886 Cil qui hors vos a mis sera desheritez." 1205 Qant Girart l'entendi. a pou Quant Gerars vit Ferrant. sel n'est forsenez. 1206 reconnut asses; 1887 Or sachies que de lui fu fo la­ ment' goulouses. 1888 "Ja Dieu, fait il, ne place, qui en crois fu -penes. I889 Que je soie ja mais ne creus ne ames I89O Ne de bele pucele baisiez ne acoles, 1891 155 Se tout maintenant n*est mon cors aventures 1892 Par quoi iert, se je puis, Fer- rans reconquestes." 1893 H I a ganchi Ferrant qui es- Lors broche le destrier com vas- toit abrivez, 1207 saus adures 1894 Et a brandie l*anste dont li fers fu quarres. 1895 Et vet ferir Gaudras qui II et Galdras se fierent sor les venoit ai'rez, 1208 escus boucles, 1896 Li espiels Galdras est par piece tronponnes 1897 Mais li espiels Gerart n*est ne frains ne quasses, 1898 Si fu ferus Galdras sor l^escu d»or bendes 1899 Eh l*escu de son col qu*il Qu*il est frains et brisies et li est frez et troez. 1209 haubers fausses, 1900 Si crue par mi le piz est li Panni le gros du cuer est li es­ aciers coulez, 1210 piels passes 1901 Et li paiens chai, qui a mort Si cue mors a la terre est’ cheus fu navrez. 1211-*-2 enverses. 190213

ELeven of the twenty hemistichs of the SB scene are foximilas (55$),

■whereas only seven of the 38 heraistichs from BC (approximately 18$) are found elsewhere in that poem. The evidence of these two passages is overwhelmingly in favor of the view that the two poems belong to different methods of composition.

Of course it is possible that some of the heraistichs of BC which

1202a: a tant es vos Girart (107, 1449), a tant es vos Guil­ laumes (208), a tant es vos Buevon (384). 1202b: si s»est haut es- cr£ez (3378). 1203a; fill a putain, Franfois (766, 1282), 1205ab: cil qui hors vos a mis sera desheritez (767)0 I206ab: qant Girart l»entandi, a pou n*est forsenez (1932), qant I'antandi Girart, a pou n*est forsenez (3032). 1207a: il lor ganchi Ferrant (3079). 1209ab: que l*escu de son col li a fret et troe (2793). 1210ab: si que par mi le cors li est l'espie coruz (290).

^ S S S a : fils a putain, fait il (1118, 2009). 1887a: quant Gerars vit son pere (2807, 3010), quant Bueves voit Gerart (2987). 1889b: qui an crois fu penes (1086). 1894b: com vassal adure (1 2 0 1 ). 1895b: dont li fers fu quarres (2531). 1897b: par pieces tronfon- nee (1712). 1902a: si que mort a la terre (2015). 156 are not underlined in the example are formulaic and would have been recognized as such had the complete text of Adenet*s poem survived.

But in using the same number of verses from which to draw evidence for the fomilaic nature of the passages I have, so to speak, given the later poem the same handicap as the earlier one. In fact there are some phrases in the SB passage which are not underlined but which do occur again in the section of the poem after verse 394-7

The actual number of fomulaic heraistichs in the SB passage is four­ teen out of twenty, but for purposes of comparison the figure of

55# is more accurate.

Expressions which are consistently found in general battle scenes are relatively few in comparison with the large number in single combat motifs.

The following have to do with acts of horsemanship,

A coite d*esperon. SB: 116b, 225b, 248b, 1343b BC: 1372b

Brochant a esperon. SB: 125b A esperon bro chant. SB: 167b BC: 685b A esperons bro chant. BC: 371**

Et bro chant les destriers. SB: 1216a II brochent les destriers. SB: 2214a

Les chevaus esperonnsnt. SB: 1692a

VJhile these expressions are the equivalent to, and sometimes identical with, the foimulas of spurring in single c o m b a t , ^ there are others

1/f1204b: dedenz les fermetez (6997). 1207b: qui estoit abrivez (4-101). 1208a: et vait ferir Girart (4-388), et vait feiir Brehais (5386).

■^See above, pp. 14-1-4-2. 157 not fcrund in duel scenes "which convey the urgency with which knights rush to the aid of their companions inthe melee,

Einz ne fina de poindre juque al'estor fomi. SB: 2192 Einz ne fina de poindre, si vint a Guielin. SB: 2196 Einz ne fina de poindre, si vint as Sarrazins, SB: 2201

Desi que a l*estour n*i ot resne tenue, BC: 1571 Desi que a 1'estour n'i fu resnes tenus, BC: 1937

During a battle it often happens that one knight becomes isolated or that the numbers become so unequal that one group has to flee,

Et Franjois les enchaucent. SB: 1235a, 1343s1 Et Sarrazin l*eschacent. SB: 124oa, 1294a, 209381 II enchauce un paien, SB: 1274a Que Sarrazin 1*enchaucent, SB: 2124a Francois les enchaucierent, BC: 3819a

It might be noted here that when the Saracens chase French knights the latter invariably escape or are reinforced, whereas "when French pur­ sue pagans they usually catch up with them and join battle.

The action of striking in general battle is rendered by more general expressions than those of single combat, atteindre a cop. en ferir a con and consuivre a cop.

Qui j'atendrai a cop. SB: 244a Cui "in ateint a cop. SB: 257a, 1266a Cui ill en fiert a cop. SB: 294a S*a cop vos puis ateindre. SB: 1426a Cui il consiut a cop. BC: 1574a

One formula of striking is used twice in SB as the conclusion to a

comparison.

Autresin con li fauz qui del pong est volez, De ferir es oisiaus qant est entalentez, Se fiert li cuens entr*aus, ne s’est aseurez. (SB: 264^66)

Ehtresi con li faus fiert es oisiaus raenuz, Se fiert li cuens entr*aus, que il n*est esperduz. (SB: 302-03)

In the first case it is Guillaume who is compared to a falcon, in the 158

second, Aymeri. The two comparisons come from succeeding laisses

(8 and 9) of S3, but they are in no m y parallel laisses since the first tells of Guillaume being caught in the middle of a swarm of

Saracens and of his rescue, and the second of how he fights after his delivery and of the capture of Buevon, Guis and Girart. It is possible that the jongleur had thought of using the comparison in /¥# laissa 8 , but then forgot afterwards that he had done so and pro­ duced it again forty verses later. In any case the proximity of the two occurrences is a point in favor of the theory of oral deli­ very and composition since this type of repetition would hardly occur in a literary work where the author could go back over his text and correct it.

Closely tied in with battle scenes is the motif of the arrival of a messenger who tells of a coming attack and rouses the armies to battle. The following formulas, which have to do with spurring, are like those already discussed in relation to single combat and general battle motifs.^

A tant es un mesage a esperon bro chant. SB: 1620

1-ialaquins de Tudele le bon destrier broeha. BC: 3646

The N»i ot resne tenue formula which occurs in general battle scenes

in BO17 is restricted to messenger motifs in SB.

Ehtreci c*a Barbatre n*i ot resne sachie. SB: 491 Ehtreci que a Sore n*i vost resne tirer. SB: 2411

l6Pp. 141-42, 149 and 156 17Above, p. 157. 159 The messenger is often wounded in breaking through hostile forces,

Et fu el cors navrez. SB: 92a, 126a Et fu navrez el cors, SB: 2651a Et fu navres oucors. BC: 148a Et moi navre oucors. BC: 184a

It is possible that the SB formulas of 92a and 126a were imitated by Adenet in BC: 148a and 184a, since the messenger scenes in which these occur take place under the same circumstances, that is as transitions from the introduction to the body of the poem. It is interesting that Adenet did vary his expressions from those of SB in this case.

Other maneuvers connected to battle scenes include the sending

out of sentry parties, the rescue of a knight in danger, and the

rallying of a body of knights for retreat. The following formulas

occur with these three motifs.

La nuit font l*eschauguete Sarrazin et Escler. SB: 2582 L*eschauguete fesoie de ci a l*esclerier. SB: 2636 L'eschaguete ferines de ci a l'esclerier. SB: 2654 L'eschargaite faisoient Sarrazin et Persant. BC: 3709

Qui vint a la rescousse a tot mil chevaliers. SB: 2642 Qui vint a la rescousse atout cinc chevaliers. SB:. 2659

Bu'eves sona un cor por sa gent ralier. SB: 2114 Mist un cor a sa bouche pour sa gent raloiier. BC: 349 Mist un cor a sa bouche pour sa gent raliier. BC: 1750 Et Bueves sonne un cor pour sa gent raliier. BC: 2887

The same idea of rallying is given in two verses at one point in

Adenet*s poem:

'Mist un cor a sa bouche, plus n*i est atargies. Jkitour lui s*est chascuns de sa gent rallies. (BC: 1981-82)

In this two-verse combination the second half of the first verse is

merely a filler. The normal Pour sa gent raliier could not be 160 utilized here by Adenet because of the rime in -ies. so he had to construct an alternate method which would place two past participles at the rime.

In calculating the number of formulas found in this chapter

I will not include epithets, which are treated in chapter 3, the formulas of discourse found under the INSULT element, which are included in chapter 5, or expressions of anger which are discussed more completely in the following chapter. Of the other expressions which are definitely formulaic, 67 are from SB (approximately 25 per 1000 verses) and 24 from BC (approximately 6 per 1000 verses).

In regard to expressions describing combat, then, SB is more formulaic than BC several times over. CHAPTER VII

OTHER ACTIONS

In this chapter will be treated all those actions of various kinds which the poets of Le Siege de Barbastre and Buevon de Con- marchis recount, except, of course, those discussed in the pervious chapter on formulas of combat and those involving discourse which are covered in Chapter V.

Among the elements of the single battle motif was one which described a character seeing or hearing that his enemy was doing something against his interests, and then becoming angry about this to the point of challenging him to a man-to-man combat. This pat­ tern, with a few changes, is employed in various other situations in the two poems. The interlocutors are not always knights from opposing armies and the outcome of the character* s perception of his enemy's doings is not always anger but may result in other reactions. When the sense of hearing is the vehicle of the per­ ception, the verb may be either oir or entendre. Oxr:

Cuens Aymeris l'oi*. SB: 333a Cuens Aymeris oi’. SS* 374a Li amiranz l'oi. SB: 417a, 1110a Malatrie l'oi. SB: 1625a

Quant 1'amirant l'oi, £©: 392a Quant il oy Buevon. BC: 651a 161 162

Quant Buevas ot ses fils, BC: 777* Qant Sarrazin 1* ofrent. SB: 803a Qant 1'ofrent paion. SB: 872a Qant Guialin l'of. SB: 1285a Qant la pueala of, SB: 1537a Qant Libanor l'of. SB: 1844a, 200?a

Et qant Buevas l*of. SB: 397a Et qant li Turs l'of. SB: 1363a

Entendre:

La contessa l'entant, SB: 360a Malatrie l'entant. SB: 1577a, 2347a, 2541a

Clarfon l'entendi. SB: 779a Clarfons l'entendi. SB: 924a Clarfons entendi. SB: 928a

Qant Clarion l'entant. SB: 625a Qant l'entant l'amustant. SB: 1690a Qant l'entant Libanor, SB: 1962a Quant l'entent Aymeris. BC: 188a Quant l'amustans l'entent. BC: 431a, 642*a, 686a Quant l'amustans antant. BC: 620a Quant l'entent l'amustans. BC: 626a Quant Clarfons l'entent. BC: 908a

Qant Girart l'entendi. SB: 1206a, 1932a, 2136a, 2332a Qant paien entendi rent. SB: 1331a Qant Buevas l'entandi. SB: 2111a, 2323a Qant Girart entandi. SB: 2491a Quant l'entendi dux Buevas. BC: 633a Quant Buevas entendi. BC: 670a Quant Buevas l'entendi. BC: 900a, 1082a Quant Franjois l'entendirent. BC: 1026a

Girart l'a entendu. SB: 1382a Qant ce a entendu. SB: 1558a Quant il ont antandue. BC: 912b

When the sense is sight, the verb is always voir:

Ou qu'il vit la contesse. SB: 356a Ou que il vit Buevon. SB: 627a Ou que il voit Franyois. SB: 809a Ou qu'il voit Malatrie. SB: 1525a, 1621a

L'amir ant voit Girart. SB: 1310a Et Sarrazin les voit. SB: 2215a 163 Malatrie les voit. SB: 2576a

Qant dus Bueves le voit. SB: 808a, 879a Qant ce virent paien. SB: 82§a Qant Clarfon les vit. SB: 860a Qant les vit Clarion. SB: 955a Qant Girart vit Ferrant. SB: 1050a Qant il vit Malatrie. SB: 1832a Qant le vit Malatrie. SB: 1945a Qant la vit l'amustant. SB: 2228a Qant la voit l'amustant. SB: 2342a Qant le voit Malatrie. SB: 2374a Quant ele voit Guillaume. BC: 100a Quant Aymeris le voit. BC: 263a, 431a, 476a Quant le voit Ermengars. BC: 295a Quant le voit Aymeris. BC: 348a Quant Aymeris la voit. BC: 555a Quant les voit l'amustans. BC: 604a Quant Guillaumes le voit. BC: 1726a Quant Bueves ot veus. BC: 1867a Quant Bueves l'a veii. BC: 967a Quant le vit la pucele. BC: 2390a Quant le voit la pucele. BC: 2762a Quant le vit l'amustans. BC: 2940a Quant Malaquins les voit. BC: 3302a Quant ele vit Gerart. BC: 3682a Quant le voit Malatrie. BC: 3702a

Et qant Bueves le vit. SB: 658a Et qant li dus le vit. SB: 66Sa, 2239a Et qant paien les virent. SB: 1201a Et qant Girart le voit. SB: 1299a Et qant Girart le vit. SB: 1901a Et qant Guion les vit. SB: 2591a Et qant je vi tot ce. SB: 2269a

These systems of hemistichs (only one of -which is found in the b position) are, like the respondre system described previously,^- flexible enough to allow any noun or pronoun subject to be included in them by metrical compensation. If the subject and its modifiers come to four syllables in length, as li amiranz or cuens Aymeris in the first ofur examples above, the verb is reduced to two syllables,

^•pp. 108-10 164 oI, and the temporal conjunction qant is omitted* If, at the other extreme, the subject is only of one-syllable length, as in the last example of entendre -where it is the personal pronoun il, the verb and the temporal conjunction together take up the other five syl­ lables of the hemistich. Since there is only one verb of seeing, and since voir is monosyllabic except in one case, thereis a greater frequency of verses opening with et qant to take up the slack in the verse -when seeing is involved.

The verb olr is also found in two other formulas.

Si que tuit l'ont ol. SB: 1329b Si que tout l'ont ay. BC: 1018b, 2150b Si que bien l'ont oy, BC: 2008b

Si com oy aves. BC: 1058b Si cam avez oy. BC: 1471b, 1805b

In verses which begin with hemistichs in which one character perceives the actions of another, the second hemistich is often an expression of anger, fear, pity or joy. These are listedbelow without regard to the content of the first hemistich.

Anger:

Ou n'a que corrocier. SB: 391b N'i ot que corrocier. SB: 397b, 2014b, 2111b, 2136b K'i a que corrocier. SB: 2097b

Le sans cuide chengier. SB: 625b Le sens cuide changier. BC: 348b

Sin a le sans marri. SB: 1299b

A pou que il n'enrage. SB: 1110b A pou n'est enragiez. SB: 2323b A pou n'est enragies. BC: 651b

A pou n'est forsenez. SB: 1206b, 1363b, 1577b, 1932b

Si n'a talent qu'il rie. SB: 1285b 165

Si n*i ot c'airer. SB: 1736b

Color prist a muer. SB: 23^7b

Pres n'a le sens derve. BC: 431b, 1673b, 29*K)b, 3853b

Le sens cuide derver. BC: 3062b

A pou qu*il n*en marvie. BC: 6 M b Tel duel an a k*a pou que il ne s'en marvie. BC: 1506

Fear:

Si ot au cuer paudor. SB: 333b

De la mort ont poor. SB: 872b

Si s*en est effreez. SB: 808b Es les vos effreez. SB: 829b Un pou fu effreez. SB: 1901b

Pity:

Sin ot au cuer tendror. SB: 879b

Joy:

N'i ot qu*eleescier. SB: 392b, 18Mb

Ne puet muer n*en rie. SB: 1690b

Le cuer an ot joiant. SB: 1625b

Eh ait grant joie eue. BC: 91^, 1568b S»en a grant joie eue. BC: 1273b Tel joie eu en a, BC: 3632b

Gerars l*ot, si an a d.e joie un ris gete. BC: 3503 De joie un ris geta. BC: 3652b

Some of these groups share common elements, and they may be grouped as follows: n ’i ot que corrocier. n*i ot quteleescier. in which the verbs are opposite in meaning; a pou n*est anragiez. a pou n*est forsenez. a pou que il n*enrage. a pou qu«il n*en marvie. in which the verbs of anger differ according to the needs of rime; le sans 166 cnd.de chengier and le sens cuide derver. in -which both verb and rime change; si ot au cuer paudor and sin ot au cuer tendror in which the noun changes the meaning of the phrase but the rime remains the same. It is possible (this is pure conjecture) that one expression in each of these groupings became formulaic and later brought about the creation of the others by the poets in order that the task of riming might be facilitated.

Stock expressions which have to do with movement are obvi­ ously very useful to the jongleur who must constantly be telling about the coming and going of heroes. As is the case with battle scenes, the arrivals and departures of armies are not usually handled panoramically in the two poems (exceptions to this will be evident in the examples below), but rather individuals are singled out for description.

Anrnjejlljr la voie:

S»a sa voie acoillie. SB: 1278a Fu lor voie accueillie. BC: 1523b Ont leur voie accueillie. BC: 1831b

jr Craler):

Ore en alez en France. SB: 365a S'en ala droit en France. BC: 1811a 0 vous ira en France. BC: 2087a K*o vous iront en France. BC: 2162a Car se ne voise en France. BC: 2237a Qu'o vous iront en France. BC: 3221a Que s'ele va en France. BC: 3245a

Iront droit a Barbastre. BC: 1647a S ’en vont droit vers Barbastre, BC: 1824a Vont s»en droit vers Barbastre. BC: 2710p

Clarion vait avant. SB: 973b, 1039a Et Aymeris s’en va. SB: 319a Et Hermenjart s*en vait. SB: 194a Vont s'en li Sarrazin. SB: Nostre Frangois s'en vont. SB: 977a Or s'en vont Sarrazin. SB; 1155a Guielin s'en ala. SB: 1293a Malatrie s*en vait. SB: 1655a Vait s'en li dus Girart. SB: 20835a, 2091a A ses neveus ala. BC: 69a Et les dames s'en vont. BC: 405a A la pucele alai. BC: 3045a Vers Malaquin rala. BC: 3409a As dames en irons. BC: 3577a As puceles se vont. BC 3684a aprochier:

De Girart aprochier, SB: 2096b L»amirant aprochier. BC: 174Sb, 2843b Sarrazins aprochier. BC: 2861b chevauchier:

Et Malaquin chevauche. SB: 2421a Nostre Frangois chevauchent. BC: 1701a, 3526a, 3582a Sarrazin chevauchierent. BC: 1828a Nostre baron chevauchent. BC: 3620a issir:

Nostre Frangois s'en isent. SB: 768a Qant cil dedanz s'en isent. SB: 1323a S'aucun Frangois s'en ist. SB: 1777a Qant des portes issirent. SB: 2164a par la porte s'en isent. SB: 2506a Des portes sont issu. BC: 1553b Des portes s'en issirent. BC: 2796a

Puis issi de Barbastre. BC: 1840a K'issus fust de Barbastre. BC: 2473a Ert issue de Barbastre. BC: 2787a Isterons de Barbastre. BC: 3468a Issirent de Barbastre. BC: 3514a passer:

m a passe avant. SB: 375a Cil est passez avant. SB: 725a Et passerons de Sore. BC: 3469a Et passeronmes Sore. BC: 349la Nous passeronmes Sore, BC: 3556a 168 repairer:

Nerbonnois repairlerent. BC: 1805a Sarrazin s'en repairent. BC: 3721a retorner (tomer):

Et Aymeris s'en tome. SB: 3Hp, 341a BC: 521a Les Frangois retomez. SB: 1069b Et Bueves s'en retorae. SB: 1245a Et Corsout se toma. SB: 1699a Gautier s'en est tomez. SB: 2432a Eh est Girart tomez. SB: 2520b Sarrazin retomerent. SB: 2588a Et Libanor s'en tome. SB: 1866a BC: 2381a Quant Frangois retomerent. BC: 484a Et Guis s'en retoma. BC: 2016a Et chascuns s'en retome. BC: 2333P L'dnt paien retomee. BC: 2975b Malaquins s'en retome. BC: 3653a

Sont retome arriere. BC: 1797b S'en retoma airier, BC: 2368b S'en retoma arriere. BC: 3294b Retome mais arriere. BC: 3475b Puis retoment arrier. EC: 396b

Se ne tomes a moi. BC: 2729a, 2839a retraire:

Retraions vers Barbastre. BC: 1865a Se traient vers Barbastre. BC: 1997P venir (revenir):

Puis revenrons arriere. BC: 1500a Puis revenrai arriere. BC: 1768a Tost revenrai arrier. BC: 3284b

Einz que venist en France. SB: 500a Se il vienent en France. BC: 2632a

Quant il vint a Barbastre. BC: 751a Tant qu'il vint a Barbastre. BC: 3265p

Car l'amustans s'en vient. BC: 471a A 1'amirant s'en vient. BC: 2070a Apres Gerart s'en vient. BC: 2727a 169

Devant les autres vint. BC: 2007a, 2725a, 3725a Devant son pere vint. BC: 2939a

Qant l'amirant i vint. SB: 305a Li amiranz vint la. SB: 339a Por l'amustant i vint. SB: 1236a Que Malatrie vient. SB: 1685a Mais 1*amiranz i vint. BC: 1975a Quant l'amirans i vint. BC: 270a

Si vint a la pucele.^ SB: 2051b Si vint as Sarrazins. SB: 2251b Puis vint a la pucele. BC: 2421a

Verbs of motion are sometimes combined with other verbs to form formulas.

Ne vos alez hastant. SB: 150b Qui molt le vait hastant. SB: 140b

S'en vet esbenoier. SB: 1828b M'irai esbenoier. SB: 1834b

S'est alez apoiier. SB: 1873b, 1876b

G'irai a lui parler, un petit m'atendez. SB: 1931 Atandez moi un pou, g'irai a lui parler. SB: 2428 Qu'il viangne a moi parler. SB: 2396b Venez a lui parler, SB: 2463b

Qui i vint acorant. BC: 1138b, 1152b

K'ales vous atendant. BC: 1147b, 3694b Moult les vois desirant. BC: 3394b Moult le vois desirant. BC: 3697b

The verbs monter and descendre. signifying to mount and to get off a horse, also fall into formulaic patterns. monter:

Et Guillaumes i monte. SB: 277a BC: 44la

^Cf. Et vait a la pucele. SB: 2047p« 170

Guillaume ont remonte. SB 285a Et 11 bers i monta. SB 1386a Or fu 11 Turs montez. SB 1611a Et monta la pucele* SB 1637a Et la pucele 1 monte. SB 1825a Et Libanor 1 monte* SB 1851a Et 11 vallez 1 monte. SB 1915a Malaquin 1 monta, SB 2408a Et 11 baron monterent. SB 2608a Corsolt ont remonte* BC 1758a Clarions en monta. BC 1928a Lors monta la pucele* BC 2703a Quant Gerars l ’ot montee* BC 2705a Et Malaquins monta. BC 3254b Bien soit chascuns montes. BC 3392a Lors montent 11 baron* BC: 3768a

Descendre:

Girart est descenduz. SB: 2564a Guillaumes descendl. BC: 391a Limbanors descendl. BC: 2133b, 2419a

For a messenger’s trip a periphrastic expression is used*

Juque au tref Libanor ne fina de coitier. SB: 1829 Juque l ’ost Libanor ne fina de coitier. SB: 1855 Juque au tref l ’amustant ne fina de coitier. SB: 2633

Me fina de coitier bears comparison with other formulas presenting a

Similar idea, na fina de poindre. which is found only in the first hemistich position, undoubtedly because there are so few words which rime with poindre, and Desi que a l ’estour n ’i ot resne tenue.3

Both of these are found in messenger motifs, as is SB: 2633 above.

Other special kinds of movement are expressed forraulaically.

The Saracen turncoats sometimes act as guides for the Christians since they know the lay of the land.

3see above, pp. 157-58. 171

Clarion les conduit. SB: 787a, 132*Ja Mnlaquin les conduit. SB: 2524a, 2609a BC: 3524a When a waman makes her entrance, someone rushes up to greet her.

SB: 375b, 1832bSi la cort enbracier. SB: 375b, 1832bSi

Before leaving camp the Saracens aim themselves.

As armes sont coru. BC: 1513a, 3857a

At the beginning of two successive laisses Malatrie stands up.

La pucele se drece o le viaire cler. SB: 1581 La pucele se drece a la clere fagon. SB: 1597

The latter verse also introduces laisse 152 later in SB (v. 5699)•

Adverbial complements of place and time and manner sometimes fill a hemistich.

Sitreci a Barbatre. SB: 455a Ehtreci c»a Barbatre. SB: 491a, 2422a Enfreci c*a Barbatre. SB: 474a Bitresi c»a Nerbone. SBi 2216a Ehtreci que a Sore. SB: 24lla

Comment l»amustans vint coiement conme lere. SB: 1460 Je vos mende la bele, coiement a larron. SB: 2462 Par la porte s*en isent coiement a larron. SB: 2506

Por coi oisites fors de ceanz sanz congie? SB: 2241 Conme osates issir de ceanz sanz congie? SB: 2246

Par devant la pucele. SB: 2375a Par delez la pucele. SB: 2600a

Ehz el chief par devant. SB: 181b, 1464b

Uh arpant et demi sont arrier recule. SB: 2682 Reculerent paien un arpant et demi. SB: 2169

Et derriere et devant. BC: 1140b Et devant et derriere. BC: 2904b

Trestout premierement. BC: 2203b, 3326a, 3411a

Si bien et si a droit. BC: 3512a, 3608a

Bien et hardiement. BC: 2832a, 3122b 172

Soz Nerbone en la pree. SB: 24b, 26b, 1565b H estoit soz Nerbone qant fui pris en la pree. SB: 1184b Es prez desoz Nerbone, SB: 67a, 105a, 348a Es prez desoz Mijainne. SB: 82b Es prez par desus Sore. SB:. 1761a Es prez desoz Barbatre. SB: 2012p Es prez devant Nerbonne. BC: 134a

Soz Nerbone en la plangne. SB: 97b

Desoz Nerbone es prez. SB: 254b Devant Nerbonne ou pre. BC: 427b Desous Nerbonne es pres. BC: 1874b

Desor l*eve de Sore. SB: 1828a Delez l*eve de Sore. SB: 1854a Deseur l'aigue de Sere. BC: 95b Seur l*yaue de darente. BC: 286b Seur l*aigue de Marsoing. BC: 313b Pres de l'aigue de Sore. BC: 2052a Pres de l’yaue de Sore. BC: 2454a Fors de l»aigue de Sore. . BC: 2622p Dedens l*aigue de Sore. BC: 2623a, 3163a Que en 1 ‘aigue de Sore. BC: 3042a

Et el brueil de pinel. SB: 2481a Ehz el brueil de pinel. SB: 2510a Pres dou brueil de Pinel. BC: 2171a, 3872a Ehs el brueill de Pinel. BC: 3255a, 343^, 3722a, 3805a Les le brueil de Pinel. BC: 3712a, 3927a Fu el brueil de Pinel. BC: 3755a Ehs el bois de Pinel. BC: 3523a C*est li bos de Pinel. 4- BC: 3597a

Au ru de la fontaine. SB: 1767a, 1782a, 2526a Au rui de.la fontaine. BC: 2952a

Desoz 1*ombre d*un pin, SB; 2171b, 2197b Tant qu*es onbres del pin, SB: 2255a

A nul jor de sa vie, SB: 104p A nul jor de raa vie. SB: 650p, 2672p

^The editor of Le Siege de Barbastre does not capitalize pinel. Albert Henry notes that the word, which originally meant simply "pine" has become a toponym in these phrases, and he does capitalize it. See Les OEuvres d*Adenet le Roi. Vol. II: Buevon de Conmarchis (Bruges: "De Tempel,11 1953)» P« 194. 173 Ja jor de vostre vie. SB: 2286a Fait il, jor de ma vie. BC: 184?b

Au jor de Panteeouste. SB: 47a Le jour de Pentecouste. BC: 285a, 522a,* 1881a

Eins le tiers jor pasant. SB: 149b Ainz catre jorz pasant. SB: 1509b Ains quo li tiers jours passe, BC: 1193b Ains le tierch jor venu. BC: 1266b

L'endemein a matin, SB: 481b due l ,andemain matin. BC: 1646a Et 1* endemain matin. BC: 1655a

Par son l*aube esclerie. SB: 1131b, 1143b Au matin par son l ’aube. SB; 1631a Le matin par son l»aube, SB: 1759a EL demein par son l*aube. SB: 1806a Hui matin par son l*aube. SB: 2251a

A l*aube aparisant. SE: 2584a, 2637a Apres l'aube aparant. BC: 1151b Apres l*aube esclairie. BC: 1837b Dusqu*au jour aparant. BC: 3698b

De ci a l*esclerier. SB: 2636b De ci a l*eclerier. SB: 2654b

Apres la mie nuit. BC: 811a, 839a Droit a la mie nuit. BC: 3488a

Plus que por une annee. BC: 880b, 1234b

The sets of expressions ■which include Es prez desoz... plus the name

of a city (and analogous phrases) and L*ajgue de... plus the name of

a river, are convenient ways of expanding the proper name of a place

into ah entire hemistich. Par son l ’aube esclerie. Au matin par son l*aube and Apres l»aube aparant are similar in meaning, but they

appear under different conditions. The first and last of these may

appear at the rime (-ie and -ant being rimes of high frequency) but

Au matin par son l ’aube is found only at the caesura. Host of the 174-

other adverbial expressions in this list occur either in the first or the second hemistich, but not in both.

Several formulas are found in an adverbial position intensi­ fying the negative meaning of a verb. When the matter under dis­

cussion is of some importance, the intensifying expression involves the riches of a city.

Je cult qu*il nel donroit por l*or de deus citez. SB: 1084 Vos n»i vosisiez estre por l*or de Monpellier. SB: 2317b Mes ele n*i do mi s t por l»or d'une cite. sb: 2365 Ne fust mie si lies pour l*or de deus cites. BC: 1083 Ne fust mie si lies pour l*or de Monpellier. BC: 2367

For lesser matters the object cited is of insignificant value.

Que il n'en puet avoir un denier mormee. SB: 585 Or n'en puis pas avoir ton denier monnee. SB: 595 Je ne pris Mahomet deus deniers monneez. SB: 1360 K ’assaus n*i puet forfaire un denier monnae. BC: 2928

La barge au roi Judas n*i vausist un denier. SB: 1648 Cui chaut qant sa desfance n'i vasist un espi? SB: 1321

Que n»i peusiez perdre la monte d*un denier. SB: 2313 Tel le dit ne valut la monte d*un denier. SB: 2322 Vous n*i peussiez perdre la monte d*un soller. BC: 3077 Ne l'aubers de son dos la monte d*un fuisel. BC: 3729

Les dieus as Sarrazins que ge ne pris un gant. SB: 982 Ne cestui ne les autres ne pris pas un boton. SB: 1026 Por les dieus ou tu croiz ne donroie un boton. SB: 2002 Sire, dist Navaris, ne vous pris une pronne. BC: 3369 Je ne pris pas nos vies un espi de fourment. BC: 3124

Qu'il ne redoutent mort une pome paree, SB: 1568 N*i forferoient il une ponme paree. BC: 875 Ne pris fol hardement une ponme paree. BC: 2999

Qu*il ne li grieve pas une fuel lie de mente. BC: 793 K*assaus n*i puet grever une fuel.He de mente. BC: 1349 Ne pris tout lor povoir une fueille de mente. BC: 2246 Car je ne me pris pas une fueille de mente. BC: 3133 175 Ne mais n ’arons sans vous qui vaille un seul festu. BC: 9P7 Celui jour n*i perdirent vaillant une estriviere. BC: 2913 Ne li valu la targe la penne d'un mantel, BC: 3728

The adverbial phrase may also evaluate distance:

Einjois peusiez est re une huchiee alez. BC: 1*1-39 Que ele avoit de lone le tret a un archier. SB: 16*19 A tant pres de ma tente conme un arc puist giter, SB: 1776

All of these phrases are implied comparisons between the object or action under discussion and the value (or distance) of the adverbial element.

Many formulas do not fall into the groupings set forth in this study. These miscellaneous phrases are listed below. Those •which have a noun as their nucleus:

Si ot en sa conpangne. SB: 2639a Et ot en sa conpangne. SB: 2656a Et sont en sa conpaigne. BC: 180a, 2152a Ot en sa conpaignie. BC: 2187b

Tote est de la lingniee. SB: J+a Tu ies del fier parage. SB: Vous estes dou lignage. BC: 79a Bien i fu li lignages. BC: 3199a A estre dou lignage. BC: 3^22p

De ce que ^amirans. BC: 826a De ce que l'amustans. BC: 2088a

Comment li amirans, BC: 859a Conment li amustans, BC: 755a

Contre les Arrabis. BC: **8*lb, 3760b

Devers lor anerais, BC: 1597b Contre lor anemis. BC: 1601b Contre ses anemis, BC: 2780b

Etami les Sarrazins. BC: 1912a, 2870a

0 lui maint Sarrazin, BC: 271a, 299a 176

H n*i ot cresti'en. BC: 402a H n*i a crestlen. BC: 2 8 9 >

Que n'aiez la pucele, SB: 24?0p J1averai la pucele. BC: 2576a Nous arons la pucele. BC:.3127a Que n°aions la pucele. BC: 3458p Se je n*ai la pucele. BC: 3 4 7 ^

Et maint autre grant prince. BC: 67a, 504a

Cil trova les Francois. SB: 97a Si trova les Francois. SB: 704a

Ne trouverons paien. BC: 871a 1-loult trouverent paien. BC: 2190a Pou trouvast on paien. BC: 2412a

Quel gent as tu trovee. SB: 131a Quel gent as tu trove. SB: 141a

Sa main a son menton. BC: 168b, 804b

De saphirs, d*esmeraudes, d'agathes, de rubis. BC: 2285a De saphirs, d* esmeraudes et de rubis prisies. BC: 3667a

L'arretueil de la lance vait en terre fichier. SBi 1870 L ’aretueil de sa lance vi devant lui fichier. SB: 2266

Par iceste raison. BC: 3040a Et par ceste raison. BC: 3177a

Tote se volente. SB: 2351b Tote lor volente. SB: 2404b Toutes ses volentes. BC: 693b, 2503b

Que ce soit verite. BC: 2079b Que ce soit verites. BC: 1065b

Those ■which have a verb as their nucleus:

La teste avra copee. SB: 36b De la teste coper. SB: 1754b Sor la teste a coper. SB: 1930a N*ait la teste coupee. BC: 3004b

Toz les menbres trencihier. SB: 912b, 2142b Por les menbres trenchier. SB: 1888b 177 De la teste tranchier. SB: 2106b Et ces testes trenchier. SB: 2158b H'ait la teste trenchie. BC: 632b

Qant de fonz fui levez. SB: 1402b, 1976b

Por son cors deporter. SB: 1762b Por lor cors deporter. SB: 1768b

Por pucele besier. SB: 2129b, 2483b

S'or estiez la sus. SB: 630a Que ge fusse la Jus. SB: 718a S'estiems or la sus. SB: 2072a S'or est£ens la sus. SB: 2548a

T'a tolue Barbatre. SB: 1108a, 1736a Tolue ra'ont Barbatre. SB: 1114a

Se les poons conquerre. BC: 723a Que je puisse conquerre. BC: 2447a Ke ne porra conquerre. BC: 3401a

Guielin encontrai. SB: 2638a Guielin encontrames. SB: 2655a

Li amiranz s'eveille. SB: 699a L'amoraive s'eveille. SB: 760a Li charteriers s'eveille. BC: 962a

De son lit ert levez. BC: 2430a De son lit s'est levee, BC: 3195a

Souz les arbres s'assient. BC: 3202a Souz un arbre s'assieent. BC: 3316a

Montez desor ce mur. SB: 2081a Montez desus ce mur. SB: 2547a

Si sailli el baucent. SB: 172b H sailli sor Ferrant, SB: 1346a Si sai.lli en Morel. SB: 2591b

Or me randront paien. SB: 201a Or me randez les contes. SB: 688a Or me rendez mon fil. SB: 2118a

A l'amirant randuz. SB: 310b A l'amirant randites. SB: 84lb 178

As paiens delivrez. SB: 282b As paiens delivra. SB: 1700b

Car Dieus nous aidera. BC: 938a, 3765a

Barons, secorons le, SB: 783a, 2'604a

Conment chascuns s'ayue, BC: 1580b Conment chascuns s'ayde. BC: 1958a

Que n*i avra mestier. SB: 153a Ne li avra mestier, SB: 624b Qui nos avra mestier. SB: 909b Tost nos avroit mestier, SB: 934a Se il en ont mestier. SB: 1662b Ore ai de vos mestier, SB: 2161b Gil ne li a mestier, BC: 4l6a

Et il si font tantost. 475a, 915a, 1035a

Faites le l£ement. BC: 2204b, 3127b

Nous le ferons ainsi. BC: 1003b Je le ferai ainsi. BC: 2130b

Par force les ont fais. BC: 404a Par force les ont fait, BC: 2894a

Conme font li Franjois. SB: 222a Bien m*avroit fet Francois, SB: 1895a Bien le font no Francois. SB: 3944a

Je vos ferai ardoir. SB: 2232a,, 2346a

Se ge m*amie pert, SB: 207a Se ge mes fiz i pert. SB: 328a, 359a

Con m*avez vergonde, SB: 2229b, 2343b

Ne fust mie si liez, SB: 1351a 083a, 1967a, 2367a Ne s'est aseurez. SB: 266b, 1386b

Toz sui desheritez. SB: 263b L*en ot desherite. SB: 584b K*en a desherite. SB: 594b Sera desheritez. SB: 767b, 1205b 179

There is one group of phrases which, although they seldom occur a second time in exactly the same form, undoubtedly consti­ tute a formulaic system. The idea expressed in all of them is the same: ••there was no hesitation'1 in the performance of a cer­ tain action. Such a phrase may be inserted after any action without affecting its meaning. Seven verbs are involved in this system: atargier (tarsier). arester. atendre, demorer. detriier and se.jomer. Three of these include substantive forms: arres- toison and arestee, demoree, atante. arester:

Ne s»i sont arestez. SB: 726b N»i ont fet arestee. SB: 1155b Et sanz arestoison. SB: 1346b Que plus ne s'i areste. SB: 2082b Ne s*i vost arester. SB: 2422b Ne s*i volt arrester. SB; 68b Sans point d’arrestison, BC: 172b Que plus n*i arresta. BC: 747b, 363^ Ne s'est pas arrestes. BC: 1088b Plus n*i sont arreste, BC: 1215b N*i fist arrestisons. BC: 1382b N*i sont plus arrestu. BC: 1550b Sans nul point d,arrestee. BC: 1740b Plus n*i est arrestes. BC: 1872b Plus n»i est arrestus. BC: 1927b Iluec plus n*arresta, BC: 2327b N*i vont plus arrestant. BC: 2416b N*i est plus arrestes. BC: 2529b N*i font arrestisons. BC: 2587b N*i sont plus arreste. BC: 2926b, 3479b, 3818b N*oi pas tans d'arrester. BC: 3031b Que plus n*i arresta. BC: 3251b Ne plus n*i arrestons. BC: 3607b atargier:

Ne li firent targier. SB: 383b Ne vos atargiez mie. SB: 695b Qui ne se tarja mie. SBs 690b, 1699b N'orent song de tardier. SB: 915b 180

Qui n*a song do targier. SB: 1825b No vos puls atargier. SB: 2269b Plus n*iest atargies, BC: 1981b Et Guis sans point targier. BC: 2877b N*i volt point atargier. BC: 3286b Plus n*i volt atargier. BC: 3423b

atendret

N*i fait pas longue atente. BC: 1353b Plus n*i a atendu. BC: 2011b Qu*ele plus n*atencli, BC: 2703b

demorer:

Sanz plus de demorer. SB: 1586b Ce ert sanz demorer. SB: 1758b Sanz point de demorer, SB: 1775b N»i veil plus demorer. SB: 1779b Et si ne demorer. SB: 2430b Gete sans demoree. BC: 863b Errant sans demoree. BC: 865b N*i a plus demore. BC: 1207b

detriier:

Plus n*i volt detriier. BC: 393b

se.iomer;

N*i a plus sejome. BC: 840b

All occurrences of this system are in the second hemistich. The

phrases add nothing to the narrative, and their sole function seems

to be to fill up verses. They provide rimes in -er (infinitive

forms of arester and demorer). -ez or -es or -e (past participle

forms of arester. demorer and se.iomer). -ier (infinitives of

detriier. atargier and targier). -ies (past participle of atargier).

-a (passe simple of arester), -u and -us (past participle of atendre.

alternate past participle of arester^). -on (arestoison). -ente

^Arestu. the variant participle, developed by analogy ■with a perfect form arestut. ■which was generated by analogy with estut from

i 181

(atente). -ons (first person plural, present of arester), -ee

(demoree) and -ie (the adverb mie). This is a structure which can be put to great advantage by the jongleur because it is made to fit so many different circumstances. Of the fifty hemistichs of this system treated here, 33 a**e from Buevon de Conmarchis.

Many of the fixed phrases of Le Siege de Baibastre and

Buevon de Conmarchis consist of sets of synonyms or nearly syno­ nymous words 'which seem, at first glance, to be linked together for no particularly good reason. Although the actions, states or objects contained in' these formulas vary greatly, I have grouped them all together because of their peculiar nature.

Some are combinations of two or three nouns.

A force et a baudor. SB: 870b A force et a vertu* BC: 905b

Chanbres i a et votes et croutes et celier. S3: 903 Qu*il n*i ramengne croute, ne vote, ne celier. SB: 919 Cerchent croutes et votes, n'i a celier remes. SB: 940

As loges et au tres. SB: 1968b De loges et des tres. SB: 2515b

Li demoine et li per. SB: 2209b Li demaine et li per. BC: 88b

De maltalent et d*ire. BC: 189a D*ire et de maltalent. BC: 645a, 1356a, 15 , _ , 2736a, 2988a, 3063a, 3780a

Font grant noise et grant hu. BC: 1559b Grant noise et grant huee. BC: 1710b, 3746a. Ot grant noise et grans hus. BC: 1915b

*stetuit. See Pierre Fouche, Le Verbe francais (Paris: Societe d'Edition ”Les Belles Lett res,” 1931), p. 361. 182

Ot grant noise et grans cris, BC: 1617b Et grant noise et grans cris. BC: 2802b

Les cris et la huee. BC: 3874b Et grant cri et grant hu. BC: 3923b

The following combine two verbs, usually past participles, or two

adjectives.

Dolant et corroceus. SB: 546b Corrocie et dolant. SB: 554b Dolente et corroucie. BC: 2972p

Departi et donez. SB: 700b Departi et donne. BC: 451b, 3806b

Bautisie et leve. SB: 850b Baptisier et lever. BC: 1427b Levee et baptisie. BC: 1855b Baptisie et levee. BC: 2979b

Ocis et afolez. SB: 928b, 1399b Ocis et afole. SB: 2344b Ocirre n*afoler. BC: 3079b

Armez et fervestiz. SB: 1251b Arme et fervesti. SB: 1323b Arroes et fervestis. BC: 940b, 2787b Anne et fervesti. BC: 1000b, 2160b Armee et fervestue. BC: 1278b

Ocis et detrancihiez. SB: 2640b, 2657b Ocis et detranchie. SB: 2644b, 2663b Ocis et detrenchies. BC: 667b, 1171b Detrenchiez et ocis. BC: 2262b

Corrociez et irez. SB: 2224b, 2235b, 2338b Corroceus et iriez. SB: 2649b Corroucies et iries. BC: 2589b

Hetenu sont et pris. BC: 506b Retenue ne prise. BC: 525b

Et dite et devisee. BC: 86lb Lors dist et devisa. BC: 3643b Bien dit et devise. BC: 3825b 183 The rest combine two adverbial elements.

Volentiers et de grez. SB: 277b, 1073b, 1097b Volontier et de gre. SB: 2437b De gre et volentiers. SB: 2250b

Isnelement et tost. SB: 1387b BC: 2706a Tost et isnelement. BC: 1318b Moult tost et moult isnel. BC: 2634b, 3723b

Tost et apertement. BC: 270b, 10?4a, 1328b, 2361a, 3294a Apertement et tost. BC: 2324a

Of these synonymous expressions, 30 are from Le Siege de Barbastre and 49 from Beuvon de Conmarchis. The phenomenon of metrical com­ pensation, by which the words of a hemistich are moved about to make room for a new entry is a mechanism for keeping certain expres­ sions down to one series of six syllables (a hemistich). When the hemistich "overflows'1 it becomes more difficult for the .jongleur to fit his poem together, for >41116 most of his ideas are already organized into blocks of one or two hemistichs, he then has to construct a hemistich with fewer than the ordinary six syllables plus what is overflowing from the first half-verse. In the synony­ mous expressions listed above, the problem is just the opposite of that which necessitates metrical compensation. Each element of a pair of synonyms, taken by itself, fits so well within the physical limit of the h emistich that there are some syllables left over — a situation which is just as intolerable, in poetic technique, as an over-stuffed hemistich. In order to take up the inadmissible gap, words undoubtedly were combined with their synonyms or with 184 other words close to them in meaning so that six-syllable blocks were foamed which had the same meaning as the original simple phrase or word. Dolant and corroceus. for example, may not have exactly the same connotations (the first inclining more toward sadness and the second toward outright hostility), but in the situations where they occur the person i-ho is described is saddened, which also makes him angry, or else he is angered, which saddens him. One element of each pair entails the other automatically, so that the poet can use the blocks of wooxls thus constituted as if they were but one word. At most they undergo inversion (de mal­ talent et d*ire. d»ire et de maltalent). and this only comes about when the rime demands it or when the hemistich occurs in the a position where rime is a consideration.

Of the 714 expressions found in this chapter, 379 are from

Le Siege de Barbastre and 329 from Buevon de Conmaarehis. Of this number 352 (approximately 130 per 1000 verses) from Le Siege de

Barbastre and 299 (approximately 75 por 1000 verses) from Buevon de Conmaarchis are formulaic. CHAPTER V H I

CONCLUSION

It should be evident from the thousands of expressions quoted in this dissertation that a formula is not a rigidly fixed phrase, but a pattern of words which can be altered by the poet to fit the immediate needs of the particular verse.

In the modification which I have called "metrical compensation" a new word which is necessitated by the larger context is ad­ mitted into the formula, to adjust either its meaning content (as in the case of the substitution of proper names) or its syntactic pattern (as when a vert tense is changed). This "substitution word" may contain more or fewer syllables than the item it replaces, and in such a case there must be compensating changes made somewhere else in the formula so that it will not exceed the limits of the hemistich. Another alteration frequently found in b hemistichs is a change in the last word of the formula so that it may fit into the rime of a particular laisse. Meaning is the essential component in the formula, but it does happen that sev­

eral phrases with essentially the same meaning are formulaic, so that at times the poet must choose between two formulas. In these

cases factors of rime and syntax are important in the poet*s

185 186 decision of -which of the alternate formulas he should use.

Some of the formulas listed in this study seem to be nothing more than the most common cliches of the Old French language.

How, it might be asked, could any author compose a poem without using such phrases as Ore ai de vos mestier^ or Apres la mie p nuit? Of course the cliches of daily speech are to be found in every poem, but whereas the traditional author will cherish these expressions, fit them into the mold of the verse and use them again and again, the literary author will only employ them with reluctance. The cliche is, after all, the nemesis of men of letters. The oral poet, on the other hand, finds it useful because it enables him to compose rapidly. Whether or not a given phrase is part of the common fund of Old French cliches is of minor importance. A useful expression will be added to the poet,s stock of formulas whether he invented it himself or simply garnered it from the daily speech of his contemporaries.

The three types of formulaic expressions, those representing actions, those which are mainly for ornamentation and tone, and those which serve to organize the structure of the work are not, however, all subject to the same kind of manipulation within the poem. The action formulas and the proper names are found whenever the poet has occasion to sing of a certain event or of a certain hero. Thus one can say that their use is determined by the progress

^Cf. above, p. 1 7 8 .

^Cf. above, p. 1 7 3 . 18? of the narrative. The organizational expressions are relatively few and their use is obvious and limited. But the epithets per­ form an important function in that they make the task of the poet much easier, much more leisurely, than it would be without them.

As an examination of the variants of the C manuscript of Le Siege 3 de Barbastre shows, the places occupied by epithets are, as it were, lacunae in the texture of the poem which can be filled in with whatever is suitable and handy* The poet can relax for a moment from the pressure of continuous composition, or use the time gained to think ahead to the next few verses.

A reading of Gormont et Isembart. r>u-mamne and Roland suggests a feature which, as far as I know, has not been noticed by previous critics, namely that one of the chief stylistic dif­ ferences between these works on the one hand and the rest of the genre on the other is that the early poems are free of epithets which take up an entire hemistich. The boring repetition of such phrases as a la chiere hardie and a la chiere manbree in the second hemistich position as many as three and four times in the course of a single laisse detracts greatly from the force of the narrative, and yet this is precisely what one finds in work after work from the twelfth century all the way down to Dieudonne

3 See above, pp. 87-90* 188 de Hongrie. the last of the fourteenth-century chansons da geste.**

In a paper read at the Colloque de Liege, Martin de Riquer speaks of the different stages in the development of the Old

French epic:

Eh lisant la Berte aus grans pies d'Adenet 11 Rois, nous avons l*impression d*etre en presence d*une con­ ception de l'art de conter tres dlfferente de celle que nous pouvons tirer de la lecture de la Chan pun de Guil- lelme. par exeraple, bien que cette chanson elle-meme soit de ja pour sa part, la refonto d»une autre plus ancienne.-’

Likewise the narrative art of Buevon de Conmarchis is of a dif­ ferent nature than that of Le Siege de Barbastre even though the latter work is approximately one hundred years younger than the

Guillaume.

Two thousand four hundred and eight of the expressions listed in the body of this study were found to be formulaic, of which

1221 were from Le Siege de Barbastre and 1187 from Buevon de

Conmarchis. Thus it can be said with certainty that 30$ of the verses of Adenet's poem contain formulas, compared to 4-5$ of the verses in the anonymous work upon which it is modeled.^ How is

^L. F. Flutre, in "Dleudonne de Hongrie. Chanson de Geste du XIIe Siecle (alias Roman de Charles le Chauve)." Zeitschrift fur romanl3che Philologie. LXVIH (1952V. 321-400. lists some commonly found "cliches" from this poem, among them many of the epithets taking up an entire hemistich which were discussed in Chapter H I of this dissertation.

^Martin de Riquer, "Chanson Jongleresque a ecouter et Chanson romanesque a lire," La Technique litteraire des Chansons de Geste: Actes chi Coll ague de~Liege (Paris: Societe d*Edition "Les Belles Lettres," 1959), pp. 81-82.

^These figures do not include what Lord calls "formulaic systems." See above, pp. 21-22. 189 this difference to be explained?

There can be little doubt that Le Siege de Barbastre is a pro­ duct of the oral tradition and that it was composed by means of formulas. For Buevon de Conmarchis the case is not so simple.

Allowance must be made for the fact that only 2696 verses of the older poem were matched with each other to determine which phrases were formulaic, as compared to 3947 of Buevon de Conmarchis.

Naturally the greater the quantity compared, the more likelihood there is of finding matching verses. When this fact is taken into consideration Adenet's work probably contains about half the pro­ portion of formulas of Le Siege de Barbastre.

And yet this amount seems to be too great for an ordinary literary work. A poem which is purely the product of a literary milieu would certainly not contain even half as many formulas as an oral work. According to Parry, Virgil's Aeneid only holds

92 repeated whole verses out of a total number of 9896. This is less than 1$, but the 1804 repeated verses of the Homeric poems comprise more than 6$ of these works.^ This is, admittedly, a difficult matter to speak about with precision since, aside from

Parry, no one has ever studied the repetitions in a purely lit­ erary work from the same point of view.

Among the formulas which occur frequently in Le Siege de

?Milman Parry, ‘'Studies in the Sbic Technique of Oral Verse- Making. I. Homer and Homeric Style, “ Harvard Studies in Classical Philology. XLI (1930), 134. The percentage figures are mine. 190

Barbastre but are not found in Buevon de Conmarchis are the fol­ lowing:

Formula Occurrences in SB

Et 113, ont respondu. 7

A sa voiz qu*il ot dere. 7

Si con vos conmendez. 5

Biau sire, volentiers. 4

Puis escr£e: "Mbnjoie." 4

Et broche le destrier. 8

A pou n'est forsenez. ^

N*i ot que corrocier. 5

Now it is not inconceivable that an oral poet could get along without these phrases, but it is difficult to see why he would want to omit them since they are all extremely useful. "Et ill ont respondu,M "A sa voiz qu*il ot clere,•• "Si con vos conmendez," and "Biau sire, volentiers" are formulas of discourse which free the oral poet from the task of finding new phrases for the manip­ ulation of dialogue. »*Puis escrSe: •Monjoie*" and Et broche le destrier" represent actions which come up time and time again in the course of battle scenes. "A pou n*est forsenez" and "N*i ot que corrocier" express the anger of a character over the actions

of another. Adenet is faced with the task of describing all these

situations frequently and there are no places where he revises the

stoiy-line of his model so as to leave out a dialogue or a battle.

And yet he spurns the use of much of the fonnulaic material that

the original poem provided him. 191 In the single combat scenes from the two poems in "which every verse was scrutinized for formulaic elements, it was found that

55# of the hemistichs of the passage from Le Siege de Barbastre but only 18# of the passage from Buevon de Conmarchis were formu­ laic* And yet one would expect the battle scenes of a purely tra­ ditional epic to be highly stylized since they form the nucleus and perhaps the oldest part of the work,8

To these indications of the literary character of Adenet* s work should be added his use of indirect discourse in the manner of non-oral narratives,9 and also the presence of the nature topos in the introduction to Buevon de Conmarchis which shows an acquaint­ ance with the courtly literary tradition.

Buevon de Conmarchis. then, seems to stand halfway between the oral tradition and the written literature of the courtly cul­ ture. We know that Adenet le Hoi was a man of letters since he is the author of a courtly romance Cleomades. But a cursory reading of his two other chansons de geste. Les Enfances Oeier and Berte aus grans pies, indicates that they share the general stylistic features of Buevon de Conmarchis.

Two possible conclusions can be drawn from the evidence presented here.

Q See above, pp. 15&-56.

9See above, pp. 109-1 0 .

■^See above, pp. 28-29. ■ 192

The first "would be that Adenet was trained to compose in the oral manner, and was therefore well acquainted with the various formulas common to the tradition, but that as a court poet he decided to write his poem rather than sing it without benefit of book. Although not an absolute necessity, the formulas would be useful to the literate poet since he still has the same verse form to fill and the same need for certain rimes. But he can go back over his work, correct it if necessary, substitute words and phrases which are more apt or more colorful than those he had originally set down. Still he does not have the constant pressure of producing a verse at regular intervals for hours at a time. As his need for the formulas diminishes, so does the proportion of his work which is formulaic. But it does seem strange that a person who began as an oral poet should go on to write a courtly romance in octosyllabic verse, which is not fitted to the use of formulas,H

The more probable conclusion is that Adenet, wishing to compose a chanson de geste. altered his written style to conform to what he knew to be the norms of the genre. Some of his for­ mulas would have come from the poems which he heard sung by

.jongleurs, others he would have-made up himself or borrowed from

^^The only Old French epic in octosyllabic verse is Gormont et Isembart. which is perhaps the oldest poem in the genre. As far as is known, there were no octosyllabic epics in the time of Adenet, 193 Le Siege de Barbastre. In this case Adenet would be writing a poem in the epic genre, with all its conventions, in much the same way as Voltaire wrote the Henriade or Ronsard the Franciade with Homeric conventions. This hypothesis would account for Ade­ net ' s liberal use of an easily imitated formulaic device like the 12 enumerative lists of objects and actions preceded by maint, and also the relative infrequency of formulas of discourse and battle, which are conveyers of the story-line and thus indispensable to the true oral poet.^-3 of course Adenet would not have completely assimilated the oral style, for this takes years of apprenticeship and once achieved would, in fact, have made him an oral poet.

The case of a poet who employs a transitional style between traditional and written poetry is not without parallel in the history of literature. Several scholars in the field of Old

English studies believe that some Anglo-Saxon poems with all the characteristics of the oral method of composition must, in fact, have been written by monks, learned men who incorporated into them Ih the sources of Christian civilization. While the poetry of

-^•^See above, pp. 44-48.

•^See above, pp. 138 and l6 0 .

*^See Ronald A. Waldron, "Oral-Formulaic Technique in Middle English Alliterative Poetry," Speculum. X X X H (1957), 792-804; Arthur G, Brodeur, The Art of Beowulf (Berkeley and Los Angeles: University of California Press, 19597, pp. 3-6; Jackson J. Camp­ bell, "Oral Poetry in the Seafarer," Speculum. XXXV (i9 6 0 ), 87- 9 6 ; Wayne 0*Neil, "Another Look at Oral Poetry in The Seafarer." Ibid.. 596-600; Robert D. Stevick, "The Oral-Formulaic Analyses of Old Ehglish Verse," Ibid.. XXXVH (1962), 382-89. 194

Cynewulf has been demonstrated to be formulaic,^ it has been pointed out that there are literary reminiscences of the Fathers of the Church in his work.^

C, M. Bowra tells of an Albanian author, Gjerj Fishta, a political figure and a highly educated man, who wrote an epic,

Lahuta e Malcis, in the language of traditional formulaic poetry.

This is a case of a man using the oral style consciously to endow his poetry with a quaint and archaic air. Toktogul Satylganov, who composed the Kara-Kirghiz epic Kedei-Khan. employs the tra­ ditional language which he learned to compose in during his youth, but transforms it into a much more subtle medium because of the very fact that he is writing. His poetry too represents a tran­ sition between oral and written style.To these poems may some day be added the Old French Chanson de Saint Alexis and other 1 O non-epic works of early Old French literature.

Albert B. Lord who has certainly had the closest and richest of contacts with the living oral tradition, denies that there can be any such thing as a transitional text on the grounds that a

•^Robert E. Diamond, •'The Diction of the Signed Poems of Cynewulf," Philological Quarterly. XXXVIII (1959), 228-41.

■^Claes Schaar, "On a New Theory of Old English Poetic Dic­ tion," Neophilolotrus. XL (1956), 301-05.

^ C . M. Bowra, Heroic Poetry (London: Macmillan, 1952), pt>. 538-41. T Q See Maurice Delbouille, Sur la Genese de la Chanson de Roland (Brussels: Palais des Academies, 1954), p. 149. 195 poet does not compose any given work by both singing and writing, and thus the oral method is incompatible with written composition.^

This is obvious, for if the poet wished to write his poem he would merely set it down on paper without the intermediaxy of performance.

I do not mean to say that Adenet le Roi was exercizing the art of oral poetxy in composing Buevon de Conmarchis. This work is a written poem. But at the same time it is evident that he was employ­ ing some of the verbal phrasings of the oral epic. Buevon de Con­ marchis is not a transitional text, but Adenet is a transitional poet.

M, Albert Henry, the editor of Adenet*s works, sees vexywell the difference between Buevon de Conmarchis and its model.

Pour le fond, nous avons vu qu'elle (Adenet*s work) oscillait entre la chanson de geste et le roman, sans se detacher completement de l'une, sans atteindre deli- berement 1*autre. Pour le style aussi, on peut parler d'une souplesse emasculee.^O

Without discussing the formulaic nature of Le Siege de Barbastre. he sees that the.later work is a hybrid creation.

The research presented in this dissertation is detailed evidence that Buevon de Conmarchis is as close to written style, as found, for example, in the courtly romance, as it is to the oral style of the epic.

~^The Singer of Tales (Cambridge: Harvard University Press, 1 9 6 3 ), p. 129.

2QLes QEuvres d TAdenet le Roi: Vol. II, Buevon de Conmarchis (Bruges: "De Tsmpel,»• 1953), P» 28. F ERRMSG F hrj *xj bcj *x) htj »xj hrj F ERMSG1 HEEL HED2 FMT2 FIRST FMT3 EMT1 FLLST ERMSG2 EHMSG3 HETEXT ENIMSG ENEMG1 SKIP START

MNGCDS(14),HEMI(13;)- AND $BUEVON AND DE CONMABCHIS$ IN RUN PREPARATION FOR TO FIND FIND FOIMULAS IN TWO THESE TEXTS - SGATRAN STATEMENT LISTING OF COMPUTER PROGRAM CODES6L7) - 0T $,$SGLTNI$,$SGLTNO$)- $0UTB SWITCH. TEXT.HEMI.MNGCDS)- (22H1 ERROR IN READING CARD)- (23H1 ERROR IN READING INTAP)- (43H1THE FOLLOWING HEMISTICHS NON-FORMULAIC)- ARE (5^H0BEL0W FOLLOW FORMULAIC HEMISTICHS MEANING WITH (6X6L6,2X,L6)- (6X 6L6, 7(1X L6))- (1H06L6,7(1X,L6))- (15H1TERMINAL ERROR)- (18H1CARD OUT OF ORDER/ (6XL4L6))- ( i f f l l F I N( )- - (ML) (ERR0R2)- PROGRAM TO PROGRAM TO FIND FORMULAS EPICIN POEMS OF TWELVE PROGRAM FOR ENCODINGPROGRAM FOR $LE TEXTSOF SIEGE DE BARBASTRE$ BL0CKV=7 BL0CKV=7 - SYLLABLE SYLLABLE VERSE - 23 SWITCH=0 - ATTACH FILES ,POOL,NMSTP,6 ATTACH FILES - BOOLEAN (TRUE, BOOLEAN )FALSE, N - O D N U IELS NSP763$NA ,OT $$NB $, $,$INPB $,$OUTA (NMSTP,7,6,3,$INPA FILE LIST (TRUE,.0.777777777777,LITERALS FALSE,0)- DEFINE ,POOL,7,650 DEFINE POOL - (A,B,INTAP,OUTAP,FOBMNO,BLOCK,TEMP,RRLC.,INTEGERS PNN EID,IP $ - OPEN REMIND ,$INPBNO (Adi),B(14),BL0CK(4200,BL0CKV),TEXT(l4), DIMENSION PROVIDED (TEXT(O).E.$ $), TRANSFER TO (ENDPM) $), TRANSFER TO - (TEXT(12)(*).0.77.NE.$OOOOOT$), PROVIDED TRANSFER TO (TEXT(O).E.$ PROVIDED $ - ,$INPBCLOSE REWIND PNRMN $UA $ - OPEN,$OUTAREMIND CALL SUBROUTINE ()«RDSID.SUBROUTINECALL (ERROKL,TEXT)- H 1 ETWITNO AE )- TEXT ON WRITTENTAPE.

APPENDIX

196

C c o a ' PETIT MOVEEM MOVEB KOVEAM MOVEA MOVEPM MOVEP vfflwni NREND ENDPM ERROR1 END ERROR2 SCAM

•TO (ERR0R2) - NUMBERS - NUMBERS (*).0.777777777700),(ERR0R2) TRANSFER TO - (I),I-0,1,I.LE.13))- HPMI(l/6)=inngcds(l)- TEXT(12)=TEXT(12)(*).0.777777777700(7-)$OOOOOA$ - HIMI(12)=TEXT(12)- - HEMl(l)=TEXT(l) HEMI(l/6)=MNGCDS(l/6)- HEMI(l)=TEXT(l/6)- H3'H(l)=TEXT(l) - H3'H(l)=TEXT(l) HE2'H(12)=TEXT(12)- TEXT(12)»TEXT(12) (*). 0.777777777700(f) $OOOOOB$ - H£MI(12)=TEXT(12) - H£MI(12)=TEXT(12) HEMI(l/6)=HNGCDS(l)- ALL CODEDALL HEMISTICHS ON ARE TAPE PROPER TOTH LINE TEXT (12 )=TEXT (12) ( * ). 0,777777777700 (-/■) $OOOOOP$ - THERE IS THERE CARD ERROR,CANNOT READ A TEXT FROM CARD BE - SWITCH=1 - POEMS ARE NOW POEMS COMPLETELYNOW ARE ON - WRITTEN TAPE. TAPE IS TAPE SCANNED FOR FORMULAIC - HEMISTICHS. NA=IP $- INTAP=$INPA PROVIDED (TEXT(6).E.$ $), TRANSFER TO (PETIT) $), TO TRANSFER - (TEXT(6).E.$ PROVIDED PROVIDED (MNGCDS(12) PROVIDED (*) .0.777777777700 .NE.TEXT(12 (MNGCDS(12)(*).0.77.NE.$000O0M$), PROVIDED ) TRANSFER CALL SUBROUTINE ()=RDSID. CALL SUBROUTINE (ERRQR1, - MNGCDS) OTRUH (MOVEA), DO THROUGH 1=0,1,I.LE.5 - RNFRT (RDTEXT) TRANSFER TO - OTRUH (MOVEAM), DO THROUGH 1=0,1,1.LE.5 - TRANSFER TO (RDTEXT) TO TRANSFER - (MOVEBM).1=0,1,1.LE.5 DO THROUGH - (MOVIE),1=0,1,I.LE.5 DO THROUGH - ( CALL SUBROUTINE - )*WRNSFC. $,H3H,13) ($OUTA WRITE ,ENDMG1OUTPUTWRITE - (MOVEPM), THROUGHDO 1=0,1,1.LE.5- ()=ENCODE. SUBROUTINE CALL (HUH, ()=ENCODE. CALL SUBROUTINE (HEMI,36)- WRITE ,OUTPUTWRITE ERMSGl - (RDTEXT) TRANSFER TO - (MOVEP DO THROUGH ), 1=0,1,1. LE. 5- ()=WHNSFC. CALL SUBROUTINE $,HHCT,13)- ($OUTA RT UPT ,EHMSG2, OUTPUTWRITE ((TEXT(I), 1=0,1,1.LE.13), (MNGCDS TRANSFER TO (SCAN) TRANSFER TO - (WEWIN) (SWITCH.E.l)TRANSFER TO PROVIDED - ()=WRNSFC.CALL SUBROUTINE $,HE*H,13)- ($OUTA TRANSFER TO (NRMD) - (NRMD) TO TRANSFER CALL SUBROUTINE ()=EUCODE.CALL SUBROUTINE (HEMI,36) - LS EID,OT $- ,$OUTACLOSE REWIND CALL SUBROUTINE ()=ENDJOB.() SUBROUTINE CALL - LS EID,OT $- ,$OUTACLOSE REWIND

6 3

) -

197

198

OUTAP»$OUTB $ - OPEN REWIND ,$SGLTNO$ - OPTAP OPEN NO REWIND , INTAP - OPQI NO REWIND ,OUTAP - C START OF COMPARISON OF HEMISTICHS. - FOEKNO-O - LINDON-FALSE - CALL SUBROUTINE ()=EINSFC. (INTAP,A, 13, EOFIN, ERROR) - READB CALL SUBROUTINE ()=RENSFC.(INTAP,B, 13,EOFIN,ERROR) - DO THROUGH (CKCDS),1*6,l,I.LE.ll - TRANSFER TO (SWLNBN) PROVIDED (A(I).NE.B(I)) - CKCDS CONTINUE - FORMNO*FOEMNO/l - DO THROUGH (KOVEL), 1=0,1,1.LE.5 - HOVEL BLOCK(FOEMNO,I)=B(I; - BLOCK(FORMNO,6)=B(12) - TRANSFER TO (READB) - ERROR WRITE OUTPUT ,EZREKSG - CLOSE UNLOAD , INTAP - CLOSE REWIND ,OUTAP - LAST CLOSE REWIND , $SGLTNQ$ - OPEN NO REWIND ,$SGLTNI$ - WRITE OUTPUT ,HED2 - RDSGL CALL SUBROUTINE ( )=RDNSFC. ( $SGLTNI$, A, 13 , ENDSGL, ' ERR0R3) - CALL SUBROUTINE ()=DE)CODE. (A, 3 6 ) - WRITE OUTPUT ,FMT2, ((A(l), 1=0,1,1.LE.12)) - TRANSFER TO (RDSGL) - ENDSGL CLOSE REWIND ,$SGLTNI$ - WRITE OUTPUT ,ENIMSG - CALL SUBROUTINE ()=H'IDJOB. () - ERROR3 WRITE OUTPUT ,EFMSG3 - CALL SUBROUTINE ()=HJDJQB.() - SWLNEN LINDQN=THUE - C HEMISTICH CODES DIFFER AND CCMPAREE HEMISTICH IS RETURNED FOR NEXT EUNTHROUGH. - CALL SUBROUTINE ()=WRNSFC. (0UTAP,B,13) - TRANSFER TO (READB) - EOFIN CLOSE REWIND ,INTAP - CLOSE REWIND ,OUTAP - PRINT TRANSFER TO (WEITEA) PROVIDED (FOEMNO. E.O) - PROVIDED (RRLC. ( ) . L. FOEMNO/^), WRITE OUTPUT ,SKIP - WRITE OUTPUT ,HEEL, ((A(l),1=6,1,1.LE.ll)) - DO THROUGH (K0VE2),I=O,l,I.LE.5 - M0VE2 BLOCK(0,l)=A(l) - BL0CK(0,6)=A(12) - C ALPHABETIC ORDERING OF GROUP OF HEMISTICHS WITH THE SAME MEANING CODE FOLLOWS, - DO THROUGH (WHFMLA), 1=0,1,I.LE.FORMNO - DO THROUGH (ALFORD), J=I/1,1, J . LE. FOEMN 0 - 199

DO THROUGH (CHKORD),K=0,1,K.LE.5 - PROVIDED (BLOCK(I,K).LG.BLOCK(J,K)), TRANSFER (SWAP) - PROVIDED (BLOCK(I,K).LL.BLOCK(J,K)), TRANSFER (ALFORD) - CHKORD CONTINUE - TRANSFER TO (ALFORD) - SWAP DO THROUGH (SWAPl),K=0,l,K.LE.6 - T3'iP»BL0CK(J,K) - BLOCK(J,K)=BLOCK(I,K) - BLOCK(I,K)=TEMP - SWAP1 CONTINUE - ALFORD CONTINUE - C FORMULA GROUP, IN ALFORDER, IS NOW TO BE DECODED - UNCODE CALL SUBROUTINE ( )=DECODE. (BLOCK(1,0), 3 6 ) - C DECODED FORMULA GROUP IS NOW TO BE WRITTEN OUT - WRITE OUTPUT ,FMT27,((BLOCK(I,J),J=0,1,J.LE.6 )) - WHFMLA CONTINUE - TRANSFER TO (WINDUP) - WRITEA CALL SUBROUTINE ()sWRNSFC.($SGLTN0$,A,13) - WINDUP TRANSFER TO (LAST) PROVIDED (.NOT.LINDON) - TEMP=OUTAP - OUTAP=$OUTO $(/)INTAP(*).0.770000 - IHTAP=$INPO $(/)TEMP(*).0.770000 - NEXT HUNTHROUGH FOR FOLLOWING HEMISTICH - TRANSFER TO (OPTAP) - END PROGRAM (FIRST) - BIBLIOGRAPHY

I. Editions of Texts

Bayot, Alphonse, ed. Gormont et Tsflmhart. Fragment de Chanson de Geste du XTT^e Siecle. Paris: Librairie Honore Champion, y a

Bran din, Louis, ed. La Chanson d*Asprepiont. Chanson de Geste du Xlle Siecler deuxieme edition revue. 2 vols. Paris: Librai- rie HonoreChampion, 1923-2^.

Cloetta, Wilhelm, ed. Les deux Redactions en Vers du Moniage Guil- laume. Chansons de Geste du XJTje Siecle. 2 vols. Paris: FiTrain Bidot et Cie,, 1906-11.

Densusianu, Ovide, ed. La Prise de Cordres et de Sebille. Chanson de Geste du H i e Siecle. Paris: Firmin Didot et Cie., I8 9 6 .

Henry, Albert, ed. Les OEuvres d ,Adenet le Roi. Vol. I: Biogra­ phic d ,Adenet — La Tradition manuscrite. Vol. H : Buevon de Conmarchis. Vol. H I : Les Ehfances Qgier. Rijksuni- versiteit te Gent, Werken uitgegeven door de Faculteit van de Letteren en WLjsbegeerte, nos. 109, 115 and 121. Bruges: "De Tempel,» 1951-56. Vol. IV: Berte aus grans Pies. Uni- versite libre de Bruxelles, Travaux de la Faculte de Philo- sophie et Lettres, no. 23. Brussels: Presses Universitaires, 1963. Henry, Patrice, ed. Les Ehfances Chanson de Geste du Xllle Siecle. Paris: Firmin Didot et Cie., 1935*

Langlois, Ernest, ed. Le Couronnement de Louis. Chanson de Geste du YTTe Siecle. deuxieme edition revue. Paris: Librairie Honore Champion, 1925.

McMillan, Duncan, ed. La Chanson de Guillaume. 2 vols. Paris: Editions A. et J. Picard et Cie., 19^*9.

Meyer, Parol and Auguste Longnon, eds. Raoul de Cambral. Chanson de Geste. Paris: Firmin Didot et Cie., 1882.

200 201

Perrier, Joseph Louis, ed. Le Siege de Barba sire. Chanson de Geste du X H e Siecle. Paris: Librairie Honore Champion, 1926.

______, ed. Le Charroi de Nlmesr Chanson de Geste du XTTe Siecle. Paris: Librairie Honore Champion, 1931•

Raynaud, Gaston, ed. Elia de Sa^nt Mile. Chanson de Geste. accom- pagnee de la redaction norvegienne traduite par Eugene Koel- bing. Paris: Firmin Didot et Cie., 1879*

RueUe, Pierre, ed. Huon de Bordeaux. Universite lib re de Bru­ xelles, Travaux de la Faculte de Philosophic et Lettres, no. 20. Paris: Presses Univer sitaires de France, I960.

Stimming, Albert, ed. Per festlandische Bueve de Hantone. Fassung II. Gesellschaft fur romanische Literatur, no. 30* Halle: Max Niemeyer Verlag, 1912 .

Van Hasselt, Andre, ed. Li Roumans de deomades. par Adenes li Rois. 2 vols. Brussels: Victor Devaux et Cie., 1865.

Weinbeck, Erich, Wilhelm. Hartnacke and Paul Rasch, eds. All scans. Halle: Max Niemeyer Verlag, 1903.

Whitehead, Frederick, ed. La Chanson de Roland. Oxford: Basil Blackwell, 19^6.

Yeandle, Frederic G., ed. Girart de Vienne. Chanson de Geste. New York: Columbia University Press, 1930*

H . Critical Studies of Le Siege de Barbastre and Buevon de Conmarchis

Becker, Philipp August. "Der Siege de Barbastre." in Beitrage zur romanisehan philologie. Festgabe fur G. Grober. Halle: Max Niemeyer Verlag, 1899*

Bovy, Arthur. Adenet le Roi et son OEuvre. Brussels: Vromant, I898,

Fripiat, M.-C. ttLes Motifs et les Formules dans quatre Chansons de Geste: All scans., la deuxieme Redaction en Vers du Moniage Guillaume. La Mort Avmeri de Naibonne et Le Siege de Bar­ bastre.11 Unpublished these de licence. University of Lou­ vain, I960. 202

Gundlach, Adolph. Das Handsdirlftenverhaltnis des "Siege de Baru bastre.” Ausgaban und Abhandlungen aus dern Deblete der romanischen Philologie, no. 4. Marburg! ELwert*sche Ver­ lag sbuchhandlung, 1883*

Keller, Victor. HLe Siege de Barbastre** und die Bearbeitung von Adenet le Hoi. Inaugural-Dissertation zur Erlangung der Doctorwurde bei hochloblicher philo sophischer Faoultat zu Jena. Marburg: B. E. Sipmann, 1875*

Roll, Erich. Untersuchungen uber das Verhaltnls des -Siege de Barbastre” zum -Bueves de Commarehiaw yon Adenet le Roi. und die Stellung der Prosafassung. Inauguraldissertation zur Erlangung der Doktorwurde der philosophischen Facul- tat der koniglichen Universitat Greifswald. Greifswald: Buchdruckerei Hans Adler, 1909.

Weeks, Raymond. "The Siege de Barbastre." Romanic Review. X (1919), 287-321; XI (1920), 349-369; XII <1921), 57-61.

H I . Literary and Linguistic Studies

Aebischer, Paul. "Halt sunt li pui e li port tenebrus,” Studi Medieval!. Nuova Serie, X V H I (1952), 1-22.

______. Rolandiana borealia; la Saga af Runzivals Bardaga et ses Derives scamdinaves compares a la Chanson de Roland. Lettres, Sciences, Techniques, Publications de la Faculte des Lettres de Lausanne, no. 11. Lausanne: F. Rouge, 1954. Alonso, D&naso. nLa primitiva epica francesa a la luz de una Mota amilianense.M Revista de Filologia Espanola. XXXVII (1953), 1-94. Bedier, Joseph. Les Legendes epiques. 4 vols. Paris: Librairie Honore Champion, 1908-14.

Bon jour, Adrien. "Poesie heroique du Moyen Age et Critique lit- teraire.»* Romania. L X X V H (1957), 243-55.

Bowra, Cecil M. Heroic Poetry. London: Macmillan, 1952.

______. "Style," A Companion to Homer, ed. Alan J. B. Wace and Frank H. Stubbings (London: Macmillan, 1 9 6 2 ), pp. 26-37. 203

Brodeur, Arthur G. The Art of Beowulf1. Berkeley: University of California Press. 1959. 9 T Burger, Andre. MLa Question rolandianne, Faits et Hypotheses." Cahiers de Civilisation Me die vale. 17 (1961), 269-91.

Campbell, Jackson J. "Oral Poetry in The Seafarer.11 Speculum. ZXZ7 (I960), 87-96.

Comfort, William Wistar. "The Literary Role of the Saracens in the French EJpic." Publications of the Modern Language Association. LV (1940), 628-59.

Creed, Robert P. “The Andswarode-System in Old English Poetry." Speculum. XXXEE (1957), 523-28.

De Cock, C. "Etude descriptive des Cliches epiques releves dans Ratailjla T/vpifar Tr Bueve de Hantone H I . Covenant Vivien." Unpublished these de licence. University of Louvain, I960.

DelbouiUe, Maurice. Sur la Genese de la Chanson de Roland (Tra- vaux re cents— Propositions nouvellesT.Brussels: Palais des Academies, 1954.

______. "Les Chansons de Geste et le Livre." La Technique lit- teralre des Chansons de Geste: Actes du Collogue de Liege (Septembre 1957y (Paris: Societe d*Edition "Les Belles Let- tres," 1959), PP. 295-40?.

Desert, J.-P. "Les Cliches dans Gerbert de Metz et Anseys de Metz. Chansons de Geste franpaises du Moyen Age." Unpublished these de Hnan c e T University of Louvain, I960.

Diamond, Robert E. "The Diction of the Signed Poems of Cynewulf," PMInlng W l Quarterly. XXXVUI (1959), 228-41.

Fawtier, Robert. La Chanson de Roland. Paris: E. Boccard, 1933.

Flutre, L.F. "Etudes sur le Roman de Perceforet." Romania. LXXI (1950), 374-92. . "Dleudonne de Honerie. Chanson de Geste du XlVe Siecle (alias Roman de Charles le Chauve)." Zeltschrlft fur Roman- ische Philologie. L X V U I (1962)7"321-400.

Foerster, Wendelin. "Etymologian,M Zeitschrlft fur Romanische Philologie. X X X H (1908), 423-55. 204-

Fou che, Pierre. Le Verbe frencais. Publications de la Faculte des Lettres de l'Universite de Strasbourg, no. 56. Paris: Societe d* Edition "Les Belles Lettres," 1931*

Frappier, Jean. "Reflexions sur les Rapports des Chansons de Geste et de l*Histoire." Zeitschrlft fur Ramanische Philologie. L X X m (1957), 1-19.

. "Les Destriers et leurs Efc>ithetes." La Technique llt- teraire des Chansons de Geste: Actes du Collogue la fSeotembre 1957) (Paris: Societe d»Edition "Les Belles Lettres," 1959), pp. 85-104-.

Godefroy, Frederic. DLctlonnaire de l»ancienne Langue francaise. 10 vols. Paris: F. Vieweg, 1888-1902.

Gsteiger, Manfreid. "Notes sur les Preambules des Chansons de Geste." Cahiers de Civilisation Medievale. II (1959), 213-20.

Horrent, Jules. La Chanson de Roland dans les Litteratures fran­ caise et espagnole au Moyen Age. Paris: Societe d* Edition "Les Belles Lettres," 1951.

Jodogne, Oner. "Sur 1'Originalite de Raoul de Cambrai." La Tech­ nique^ litteraire des Chansons de Geste: Actes du Collogue de Liege (Septambre 1057) (Paris: Societe d ,Edition "Les Belles Lettres," 1959), pp. 37-58.

Jordan, Leo. "Von der Fassung des Problems, dem Beweisgang, Wlrk- Hriike-^t. Moglichkeit und Irrtum In der Sprachwissenscihaft." Archivum Romanicum.H (1925). 292-98.

Jpusse, Marcel. Etudes de Psychol np^e Unguistique: le Style oral rvthmique et mnanoteehnique chez les Verbo-moteurs. Archi- ves de Philosophie. vol. H . cahier 4. Paris: Beauchesne, 1925. Lambotte, F. "Les Formules epiques dans la Chanson de Roland. Manuscrits d*Oxford et de Venise." Unpublished these de licence. University of Louvain, I960.

Langlois, Ernest. Table des Noms oronres de toute Nature compris dans les Chansons de Geste imprdmees. Paris: Librairie Eni.le Bouillon, 1904-.

Le-Gentil, Pierre. La Chanson de Roland. Paris: Hatier-Boivin, 1955. 205 Lejeune, C. "Fornrules et Technique foxraulaire des Combats singu- liers dans la Litterature epique." Unpublished these de licence, University of Louvain, i9 6 0 .

Lejeune, Rita. Recherches sur le Theme; Les Chansons de Geste et l'Histoire. Bibliotheque de la Faculte de Philo sophie et Lettres de 1*Universite de Liege, no. CVHI. Paris: Faculte de Philosophie et Lettres, 19*18.

______• “La Naissance du Couple litteraire •Roland et Olivier. •“ Annuaire de l BInstitut de Philologie et d'Histoire orien- talesr X (iqqo^/ T76-400.

• “One Allusion meconnue a une Chanson de Roland." Romania. LXXV (195*0, 1*15-6*1.

______• “Technique fonnulaire et Chansons de Geste." Moven Age. LX (195*0, 311-3*1-.

______. "Aetualite de la Chanson de Roland.M La Nouvelle Clio. V H (1955), 207-2 7 . • Levy, Raphael. Chronologic approximative de la Litterature fran­ caise du Moyen Age. Beihefte zur Zeitschrift fur Roman- ische Philologie, no. 98. Tubingen: Max Niemeyer Verlag, 1957. Lonnrot, Elias. The Kalevala. or Poems of the Kaleva District. A Prose Translation -with Foreword and Appendices by Fran­ cis Peabody Magoun, Jr. Cambridge; Harvard University Press, 1 9 6 3.

Lord, Albert B., ed. SerboCroatian Heroic Songs, collected by Milman Parry. Vol. I; Novi Pazar: English Translations, with musical transcriptions by Bela Rartok and prefaces by John H. Finley, Jr. and Raman Jakcbson, Vol. H : Novi Pazar: Serbocroatian Texts, with introduction and notes by the editor and with a preface by A. Belie. Cambridge and Belgrade: Harvard University Press and the Serbian Academy of Sciences, 1953-5*1.

______. The Singer of Tales. Harvard Studies in Comparative Literature, no. 2*f. Cambridge: Harvard University Press, I960.

______. "Homer and Other E£ic Poetry." A Companion to Homer. eds. Alan J.B. Vfa.ee and Frank H. Stubbings (London: Mac- millan, 1 9 6 2 ), pp. 179-21*1-. 206

Lot, Ferdinand. Etudes sur les Legendes epiques francaises. With an introduction by Robert Bossuat. Paris: Librairie Honore Chanqiion, 1958*

Louis, Rene. De l'Histoire a la Legende. Vol. I: Girart. Comte de Vienne („ . . 819-877) et ses Fondations monastiques. Vols. II and H I : Girart. Comte de Vienne dans les Chansons de Geste: Girart de Vienne. Girart de Fraite. Girart de Roussillon. Auxerre: Bureaux de l'Lmpriraerie Mo dome, 1946-/1-7.

Magoun, Francis P., Jr. "The Oral-Formulaic Character of Anglo- Saxon Poetry." Speculum. XXVHI (1953), 446-67.

______. "Bede's Story of Caedman: The Case History of an Anglo- Saxon Oral Singer." Speculum. XXX (1955), 49-63.

Menendez Pidal, Ramon. La Chanson de Roland et la Tradition epi­ que des Francs, deuxieme edition revue et mise a jour par 1*auteur avec le concours de Rene Louis, traduite de l'es- pagnol par I.-M. Cluzel. Paris: A. et J. Picard et Cie., I960.

Nichols, Stephen G.,. Jr. Formulaic Diction and Thematic Composi­ tion in the Chanson de Roland. University of North Caro­ lina Studies in Romance Languages and Literatures, no. 36. Chapel Hill: University of North Carolina Press, 1961.

O'Neil, Wayne A. "Another Look at Oral Poetry in the Seafarer." Speculum. XXXV (i96 0 ), 596-6OO.

Paris, Gaston. Histoire poetiaue de Charlemagne. Paris: Bouillon, 1865.

Parry, Milman. L»Epithete traditionnelle dans Home re. Paris: Societe d'Edition "Les Belles Lettres," 1928.

. Les Formulas at. la Metriaue d'Homere. Paris: Societe d'Edition "Les Belles Lettres," 1928.

. "Studies in the ^pic Technique of Oral Verse-Making, I: Homer and Homeric Style." Harvard Studies in Classical Philology. XLI (193°), 73-1W. ______. "Studies in the Epic Technique of Oral Verse-Making, H : The Homeric Language as the Language of Oral Poetry." Har­ vard Studies in Classical Philology. XT ■HI (1932)* 1-50. 207 Riquer, Martin de# Les Chansons de Geste francaises. deuxiane edition entierement refondue, tradulte de l'espagnole par I.-M. Cluzel. Paris: Nizet, 1957*

"Epopee jongleresque a ecouter et epopee romanesque a lire." La Technique litteraire des Chansons de Geste: Actes rfii Collogue H r (Serptepibre 1957) (Paris: Societe d*Edi­ tion "Les Bellies Lettres," 1959), pp« 75-8^»

Rosmeulen, S. "Les Cliches dans les Chansons de Geste ETHe rie Saint- tuna et Folque rie CnnH-5a.»» Unpublished these de licsncer University of Louvain, I960.

Roy, Bnile. La Versification et le Style de Hugues Capet, de Baudouin de Sebourc et du Bastard de Bouillon. Publica­ tions de l'Universite de Dijon, no. 2. Dijon and Paris: A. et J. Picard et Cie., 1929#

I^ychner, Jean. La Chanson de Geste. Essai sur I'Art epiaue des Jongleurs. Geneva: Droz, 1955*

______. "Observations sur la Versification du Couronnecient de Louis." La Technique litteraire des Chansons de Geste: Actes riii Colloqne de liege (Septembre 1957) (Paris: Societe d 1 Edi­ tion "Les Belles Lettres," 1959), PP» 161-82.

Schaar, Claes. "On a New Theory of Old Ehglish Poetic Diction." Neonhilologus. XL (1956), 301-05.

KLootmaekers, M. "Etude descriptive des Cliches epiques relevees dans Otinel. La Prise de Cordres et Sevile. Le Moniage Guil­ laume. " Unpublished these de licence. University of Lou­ vain, I960.

Stevick, Robert D. "The Oral-Fonaulaio Analyses of Old Siglish Verse." Speculum. XXXVII (1962), 382-8 9 .

Thompson, Stith. Four Symposia on Folklore. Bloomington: Indiana University Press, 1953*

Troch, H. "La Stylistique epique. Essai stir l'Srploi des Formulas et des Cliches dans les Chansons de Geste: Aiol et Mirabel. Aye d*Avignon. Aspremont. Fierabras. Unpublished these de licence. University of Louvain^ 1 9 ^0 .

Tyssens, Madeleine. "Le Probleme du Vers orphelin dans le *Cycle d ,Aliscans* et les deux Versions du Moniage Guillaume." La Terihrpnue lntt4ra-ii»e des Chansons de Geste: Actes du Collogue rift Triage (Septembre. 1957) (Paris: Societe d* Edition "Les Belles Lettres," 1959), pp. ^-29-56. 208

Van de Walle, A.M, “Etude comparative des Cliches epiques dans Eloovant. Do on de la Roche. Gui de Nanteuil. Amis et AmilflH et La Chanson des Saxons. Chansons de Geste des X H e et X H I e Siecles." Unpublished these de licence. University of Louvain, 1960.

Von Wartburg, Walther, ed, Franzosisches Etvmologisches Worter- buch. Vols, I-V, VII-IX, XEV and XVI, and fascicles 6z, 6 6 , 72-75, 78-81 and 88-91 have so far appeared. Various publishers from Bonn: Fritz KLopp Verlag, 1928 to Basel: Zbinden, 1963.

Waldron, Ronald A, “Oral-Formulaic Technique and Middle English Alliterative Poetiy." Speculum. XXXII (1957), 792-804,

Webber, Ruth House, Fonnulistic Diction in the Spanish Ballad. University of California Publications in Modern Philology, XXXTV, 175-278, Berkeley: University of California Press, 1951.