Version classiqueVersion mobile

Mélanges de langue et littérature françaises du Moyen Âge offerts à Pierre Jonin

Jordan Fantosme's experiments in prosody and design

R.C. Johnston

Texte intégral

1Every scholar who has worked on Jordan Fantosme's Chronicle (composed about A.D. 1175) and who has not rejected the decasyllabic section as spurious has taken it to be a 'mixed-prosody text', i.e. one not written in the same metre throughout. Examples of this genre occur in Medieval Latin (e.g. Abelard's planctus Dolorum solacium), in Old Provençal (e.g. Le Jeu de sainte Agnès), and in Old French (e.g. Le Roman de Rou, Aiol, Piramus et Tisbé, et.) while in Anglo-Norman Ph. de Thaon used hexasyllabics and octosyllabics in his Lapidary, Elias of Winchester wrote in five different metres, one for each of the Prologue and four books of his Afaitement Caton, and Matthew Paris wrote his Saint Auban mainly in Anglo-Norman alexandrines but, says his editor (Harden 1968), 'on occasion it would appear that the author quite consciously changes the metre for a specific purpose'. Chronologically Hordan Fantosme comes between Elias of Winchester, c. 1140, and Matthew Paris, c.1235.

2That the Chronicle is a mixed-prosody text is agreed, but there is no agreement on what the mixture is. Rose (1880) and Becker (1944) opted for decasyllabics plus alexandrines ; Macaulay (1899) and Macdonald (1961) independently saw the text as decasyllabics plus lines in which 'each half verse(had) two accents but an uncertain number of syllables', as Macaulay phrased it; Suchier (1913) found lines of 10, 12, 14, and 16 syllables; Vising (1882 and 19é") concluded that the variations in syllable-count were incomprehensible and that Fantosme was satisfied with lines of approximately the same length ; Legge (1971) recognised A.-N. alexandrines, decasyllabics, and a short passage of 14-syllable lines.

3My approach (Johnston 1974) is that of Macaulay and Macdonald in so far as I consider not whole lines but hemistichs, but I differ fundamentally from them in aligning myself with all other cri tics and considering the hemistichs, and indeed the whole lines, to be syllabic and not accentual in construction.

4I take the final hemistichs of the lines first, for these, except for lines 2039-59 (in the numbering of Howlett 1886) are treated in the same way throughout the whole poem. They carry the rhume-words that link lines into mono-rhymed laisses and separate one laisse from the next. They give unity and cohesion to the whole work. They correspond to six-syllable hemistichs of Continental French, where each one has six syllables if the rhyme-word is masculine plus a final supernumerary -e if feminine. In Anglo-Norman this count was not rigorously observed. Quite apart from the possibility that an internal feminine e might drop from pronunciation and convert six syllables for the eye to less than that for the ear, many final hemistichs have more or less than six syllables and the count may be termed 'six or thereabout'. In his treatment of final hemistichs, Jordan Fantosme is completely in line with the normal practices of his A.-N. contemporaries.

5When we turn to laisses of decasyllabic or of alexandrines lines, we find little to comment on or to cause surprise in Fantosme's handling of them.

Oëz, seignurs, qu'avient de trop ultrage
Ke lur avint d'Escoce la salvage !
Bel fud lu tens senz nul malveis orage.
Li reis d'Escoce esteit de fier corage,
Chevalier bon e de grant vasselage.
De devant lui vint errant un message;
Chanuine fud, si saveit le language,
Hastivement li cunta sun damage.
(Laisse 78, lines 706-13)

6True, not all of the 120 decasyllabic lines are as accurately counted as these, but Fantosme is A.-N. in not being rigorous about the fours and the sixes, and some of his aberrant lines, in this section as elsewhere, are written deliberately as 'signals' to indicate a change in subject-matter or metre, as is the case in the first line, 646, of the decasyllabic section Li reis fait sa gent apparaillier.

7In the following alexandrine laisse two of the thirteen lines are incorrect by continental standards but acceptable in England, and the first of them could be defended as a 'signal', while the other scans 5/7, and may, along with the far commoner 7/5 lines and the rare 8/4, be allowable as A.-N. variants of a twelve-syllable line.

Mes un chevalier noveaus lur iert le jor venuz.
Ore oiez de ses fez e de ses granz vertuz!
P Puis que ses cumpaignuns se furent tuit renduz,
Remist il en la tur e saisi dous escuz
Sis pendi as kerneaus. Lungement s'est tenuz,
E lança as Escoz treis espiez esmuluz ;
A chascun des espiez ad un mort abattuz
Quant ceus li sunt failliz, si reprent peus aguz
E lançad as Escoz, si en ad cunfunduz.
E tuz jors vait crïant: 'Ja serrez tuz vencuz !'
Unc d'un sul vassal ne fud estur mielz tenuz !
Quant li fus li toli le defens des escuz.
Ne fait pas a blasmer, s'il s'est idunc renduz.
(Laisse 155, lines 1499-1511)

8The above are fair specimens of the twelve laisses of decasyllabics and the seventy laisses of alexandrines, which together make up over 800 of the 2065 (2071 in Howlett) lines of the Chronicle ; they are metrically quite unremarkable in the productions of A.-N. poets of the eighth decade of the twelfth century.

9It is the rest of the poem (135 laisses, some 1350 lines) that has aroused all the pother about Jordan Fantosme's versification. It is a fair assumption that these remaining laisses are not likely to correspond with accepted A.N. norms, and arguable that they must be something not found, or not hitherto found, in the work of A.-N. or indeed O.Fr. writers, for otherwise any or all of the scholars who have studied Fantosme would have recognised and identified his metres. It would be unreasonable to assume that Fantosme, a model of poetic technique so far, had suddenly become innumerate or that a common ancestor of the Durhamand Lincoln manuscripts had gone berserk and sabotaged the syllable count of over half the poem while leaving grammar and sense intact.

10Compare the following apparently alexandrine laisse with

11155 above :

Quant cel cunseil fud duné, li reis nel crut mie.
Ne remaindra la guerre pur perdre Albanie.
Mes il voldrad ultre mer enveier un espie
Pur veeir le cuntienement le pere en Normendie,
E puis avant en Flandres al fiz en qu'il se fie,
Ses briés e ses messages ki haltement li die
Cume li reis sun pere par dit me cuntralie
E par itel manace cum vus avez oie.
E s'il volt tenir cuvenant par fiance plevie,
Ne larrai ne li face prueçainement aïe.
Si nus enveit de Flandres ses Flamens od navie,
Par cent e par cinquante de cele gent hardie,
Jo lur livrai la veie la gent qui nus guerrie,
Cil assaudrunt les chastels par force d'establie
(Laisse 36, lines 408-21)

12Of the initial hemistichs nine have six syllables, three have seven, two have eight. Does the comparison indicate madness ? If so, though this be madness, yet there is method in it. The nine short hemistichs are all feminine, ending in a syllable containing a supernumerary -e, the five long ones are masculine. There are in the Chronicle 92 laisses of this kind and the distinction holds throughout : the masculine hemistichs (m) are long with 7 or 8, occasionally 9 or 6 syllables, the féminines (f) are short with 6, occasionally 5, syllables. What Fantosme has done is to sort out the confusion of the 'six or thereabout' count of the A.-N. alexandrine, and to turn it into a regularised Jordanian alexandrine line.

13Closely allied to these are at least nine laisses whose alexandrines are Jordanian but which have as an extra characteristic a tripartite division indicated by sense-breaks and reinforced by patterning of m and f.

Ore chevalche a seür le grant ost d'Engleterre.
Li sires Richart de Luci - nul meillur n'estuet querre -
Bien aïdet sun seignur a maintenir sa guerre.

E si set en buisine triewes e pais requerre
La u il veit la force e li estuet requerre.
(Laisse 82, lines 766-70)

The laisse divides : m / m m / f f.

Mult iert de grant affaire dan Humfrei de Boün,
Li barun de Northumberlant en sunt si cumpaignun.

Arstrent tut Berewic a flambe e a tisun
E une grant partie de terres envirun,
Ke perent en lur marches, crüel cume leün.

Mes dan Richart de Luci n'ad suin de tel sermun,
Si dit en sun language : 'Sire Humfrei de Boün,
A ! si Deu n'en prenge cure, n'i frum si perdre nun.'
(Laisse 86, lines 801-9)

14Here we find : f m / m f f / m f f.

15Nearly at the end of the Chronicle come three laisses, 214-6, lines 2039-59, which I call 'double stanzas', in which both hemistichs, initial and final, vary the syllable-count in such a way as to produce patterns characteristic of lyric poetry :

Or cumande le rei d'Escoce hastivement a sei mener,
Kar li vint une novele : passer li estuet mer.

Roem est asise, sa cité ; ne voldra plus demurer.

Il enmeine David od lui, si vait envers la mer.
E Brïen s'en est turné, ki n'ad cure de demurer.

E ad dit a sun seignur ignelement face amener
Le rei d'Escoce a Suhantune, li reis volt qu'il passe mer.
(Stanza 215, lines 2049-55)

16The initial hemistichs:8 7/8 / 8 7 / 7 8

17and the finals: 8 6/7 /6 8/8 7

18So far I have given samples of the following groups: 12 decasyllabic-line laisses, 70 A.-N. alexandrines, 92 Jordanian alexandrine laisses plus 9 with a special feature making 101 in all, and three double stanzas: total 186. These figures are not unshakable, because it is possible to scan a few laisses in two ways.

  • 1 The m resulting from elision is designated mf.

19Of the 217 laisses comprising the whole poem there remain 31, on two of which I am not ready to express conviction, but 29 of them seem to be characterised by a different treatment of the feminine -e that ends the initial hemistich at the caesura; it is syllabicwhen a consonant follows either in the same word or at the beginning of the final hemistich, and is elided before a following vowel1 It is not supernumerary. If we look back for precedents, we find that for Elias of Winchester final -e of a line or hemistich was always syllabic. Thus in the section of hexasyllabics grouped in tail-rhymed couplets lines 267-72 run :

La deit hum travaillier U l'um set sun luier
Receivre prestement
E d'illoc retrairë De servise fairë U l'um
nul bien atent.

20In the 8 + 6 quatrains with internal rhymes :

N'aies vers nului envië N'en bien n'en destrescë;
Ki a lui prent cumpainië Tuz tens ad tristescë
(Lines 387-90)

21In Horn this -e can be supernumerary, syllabic before a vowel or a consonant, elided before a vowel; (Pope 1964 : 24-27). Fantosme, in these 29 stanzas, adopts an unbending technique, which is also found much later, in the Cinkante Balades of John Gower :

Et vous, ma damë, croietz bien cela (11,7)
O noble dame, a vous ce lettre irra (11,25)

22I use the term 'stanza' and not 'laisse' because, as in the double stanzas, the arrangements of hemistichs, but here the initial hemistichs only, create lyric-type patterns in their syllable-count, and expand this to 'Jordanian stanza' because I thtnk Fantosme was its inventor. Matthew Paris followed in his wake in stanzas 4, 8, and 10 of his Vie de saint Auban.

23The first two stanzas of the poem illustrate how Fantosme uses his lines. In l we have three verses, each metrically different, framed by single lines. Scanning the initial hemistichs we get : seven-syllable framing hemistichs enclosing 6 8 7 ' 6 8 7 / 7 8 ' 7 8 / 7 7 7 ' 8 8 8. The framing-lines are isolated by sense-breaks : a break occurs in the middle of each verse ; the sense runs over from verse to verse. If the simple definition of Verrier (1932) 'un poème dont les vers se divisent en groupes de mètre différent' be accepted, what we have here is an embryo descort, based on the Mediaeval Latin sequence as used by Abelard, and the first in a vernacular, antedating the O.Fr. descorts of Gautier de Dargies and Colin Muset, which, according to Jeanroy, themselves antedate the first descort in O. Pr.

Oëz veraie estoirë : que Deu vus beneïe !

Del mieldre curuné qui unkes fust en vie
Talent m'est pris de faire vers, dreiz est que jos vus die.
Celui tieng a sage qui par autre se chastie.
'Gentil rei d'Engleterre a la char tres hardie,
Al curuner de vostre fiz ne vus suvienge il mei
Ke l'umage de ses meins le rei d'Aubanie

Li feïstes presenter sens fei aver mentie?
Puis lur deistes ambesdous : "Deus les maldie
Ki de vus departirunt amur ne drüerie !
Encuntre tutes genz del mund en force e en aïe
Od mun fiz seiez tenant, salve ma seignurie !"
Puis entre vus e vostre fiz mortel nasquid envie,

Dunt maint gentil chevalier ad puis perdu la vie.
Maint hume deschevalchié, mainte sele voidie,
Maint bon escu estroé, mainte bruine faillie.
Aprés icest curunement, e aprés ceste baillie,
Surportastes a vostre fiz auques de seignurie,
Tolistes lui ses volentés, n'en pot aver baillie.

La crut guerre senz amur: Damnesdeus la maldie !'
(Stanza 1, lines 1-20)

242 can be edited in two ways, either as a Jordanian laisse with m 7 and f 6, or as a Jordanian stanza giving a tripartite lyric of the Provençal type with frons and pes each 7 8 7 7 and cauda 7 6. Not only is this vastly superior in deftness to the m m m f m m f f f f of the laisse, but it is a habit of the poet to introduce a new theme - and this is the first stanza concerning the events of 1173 - by an outstanding piece of prosody, and, also, the following laisse 3 is a conspicuous example (so conspicuous that it must be intended to alert the reader to something novel) of the Jordanian laisse, having eleven f and only one m. Here is stanza 2 :

Reis de terre senz honur ne set bien que faire:
Nu sout li juefnes curunez, li gentilz debonaire.
Quant ne pot ses voluntez acumplir pur sun paire,
Pensout en sun curage qu'il li fereit cuntraire.

Turnad s'en tut a celé, passad un gué de Leire ;
Desi qu'il vint a Saint Denis ne volt mangier ne beire.
Cuntad al rei de Francë trestut le suen afaire.
Mandent celui de Flandrës, Felipe le Puignaire,
E Maheu de Buluinë, qu'il venist od Sun fraire.
Grant fud cele asemblee, une ne veïstes maire.

25It seems likely that a poet ingenious enough to use five different metres (or six if one differentiates between the 92 J.L. and the 9 tri-partite J.L.), and to be novel and experimental in three or four of them, worked to an overall plan and did not scatter these metres haphazardly about his poem. What is the evidence for design ?

26The Chronicle itself is tripartite, each part dealing with events of different years and having different metrical characteristics. The introductory background stanza treats of happennings before 1173 in the framed descort of Augustan impressiveness printed above. Part I, 2-109, 11. 21-1101, devoted to the events of 1173, and Part II, 110-217, 11. 1102-2071, to those of 1174, have an equal number of laisses, one hundred and eight. The introductory stanza, the first stanza of Part II and the last stanza of the whole poem rhyme in -ie and have descort features, while a Jordanian stanza begins Part I.

27In Part I Jordanian laisses greatly outnumber A.-N. alexandrines ; the reverse is true of Part II. There are no decasyllabics in Part II and no double stanzas in Part I.

28The Jordanian and double stanzas, with their affinities with the lyric, differ markedly from the laisses, whose associations are with epic and narrative literature, and they are differently used, superimposing colour and contrast on the factual information they convey. Thus it is that they appear in the metrical fireworks of the Introduction and the six stanzas of the finale, they contrast with laisses to compose set-pieces in Part II, and they occasionally serve to signal changes of subject-matter. An obvious example occurs in the middle of the graphic description of the interview of the bishop of Winchester with Henry II. The rhyming quatrains that give the swift exchange of question and answer assonate in pairs (thus 165 sunt etc. assonates with 166 traïsun etc. and 169 Karduil etc. with 170 pris) and are written in alexandrines. But 167 and 168 report william the Lion's attempt to bribe and threaten Carlisle into surrender ; they rhyme together and are in stanza form. Here they are, with their surrounding laisses :

Ad dunc Robert de Vaus faite nule traïsun ?
Ad il Carduil rendu ? Ne dites si veir nun !'
'Mes le tient noblement cume gentil barun.
De sun grant desturbier dreiz est que vus dium.

Li reis d'Escoce vint l'autrier par Carduil chevalchant,
E dan Robert de Vaus durement maneçant,
Demande lui le chastel par itel cuvenant
Qu'il li durreit asez dunt il serreit manant.

E s'il (i) ço ne fait des idunc en avant,
Tuz les ferad afamer li petit e li grant.'
'Par ma fei,' dist li reis, 'ci ad bon cuvenant.
En poi d'ure Deu laburë, ço dit le mendiant.

Ke fist dune li Escot ? Asiega il Karduil ?
'Nenil, sire, si vus pleist, ainz fist greinur orguil :
Ainz ad pris Appelbi, dunt jo forment m'en duil,
E le chastel de Burc, bien acuinter vus voil.'
(Lines 1572-87)

29Association of metre with subject-matter runs through the whole poem and in the space available anything like a complete treatment is out of the question. The best that can be hoped for is that a broad outline with a few details will convey a reasonably accurate impression.

30The narrative of Part I breaks down into six sections ; two of them are homogeneous, four are not. The two are 70-81, 11. 646-765, the Scots at the siege of Carlisle, and 82-89, 11. 766-838, Bohun and de Lucy in Northumberland. The metres are uniform : decasyllabic laisses for the first and tripartite Jordanian alexandrine laisses for the second ; the first is commonplace, the second may be better appreciated if details of the positioning of m and f are given. 82 m mm ff, 83 fmfm ffff ffff, 84 mmm fff fmf, 85 mfmf f mfmf, 86 fm mff mff, 87 mfff f fffm, 88 mmf ffmf ffm, and 89 mf mm mf. In 87 reversal is used, in 88 opposition of 2m 1f to 2f lm ; both principles are used elsewhere. In the four sections where there are very evident subsections of matter, the flow of Jordanian laisses is interrupted at the beginning and/or the end of each subsection as can be illustrated from the first section 2 to 21, 11. 21-240, which deals with the outbreak of warfare on the Continent : 2 J.S. 3-9 J.L. Henry's enemies unite and there is fighting in Normandy ; LO J.L., 11 J.S. Fantosme reflects ; 12 J.S., 13-20 J.L., 21 alex. Revolt crushed in Brittany. Then follows the second section, 22-45, 11. 241-467, which lists diplomatie moves, envoys' journeys, and debates.

31Part II is differently orchestrated. 110-156, 11. 1102-1528, show how William is lured on by fire.-flies of hope like the successes of his brother David in the Midlands, the adhesion of Roger Mowbray, the surrender of Appleby, and the reduction of Brough against the grimmer warnings of failure to capture Wark, and to obtain the surrender of Carlisle, to risk all in Northumberland. These shifting scenes are narrated as before in J.L. separated off into sections and subsections by J.S. and alex. The tide turns in 157-174, 11. 1529-1633, the set-piece of purposefully opposed alex. and J.S. the report of the bishop of Winchester to Henry II. After this, 175-211, 11. 1634-2028, the road winds downhill all the way for William. His Scots do awful damage, but he is defeated and captured near Alnwick, the Barons' Revolt collapses, Henry returns to England, the English triumph. This long sweep of events is told in solid alexandrine laisses, with two vital moments picked out by descorts: 182, 11. 1701-14, a Jordanian stanza, highlights the explanation and the cause (in Fantosme's theological view of history) of William's downfall, namely the sacrilege of the Scots, and 211, 11. 2013-28, a Jordanian laisse, shows us Henry hearing the news of his victory. Descorts have also been used for the introduction of David, the rallying of Mowbray to William's cause, and the move to Carlisle under Mowbray's auspices.

32My remaining space must be devoted to a ridiculously exiguous mention of the set-pieces. As well as the bishop and king interview this technique of metrically relating laisses and stanzas to one another is used for the hostilities in Northumberland, 115-123, 11. 1137-1206, and the resistance of Carlisle contrasted with the submission of Appleby, 140-151, 11. 1360-1468. In the Northumberland episode four sestets of J.L. are distinguished from five J.S. When they are set together it is seen that the single m that occurs in the first four lines of each sestet shifts its position regularly from first to fourth line.

33There is also a patterned relationship between m and f and syllable count in the final lines of all the sestets.

34The matter and the manner of the telling have made many people regard the Chronicle with affection. What has been lacking has been a clue to the technical abilities of Jordan Fantosme. Without it there is a colossal dissonance between the education and training of the man and the impression created by the totally inadequate manuscript presentation of his work resembling the product of a careless and disordered mind. Examination hemistich by hemistich and syllable by syllable really does clear the air. It shows in detail what the problems are,, and, what is equally important, where there are no problems. In his alexandrines he is often more continentally regular than his near Anqlo-Norman contemporaries, such as the author of the Roman de tute chevalerie. In his Jordanian alex alexandrines, he has systematised the irregularities of insular practices. In the stanzas he has transformed irregularity into patterned order by the positioning of impeccably counted hemistichs. In two more departments Fantosme is more regular and more sophisticated than his contemporaries: he is capable of patterning caesural types, and this I have not seen in any other A.-N. poet; where all O.Fr. and A.-N. poets write successive rhymed laisses and couplets that assonate together by mere inadvertence and lack of strict policy (see Johnston 1976), Fantosme utilises assonating rhymes to link up laisses and stanzas into deliberately shaped patterns.

35The answer to the charges of 'irregularity' and even innumeracy often laid at his door in the nineteenth and twentieth centuries must be a firm rebuttal. His reputation as a brilliant teller of tales is well established; his place as an accomplished prosodist is even more eminent.

LITERATURE AND ABBREVIATIONS

36(No references are given for works in the SATF and CFMA series.) Abelard, P. Dolorum solacium in Gaselee, S. 1928. The Oxford Book of Medieval Latin Verse. Oxford.

37Foster, B. (ed.) Le Roman de tute chevalerie. ANTS, 29-33. London. Harden, A.R. (ed.) Matthew Paris, La Vie de Seint Auban. ANTS, 19. Oxford.

38Howlett, R. (ed.) 1886. Chronicles of the Reigns of Stephen, Henry II, and Richard I ; Rolls Series, 82. III. London.

39Johnston, R.C. 1974. The Versification of Jordan Fantosme. Oxford.

40Johnston, R.C. 1976. Sound-related couplets in Old French in Forum for Modern Language Studies, 12, no.3.

41Legge, M.D. 1971, Anglo-Norman Literature... (Corrected reprint of the first edition, 1963). Oxford.

42Macaulay, G.C. (ed.) 1899. John Gower, The French Works. Oxford.

43Macdonald, I. 1961. The Chronicle of Jordan Fantosme in Studies in Medieval French presented to Alfred Ewert. Oxford (which see for bibliographical references to critics.)

44Pope, M.K. and Reid, T.B.W. (ed.) 1964. The Romance of Horn. ANTS, 12-13. Oxford.

45Stengel, E. (ed.) 1886. L'Afaitement Caton .... Ausgaben und Abhandlungen, 47. Marburg.

46Verrier, P. 1932. Le versfrançais. Paris

________

47alex. alexandrine. A.-N. Anglo-Norman. ANTS A.-N. Text Society. f feminine hemistich(s). m masculine hemistich(s). J.L. Jordanian (alex.) laisse. J.S. Jordanian stanza. O. Fr ; Old French. O. Pr ; Old Provençal.

Notes

1 The m resulting from elision is designated mf.

Table des illustrations

URL http://books.openedition.org/pup/docannexe/image/3720/img-1.jpg
Fichier image/jpeg, 18k

Le texte et les autres éléments (illustrations, fichiers annexes importés) sont sous Licence OpenEdition Books, sauf mention contraire.

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search