Estudi introductori - Einleitende Studie - Introductory study - Estudio introductorio - Introduction - Studio introduttivo - Estudi introductòri

Lewent, Kurt. An Old Provençal Chanson de mal mariée . "Romanic Review", 37.1 (1946), pp. 3-19


The poem here studied is a composition of the Catalan troubadour Cerveri de Girona, published by F. A. Ugolini in his diplomatic edition of Il Canzoniere inedito de Cerverì di Girona, Rome, 1936 (Memorie della R. Accademia Nazionale dei Lincei, Classe di Scienze Morali, Storiche e Filologiche, Anno 333, Serie VI, volume 5, 513-683), Number 102, page 672. The MS, known as Sº, now belonging to the “Biblioteca de Catalunya” (Number 146) in Barcelona, calls the poem a gelosesca. The name is derived from gelos and formed after the model of sirventesca, (1) originally the feminine of an adjective in -esc (cf. sirventes joglaresc) with a noun, such as canso, to be supplied. In MS Sº, almost all of Cerveri’s poems are provided with titles, some of them occurring only there. These titles do not always designate a literary genre, such as canso, sirventes, vers, tenso, alba etc.; many of them are of individual coinage and meant to illustrate the contents of the poem. (2) The question is to which of the two categories of titles gelosesca belongs.

There are no other known poems bearing the same name. Theoretical works dealing with Provençal poetics, it is true, mention the gelozesca. The Leys d’Amors (ed. Anglade, Bibl. méridionale, Série 1, Number 17-20; volume II, 185) name it among the dictats no principals (the less important genres), giving this additional remark: Alcu fan gilozescas al compas (i.e. “after the metrical scheme”) de dansa o de chanso. Of the dansa they say: ... dansa no ha mays tres cobblas estiers lo respos e la tornada. The Doctrina de compondre dictats (ed. Paul Meyer, Romania, VI, 353 SS) (3) gives a somewhat more detailed definition (page 357, number 14): Si vols far gelozesca, deus parlar de gelozia reprenden e contrastan de fayt d’amor; e deu haver responedor, e quatre cobles, e una o dues tornades, e so noveyll o estrayn ya feyt, (4) and later on: Gelozesca es dita per ço car gelozamen parla de ço que dir vol, contrasta[n] ab alguna persona en son cantar(page 358, Number 30). The second explanation of the Doctrina, not quite clear in all its wording, seems to add hardly anything new to the first. The definitions of the Leys and the Doctrina, however, are contradictory in one respect. The former gives the gelozesca, like the dansa, three stanzas, the latter gives it four. Setting aside the question of the number of stanzas we could say that those definitions, and especially that of the Leys, are in full agreement with Cerveri’s poem, and we venture to claim that the two treatises on poetics did not know more specimens of the gelozesca than we do, their instructions having thus been made according to Cerveri’s poem. This would imply that the word gelozesca does not designate a literary genre, but is one of those individual titles that MS Sº uses to characterize the contents of a song. However this may be, our gelozesca deals with the gilos (5) (the jealous husband), traditional figure of the chanson de mal mariée, of which it is a typical representative. This literary genre has been made the topic of a monograph by Rudolf Daehne: Die Lieder der Maumariée seit dem Mittelalter, Halle, 1933 (Romanistische Arbeiten, ed. by Karl Voretzsch, Number XX). Cerveri’s song, indeed, shows all the characteristics of that kind of poem: 1. the husband is old, ugly, and of disgusting manners; 2. he is of bad character: false, malicious, jealous; (6) 3. he is a vilain, (7) interested in nothing but his farming; 4. The wife complains bitterly of being married to such a man; 5. she blames her friends for having given her to that abominable man; 6. she curses her husband for depriving her of the pleasure of being with her lover, wishing the former all kinds of disaster, even death; 7. her lover will help her.

However, besides these features there are others in our poem which, up to now, have either not been found in so early a period (a) or have no parallel at all in songs of this kind (b). With reference to (a): 1. the wife’s mother sympathizes with her unhappy daughter; she even tries to help her; (8) 2. the fact that the husband, instead of making love to his wife, turns his back to her, sleeping and snoring, is a trait that Daehne (pages 49-51) reports as not appearing in French songs before the fifteenth century and as becoming of greater importance in the sixteenth and seventeenth centuries only. With reference to (b): I. in Cerveri’s poem the woman does not only wish her husband to die, (9) but she makes preparations to kill him herself; 2. she means to achieve this by employing magic and witchcraft.

Originality of ideas was generally not the aim of medieval poets. Nevertheless, the motifs named under (b), being without parallels in the chanson de mal mariée, may indeed constitute Cerveri’s own contribution to the genre. Particularly the means by which the woman is going to rid herself of her husband seems to follow a line of thought which apparently was quite familiar to our poet. He indeed makes not infrequent allusions to matters of popular belief. He knows the fabulous country of Prester John (Gr. 434a, 82), he mentions several facts of legendary zoology (Gr. 435a, 31), he refers to the story of the old servant eating a snake (Gr. 434, 15), he names the legend of Saint-Genis who was once a jongleur and the candle of Arras (Gr. 434, 57), he makes Saint-Mark and a miraculous power of his the basis of a humorous song (10) (Gr. 434a, 72), and he seriously warns us of old women, whose aspect may deprive us of the faculties of seeing, hearing and speaking (Gr. 435a, 56). If Cerveri really invented the above-named motifs, which I think extremely likely, he was well inspired, because they excellently fit into the framework of the so-called popular poetry to which the chanson de mal mariée belongs.

The characteristics of the first category (a) are, in all probability, less original. Since they are found in specimens of the genre posterior to Cerveri’s chanson de mal mariée, he may well have picked up those features in songs that existed at his time and were known to him, but that we have no knowledge of because they have been lost since. The question is whether Cerveri underwent the influence of Old Provençal or Old French songs of this kind. The influence of Provençal poetry would not be surprising in an author who composed about a hundred poems in that language. But this assumption is opposed by the fact that songs of the unhappily married woman are extremely rare in southern France. They are even less numerous than Daehne (pages 10 ss..) makes us be-lieve. The poem of Jausbert de Puycibot (Gr. 173, 9 ed. Shepard, page 28) does not belong here, though the four lines quoted by Daehne (page 6) speak of the brutality of his beloved lady’s husband; it is a regular canso. Neither are the 12 lines of the Kalenda Maya inserted in, and adapted to the metre of, Flamenca really a chanson de mal mariée. Sung by the girls of the village, they praise in general terms illicit love in spite of the gilos, but do not contain a word of complaint on the part of an unhappy wife. The same is true for the two anonymous songs A l’entrade del tens clar (Gr. 461, 12) and Li jalous partout ount fustat (Gr. 461, 148-a). The poem D’una don’ai auzit (Gr. 234, 8), attributed to Guillem de Saint-Leidier by MS C, to Peire Duran by MS R, (11) cannot be considered as a literary product at all, being nothing but an imaginary dispute of utter obscenity between a man and his wife about one single detail of their married life. Can se reconian auzeus (Mahn, Ged. 728) is a French song, superficially Provençalized, by Thibaut de Blizon. As the last of “genuine” chansons de mal mariée Daehne names a song beginning S’anc fui belha ni prezada. He does not seem to have ever seen more of this poem than the first line just quoted. All his knowledge about it comes from Gaspary, Geschichte der italiaenischen Literatur, I, 101. Indeed, Gaspary gives the line in question —and nothing else— saying that it is the beginning of a Provençal song in which the lady complains about the vilan to whom she has been married because of his riches, but finds consolation in her lover and the faithful watchman who protects their love from any danger. What Gaspary does not say and Daehne consequently does not know is that the summary given by him is not that of a whole poem, but of only one stanza (12) out of five of an alba composed by Cadenet (Gr. 106, 14) and published and discussed by Appel in his edition of that poet (Halle, 1920, pages 76-83). That this stanza brings some motifs characteristic of a chanson de mal mariée there is no doubt, and Appel does not fail to call our attention to this fact. The fifth stanza, too, bears some similarity with this kind of song. (13) These facts led Appel to believe that the poet tried to create something new by fusing into one the two genres of the alba and the chanson de mal mariée. However that may be, there is no denying that, providing his poem with the traditional refrain of an alba, Cadenet had meant it to be an alba.

From this critical review of the songs claimed by Daehne as Provençal chansons de mal mariée it may be concluded that, besides Cerveri’s gelozesca, there are only two real specimens of that literary genre in Old Provençal: (14) Coindeta sui, si com n’ai greu cossire (Gr. 461, 69) and Quan lo gilos er fora (Gr. 461, 201), both of them published by Appel in his Prov. Chrestomathie (Numbers 47 and 45). Jeanroy does not know of any others either; cf. Les Origines de la Poésie lyrique en France au Moyen Âge, 2nd ed., Paris, 1904, pages 86-87 and Poésie lyrique des Troubadours, Paris, 1934, I, 303. Each of these poems is known only from one single MS (Q and X), and it may therefore appear doubtful whether Cerveri had known them. On the other hand, seeing in Cerveri’s gelozesca an imitation of French songs of this type presents another difficulty.According to Am. Pagès, La Poésie française en Catalogne du 13e siècle à la fin du 15e, Toulouse, 1936 (Bibl. meridionale, I, 23), literary relations between Northern France and Catalonia were not very close until the fourteenth century. He names, it is true, Cerveri as one of the earliest links between the two countries. The principal proof he gives for his claim is the fact that Cerveri makes use of the Alexandrine verse (page7). This proof, however, is not convincing. The Alexandrine metre is not unknown to Provençal literature. The best-known epic specimen of this kind of verse is the Albigensian Chronicle. (15) It occurs in lyric poetry too. Maus, Peire Cardenals Strophenbau, Marburg, 1884, page 79, names fourteen poems in Alexandrine verse (Sordel, Bertran d’Alamanon, Peire Bremon, Dalfin d’Alvernha, etc.). All of them have in common the fact that they are composed in monorhyme stanzas, which seems to point to an epic origin of this metre. Now, the two poems of Cerveri that are written in the Alexandrine metre follow the same rule. (16) This conformity of Cerveri’s with Provençal usage is strong evidence against Pagès’ opinion. Another argument advanced by Pagès is this one. Among the poems written by Ramon Lull in Alexandrine verse there is one called the Desconort which is to be sung in el so Berart. This is Berart de Monleydier, the hero of a lost epic, who is also mentioned by Cerveri in his didactic poem Ben dit, mal dit (written in lines of 6 syllables!) together with David, Plato, Roland, Perceval, and others (see Suchier, Denkmaeler der prov. Lit. und Sprache, Halle, 1883, page 269, l. 500). For Pagès it is sure that the lost chanson de geste was French and that Cerveri must have known it. But the fact that Cerveri mentions Berart does not mean anything because allusions to Berart are found several times in troubadour poetry, the first of them as early as Marcabrun, and there is no proof that the lost epic was written in French. Birch-Hirschfeld, Ueber die den prov. Troubadours . . . bekannten epischen Stoffe, Halle, 1878, page 70, thinks it, on the contrary, quite possible that it was a Provençal poem. Pagès might, however, have supported his point by emphasizing two other facts. 1. In one of his pastorelas (Gr. 437, 7c, ll. 10-11, ed. Kleinert, Vier bisher ungedruckte Pastorelen des Troubadours Serveri von Gerona, Halle, 1890, page 20) Cerveri says:

C’a mos oyls tan plazen no vis,
nems en França n’en Castela.

Taken literally these lines seem to prove that Cerveri had been in France and in Castile. But the question is whether the word França is really to cover only the northern part of Gaul and, if this is the case, whether we do not face here one of those stereotyped hyperbolical expressions so common in troubadour lyrics, nothing being known of Cerveri’s stay in Castile either. 2. Cerveri composed two lines in French forming part of a cobla en. VI. lengatges (17) (Gr. 434a, 40) which run thus:

E voldroye touz les (18) jorz de ma vie
dames trover o pris de tota (sic) jan.

Here again we dare say that Cerveri followed Provençal rather than French tradition. When he composed that poor stanza he is not unlikely to have had in mind the famous five-language descort by Raïmbaut de Vaqueiras (Gr. 392, 4 ed. Appel, Prov. Chrestom., Number 37).

There is another reason to presume a closer connection between Cerveri’s gelozesca and Provençal poems of this kind. Like the two Provençal representatives mentioned above, Cerveri’s chanson de mal mariée has the form of a dansa, i. e. a dancing song with a refrain put before the first stanza and supposed to be repeated after each stanza by a choral group, while the stanzas themselves were probably recited or sung by a soloist. Now, Pagès (l. c. 135), pointing out the unpopularity in Catalonia of the French rondeau and ballade, holds (without giving reasons for his opinion) that this was due to the fact que la dansa provençale s’y était implantée depuis Cerveri de Girone et qu’aucune des autres chansons à danser n’a pu, durant plusieurs siècles, la détrôner. We have already indicated that we know only very few specimens of such songs, especially the chansons de mal mariée. But Cerveri had been in southern France, at the court of the counts of Foix (see Massó Torrents, l. c. p. 182). There he may have heard not only the two Provençal chansons de mal mariée of which we have knowledge, but also other poems of the same kind which have not come down to us. The fact that Provençal troubadours, such as Jausbert de Puycibot and Cadenet, used motifs characteristic of the chanson de mal mariée make us believe that this genre was far more popular in southern France than its few written specimens indicate. Most of them probably were refused the honor of being written down together with the representatives of “nobler” genres.

So the odds are definitely in favor of Cerveri’s having undergone the influence of Provençal rather than French models. The metrical structure he gives his poem seems to confirm this assumption. The stanzas of the two other Provençal chansons de mal mariée are in a very simple metrical form, probably much too simple for a man like Cerveri who loved to display his versifying skill. He is, however, not likely to have invented the metrical scheme of his gelozesca himself. There is one Provençal poem whose stanza has exactly the same form, though its rhymes are different from those of Cerveri’s song, (19) i. e. 10a’ 10b 10a’ 10b 10a’ 10b 10a’ (Maus, l. c. No. 210). It is a poem by Montan (Gr. 306, 2). This author was a contemporary of Sordel, with whom he exchanged a cobla, and therefore prior to Cerveri. His “poem” is of incredible obscenity. That does not mean that Cerveri was not allowed to borrow its metre, which is the more likely because his own poem does not quite conform to courtly usage either. Moreover, Montan himself is likely to have had a model for his “poem,” which is a tenso, though a fictitious one, and therefore not subject to the rule of the canso, which had to be original in its metrical structure. This model, it is true, has not come down to us but may have existed (20) and been imitated for a second time by Cerveri. This presumably lost original is supposed to have had the same rhymes as our poem.

There is not much to be said about the versification itself. The decasyllabic lines, in our opinion a little too heavy and solemn for the tenor of the poem, are mostly built in strict observation of the metrical rules, i.e. with a caesura after the fourth syllable, which, when unstressed, forms what has been called a lyric caesura (ll. 8, 10, 13, 15). There are, however, two lines which seem to deviate from those rules, showing the structure 6+4 instead of 4+6. They are l. 3 and l. 23. In l. 3 (E·m dè mal pus lo pris. Car tan me dura?), the real break is after the sixth syllable. There is a pause after mal, too; but this word being the third syllable, the caesura cannot be placed after it. The case of l. 23 (Qui·l ve, si·s vol tardar de mal destrich) is less obvious. The sense admits two natural stops: after ve and after tardar. No caesura is possible after ve, this word forming the second syllable. The regular caesura should be after vol. The question is whether or not vol can be separated from the infinitive tardar depending on it. At any rate, the pause after tardar (6th syllable) seems to be weightier than one after vol. As Cerveri makes use of a decasyllable type 6 +4 in l. 3, I see no reason why he should not have done the same thing in l. 23. After all, he was not the only poet to insert decasyllables of that type among others of regular structure. (21) I gave quite a few examples in Studi Medievali, n.s. IX, 144-145.

Exceptional as these cases are compared with the mass of regular decasyllables, they no doubt did exist. They may have been used with the idea of giving the flow of equally built lines an occasional and welcome change of rhythm.



1) Another of Cerveri’s poems is entitled pegesca, and it is the author himself that gives it this name in the tornada (Ugolini, No. 101, p. 671). I suppose it is a humoristic derivation from pec (pega) “fool.” ()

2) Cf. Storost, Litbl. f. germ. u. rom. Phil., 60 (1939), 519.()

3) The date of this work is not to be deduced with certainty. In the only MS that contains it, the Doctrina follows the Razos de trobar by Raimon Vidal de Besaudun. If Raimon is the author of the Doctrina, too, which P. Meyer does not think impossible, this work would belong to the first half of the 13th century, and its author could then not have known Cerveri’s gelozesca. But Raimon’s authorship is anything but sure. Meyer himself sees a difference of style in the two works, and Milá thinks the Doctrina belong to the end of the 13th century. It may even be still more recent.()

4) The gelozesca may have a melody (form) of its own or be composed after the model of another song.()

5) After all, any poem doing this, especially the poems discussed below, might have been called gelozesca, but Cerveri’s song is the only one that actually did get that name.()

6) The reproach of jealousy is, of course, inherent in the genre, though in our poem the man does not seem to do anything that might prove him to be jealous. On the contrary, his behavior described in ll. 12-15 and l. 24, rather indicates that he does not care for his wife in the least.()

7) Cerveri does not use this expression, but the man is one (cf. ll. 23-24).()

8) In our song she does it by a sort of incantation. In a modern Provençal song, the mother stays still more in the background; she advises her daughter to do the same thing that she had done herself when she was young: Fai lou couioul, ma filho, que tun paire ja n’es (Daehne, p. 145).()

9) Cf. the following line from one of the two Provençal chansons de mal mariée till now the only one extant: Qu’eu prec la mort qe·l venga tost aucire Gr. 461, 69 (Appel, Prov. Chr., No. 47) II, 4. A little more audacious is a woman in a French song of the 18th century from which Daehne quotes these lines:

Je voudrais qu’il vienne un édit
D’écorcher tous les vieux maris.
J’écorcherais le mien aussi.

But her activity is only hypothetical.()

10) See my edition (Une chanson humoristique de Cerveri de Girona) in Ann. du Midi, 51 (1939), pp. 285-294.()

11) See my article Abseits vom hohen Minnesang in Studi Medievali, n.s. IX, 131.()

12) This is the stanza in question:

S’anc fui belha ni prezada
ar sui d’aut en bas tornada,
qu’a un vilan sui donada
tot per sa gran manentia;
     e murria,
s’ieu fin amic non avia
cuy disses mo marrimen,
     e guaita plazen
     que mi fes son d’alba.

Some MSS do not have this stanza, beginning with Eu sui tan corteza gaita, and it is with this line that the song is listed in Bartsch’s Grundriss and Pillet-Carsten’s Bibliographie. This is probably the reason why Daehne could not identify it.()

13) The fifth stanza runs thus:

Ja per guap ni per menassa
que mos mals maritz me fassa,
no mudarai qu’ieu no jassa
ab mon amic tro al dia,
     quar seria
desconoissens vilania,
qui partia malamen
     son amic valen
     de si, tro en l’alba.()

14) There is another poem of Cerveri’s (Gr. 434a, 34; Ugolino No. 99, p. 669), called viadeyra, which is no chanson de mal mariée, though it shows some features of the latter. This is the text:

No·i prenatz, lo falls marit, ia, na Delgada,
          no·l prenatz, lo fals-jurat,
3                    que pec es, mal enseynat,
          no·l prenatz, lo mal marit,
6                    que pec es ez adormit,
          que pec es, mal enseynat;
9                    no sia per vos amat,
          que pec es ez adormit,
12                  no jaga ab vos el lit,
          no sia per vos amat.
   15         Mel[h] val cel c’avetz privat,
          no jaga ab vos el lit,
18         mais vos y valra l’amich,

4. ya]y   12. ya]ia.   15. mel.   18. Mes vol.()

15. The Thezaur by Peire de Corbian (ed. Jeanroy-Bertoni, Ann. du Midi, 23, 289 ss., 451 ss.) is also written in this metre (520 lines all of them ending in -ens.) A different kind of dodecasyllable is used in the metrical version of the Practica Chirurgica by Roger of Parma; the lines have a caesura after the fourth or the eighth syllable (see Stimming in Groeber’s Grundriss, II. 2, 42-43).()

16) There are indeed rare examples in Cerveri’s lyric poems of Alexandrine verses alternating with those of different length. Gr. 434a, 20 is f.e. composed of stanzas comprising 4 decasyllabic and 2 Alexandrine verses. But this, of course, is no proof that Cerveri must have borrowed this metre from French poetry. Even if Cerveri had not had the abovenamed Provençal models, an author such as he, who, concerning lengths of lines, tried almost everything from monosyllabic to endecasyllabic verse, might well have hit upon the dodecasyllable by himself. In this he could have been guided by stanzas like those of Gr. 434a, 15. Here, the sixth syllables of all the lines are bound together by an internal rhyme, which allows the stanza to be considered as consisting of 10 lines of 6 syllables each instead of 5 Alexandrine verses with internal rhymes.()

17) Ugolini No. 93, previously edited by Massó Torrents (l. c. 230), who identifies them as Castilian, Provençal, French, Gascon, Sicilian (?), Catalan.()

18) Ugolini has le, Massó Torrents les.()

19) So we have no definite proof that Cerveri imitated the form of a Provençal song with his gelozesca. He did do it with the following of his poems in which there is identity of rhymes: 1. Gr. 434 a, 23 = Giraut de Bornelh Gr. 242, 17; 2. Gr. 434 a, 28 = Raimon Jordan Gr. 404, 11 (cf. Kjellman’s edition p. 121); 3. Gr. 434 a, 36 = Raïmbaut de Vaqueiras Gr. 393, 23; 4. Gr. 434 a, 48 =Uc Brunet Gr. 450, 3; 5. Gr. 434 a, 75 = Bern. de Rovenac Gr. 66,4; 6. Gr. 434 a, 81 = Arnaut de Maruelh Gr, 30, 16.()

20) Montan’s “poem” is a travesty of courtly love. The existence of a canso being the metrical model of this tenso would greatly add to its cynic nature and to the pleasure of certain hearers recognizing the “noble” model of this ignoble product; cf. Vossler, Rom. Forsch., 51, 266 and my remark Neuph. Mitteil., 34, 241.()

21) It is well known that the epic of Girart de Rossilho exclusively uses decasyllables of the type 6+4.()








Institut d'Estudis Catalans. Carrer del Carme 47. 08001 Barcelona.
Telèfon +34 932 701 620. Fax +34 932 701 180. - Informació legal