‘Mais où va donc mon sonnet ?’: Verlaine and the Fixed Form of Petrarch

By Alan English

Paul Verlaine was one of the most prolific sonneteers of the nineteenth century, producing over two hundred and forty poems based on or adopting a version of this fixed form. ‘Monsieur Prudhomme’[1], his first published poem (1863), is a sonnet, and, in a neat symmetry, so too is ‘Désappointement’[2] his last poem, which is dated 31 December 1895, just a little more than a week before his death. In the intervening decades, the large majority of Verlaine’s recueils – albeit with some significant exceptions – include sonnets, and attest to the poet’s longstanding, though ambivalent and, at times, curious, relationship with this form. This volume of sonnet production far outstrips that of Charles Baudelaire, Arthur Rimbaud or Stéphane Mallarmé (even combined), and yet, very few of Verlaine’s sonnets are as successful and as influential as many of those of his contemporaries. Furthermore, as their near-complete absence[3] from Fêtes galantes (1868), La Bonne Chanson (1870) and Romances sans paroles (1874) demonstrates, it is often precisely at those moments when Verlaine’s poetic expression is at its most original that he seems farthest away from the sonnet, both expressively and creatively.

It is not therefore surprising that many studies of the modern French sonnet relegate Verlaine’s contribution to the margins, highlighting particular successes, but lamenting the poet’s descent into the fundamentally ‘anti-sonnet’ type of prosaism characteristic of, for example, many of the ninety-six occasional or circumstantial sonnets of Dédicaces (1894), the reading of which is never likely to be among the most satisfying or interesting experiences for French poetry enthusiasts. As Jacques Borel remarks most diplomatically in his introductory notes to the collection, ‘Verlaine semble avoir définitivement perdu [ici] le secret de son propre chant.’[4]

However, it is important not to completely dismiss Verlaine’s poorest sonnets: to restrict consideration to the few successes would not only distort his contribution, but would misrepresent and disguise what was, in terms of his complete works, an unsteady, changeable and clearly ambivalent relationship between poet and fixed form. As David Scott sees it, there is an intrinsic mismatch between Verlaine and the sonnet, in that the poet was no more ‘temperamentally suited’ to this fixed form than its ‘rigid and complex structure’ was ever likely to provide the best framework for composing the type of atmospheric, mood and visceral poetry with which Verlaine is most commonly identified.[5] This is undoubtedly true, and yet, not least because of this mismatch and the ironic attitude it engenders in the poet, the sonnet would also provide an ideal space in and through which the poet could extend and develop, at the level of the fixed form, the sort of metric and prosodic experiments that are such a recurring and characteristic feature of his use of the French verse-line, particularly the Alexandrine. Further, it can be argued that the sonnet was in fact a necessary and unavoidable choice for Verlaine, not just because it represents tradition and convention, but because it embodies the very concepts of metricity and poeticity, playing a similar role in this respect to the Alexandrine, which Benoît de Cornulier identifies as one of those ‘modèles culturels ou stéréotypes immédiatement reconnaissables hors contexte.’[6]

In attempting to explore these ideas, it is useful to start by briefly considering the forms and characteristics of the modern sonnet, for one thing becomes very clear: despite its status as a ‘fixed form’, and its reputation as a rigorous mathematical model – Baudelaire refers to its ‘beauté pythagorique’[7] – the sonnet reveals itself to be remarkably ‘unfixed’ and, one could even venture to say, ‘free’. To begin with, there is more than one standard type: the Franco-Italian or continental sonnet, itself a fusion of two traditions, which has two formal units – an octave, based on two identical quatrains with embraced rhymes (abba abba), and a sestet introducing either two or three new rhymes sounds, with tradition favouring ccd ede and ccd eed as the regular forms; and there is the Elizabethan or Shakespearean sonnet, which has three quatrains based on different crossed rhymes and a final rhyming couplet, abab cdcd efef gg. Furthermore, by the time Verlaine wrote his first sonnet, many variations in the standard forms were also commonly used: so, it is possible, for example, for the octave’s rhymes to be alternating, not embraced (abab abab or abab baba), to be based on four sounds, not two (abba cddc or abab cdcd), or to follow different arrangements in the two quatrains (abba abab, abab cddc). Greater variety still is to be found in the rhyming scheme of the sestet: (cdc dcd), (ccd ccd), (ccd, eed), (cdd, cee) and others all appear, for example, in Les Fleurs du mal (1857).

Therefore, by the second half of the nineteenth century, the sonnet was structurally more flexible and variable than perhaps at any other time in its history. This variability can be linked to its multiple origins and the cross influences of its many traditions, but as Clive Scott argues, it is important to identify in it too a fundamental and defining characteristic of this ‘fixed’ form:

[The sonnet’s] tenacious persistence through the centuries – with the notable exception of the eighteenth – has meant that variations on the [standard] patterns […], which have been manifold, have not so much been deviations from the sonnet as extensions of the sonnet’s freedom, a freedom whose only qualifying limit is ultimate recognisability as a sonnet.[8]

One essential element of the (continental) sonnet’s ‘recognisability’ or ‘sonnetedness’ is its asymmetric, dialectical, eight/six line-arrangement – an embedded reminder of its origins, in the early thirteenth century, as the combination of two pre-existing and separate poetic forms – the eight-line strambotto, one of the oldest types of Italian verse, and the Sicilian six-line sestina. The ‘soldering’ point or volta between the two sets of stanzas remains, in the modern forms of the sonnet, a space invested with significant poetic and metric potential, a kind of inter-stanzaic caesura or fulcrum around which several dialectical stanzaic relationships can be created. It is also the ‘face-off’ point between two clearly opposing prosodic worlds – the structurally complete, self-sufficient and unrushed quatrains, on the one hand, and the individually incomplete, dependent and more urgent tercets, on the other. Hence, each part of the sonnet has tended to take on different, and even opposing, functions, with the quatrains often presenting and describing, and the tercets revealing and resolving, as Théodore de Banville describes in his Petit Traité de poésie française:

Ce qu’il y a de vraiment surprenant dans le Sonnet, c’est que le même travail doit être fait deux fois, d’abord dans les quatrains, ensuite dans les tercets, – et que les tercets ne doivent pas répéter les quatrains, mais les éclairer, comme une herse qu’on allume montre dans un décor de théâtre un effet qu’on n’y avait pas vu auparavant.[9]

These structural differences – and the asymmetries and potential tensions they create – are defining characteristics of the sonnet, which, despite what the term ‘fixed form’ might evoke or imply in terms of rigidity, assuredness and structural certainty, is in fact a fundamentally ambivalent form – ‘magnifique, prodigieusement belle, – et cependant infirme en quelque sorte,’[10] as Banville puts it. The challenge for the sonneteer is to balance the structural and formal tensions, and semantic expectations, created by the interplay of the sonnet’s component parts, to decide how and whether to foreground or attempt to disguise its asymmetrical and dialectical potential, and, critically, to justify and motivate any deviations from its conventional forms. The result should, as Clive Scott says, provide a sense of the necessity or appropriateness of the choices made:

The reading of a sonnet is an experience fraught with uncertainty […]. The sonneteer must capitalise upon this anxiety, but at the same time must create in the reader, as the poem comes to a close, a strong sense of the inescapability of the chosen scheme, so that the most eccentric variation will have the feel of a fixed form. The sonneteer’s art might be described as a fourteen-line postponement of the reconciliation between projective unpredictability and retrospective fittingness.[11]

Measuring and comparing Verlaine’s sonnets against these characteristics reveals, perhaps surprisingly, that he is in many respects a traditionalist. For example, while up to eighty different rhyme scheme formats appear across his more than two hundred and forty sonnets, very nearly half of all these poems present just five patterns which include the two conventional formats abba abba ccd ede (thirty-six occurrences) and abba abba ccd eed (twenty-nine occurrences), as well as two which had already been used by Baudelaire and Musset: abba abba cdd cee (twenty occurrences) and abab abab ccd ede (eighteen occurrences), and abba abba cdc dee (thirteen occurrences), which also appears in Les Fleurs du mal. The most questionable and most ‘unsonnet-like’ arrangements are the rarest, making just ‘guest’ appearances once or twice: abbb accc dde ffe (in ‘Cordialités III’[12]), abba ccca ddd eee (in ‘Gabriel de Yturri’[13]) and abab cdcd eff ghc, in the appropriately named tredecasyllabic ‘Sonnet boiteux’[14] whose final tercet’s particularly chaotic scheme seems to give prosodic expression to the unhappy and turbulent time the poet is remembering having spent in London with Rimbaud, towards the end of 1872.

Verlaine is conservative too in terms of varying the sonnet’s presentation. There are just ten inverted sonnets, which all display sonnet-like rhyme schemes, for example: abb acc deed deed in ‘Résignation’[15] (dating from 1866, the same year as the first Le Parnasse comtemporain which includes Baudelaire’s only inverted sonnet, ‘Bien loin d’ici’, first published in 1864 in La Revue Nouvelle); and aab ccb deed effe, in the more playful and light ‘A Louis et Jean Jullien’[16], which interestingly mentions the ‘sonnet les jambes en l’air’ and whose first quatrain’s burlesque rhymes contrast obviously with the lazy and uninspiring rhymes in the other stanzas:

Savantissimo Doctori
Bonnissimoque Scriptori,
Au frère et puis encore au frère
Ce sonnet les jambes en l’air
Qui commence à chanter son air
En pur latin de feu Molière !

Ce sonnet pour dire à tous deux
Sur un ton badin mais sincère
Que je les aime bien et serre
Leurs loyales mains à tous deux.

Louis, malgré le sort contraire,
Salut à vous qui guérissez,
A vous aussi qui punissez
L’ordre bourgeois, Jean, mon confrère.

Only two other sonnets offer presentational variations: ‘A Félicien Champsaur’[17], whose tercets are presented typographically as a sestet cddcee, and ‘L’Allée’[18], the disguised sonnet of Fêtes galantes, whose fourteen lines are not separated into visual quatrains and tercets, but are nonetheless arranged in dialectical sestet/octave form: ababaa cddcefef.

Other licences which Verlaine takes with the sonnet – such as dispensing with the principle of alternating masculine and feminine rhymes and using assonance in place of rhyme – are often just experiments and do not mark new departures in his conception of the fixed form.[19] However, there are some processes and features characteristic of many of Verlaine’s sonnets whose effects are quite subversive and destabilizing, which I mention here briefly.

First, Verlaine often seeks to undermine the volta, one of the sonnet’s distinguishing characteristics, and comparable in prosodic and metric importance within the fixed form to the Alexandrine’s caesura. A particularly obvious example is provided by the dramatic inter-contamination of second quatrain and first tercet in ‘Une grande dame’:

Ses yeux froids où l’émail sertit le bleu de Prusse
Ont l’éclat insolent et sur du diamant.
Pour la splendeur du sein, pour le rayonnement
De la peau, nulle reine ou courtisane, fût-ce

Cléopâtre la lynce ou la chatte Ninon,
N’égale sa beauté patricienne, non !
Vois, ô bon Buridan : « C’est une grande dame ! ». [20]

As the same sentence runs from one stanza to the next, straddling the volta with a syntactic unit, it carries meaning through from the quatrains, unhindered by the semantic reorientation or change of perspective traditionally marked by the volta. While the technique is not very frequently used by Verlaine, its use is noteworthy in this early poem, as it prefigures the loss of form and the undermined dialectical structure characteristic of many of his later and poorest sonnets, like ‘A Irénée Decroix.’[21]

Second, Verlaine frequently obscures metre identity through a number of techniques aimed at disguising verse-line isosyllabism in the sonnet. In ‘A la meme’,[22] for example, a seemingly random and approximate approach is taken to metre, with up to six different ones appearing: 10/11/10/11 // 10/13?/10/12 // 9/10/12 // 10/10/14; similar exercises in, for example, the quatrains of ‘Pour S’,[23] 9/10/10/9 // 10/9/10/10, underline how destructive metric imprecision and unpredictability is to the recognisability of the sonnet: the proportionality of the eight/six dialectic and the internal balance achieved through the repetition of both stanzaic forms Q/Q/T/T are both lost; the uncertain contours of the verse-lines create a macrostructure which too is ambiguous and unsteady. Similar effects can achieved through the use of unconventional metres, like Verlaine’s famous impairs, which place at the centre of a highly-codified and regulated form, unstable and uncertain structures lacking precedent and definition, and whose qualities are not reminiscent of the sonnet, as Verlaine himself describes them in ‘Art poétique’[24]: ‘l’Impair / Plus vague et plus soluble dans l’air, / Sans rien en lui qui pèse ou qui pose’. Other metres, even when used exclusively, can disguise the sonnet: the dimensions of the tiny tetrasyllables of ‘A Bibi-Purée’[25] and the pentametres of ‘A Léon Cladel’[26] restrict the capacity of the quatrains and tercets to fulfil their traditional organizing and semantic roles, and disguise the identity of the sonnet, while much longer metres are particularly destabilizing, as demonstrated by the tredecasyllabic lines of ‘Sonnet boiteux’[27] and, especially, the fourteen-syllable lines of, for example, ‘Laurent Tailhade’.[28] In these cases, the long verse-line starts to approach the dimensions of prose, not least typographically and visually, and the sonnet tends to take on more complex or elaborate syntactical structures, at the expense of losing its sense of concentration, particularly in the tercets.

Third, Verlaine’s habitual recourse to the sonnet for occasional poetry brings an unsettling lightness and spontaneity, often bordering on the frivolous, to a form more conventionally associated with the rigour and logic of syllogistic argument. Often written on demand or on a whim, without much obvious effort or thought, these texts give the impression of seeking to compensate their prosaic nature by having the sonnet almost ‘confer’ poeticity on them by dint of its status as a ‘modèle culturel.’[29] The results are uniformly uninteresting and particularly unmemorable, as can be seen in these lines from ‘A mon amie Eugénie’, a sonnet written ‘pour sa fête’:

Contrariante comme on l’est peu, nom de Dieu !
Tu n’en fais qu’à ta tête, – et moi rien qu’à la mienne
Non plus – et je suis tel que je suis, quelque peu
Que je sois, et j’y reste en dépit de la tienne

De tête, et, nom de Dieu, j’adorerais ce jeu,
S’il ne me tuait pas en manière de tienne
Plaisanterie et de ta part et de la mienne,
Je dis un peu ce qu’il faut dire, nom de Dieu.

Je ne suis pas ni comme il faut, ni de génie,
Mais je me souviens qu’on te prénomme Eugénie
Et je me rappelle aussi que c’est aujourd’hui

Ta fête, et qu’il faut encor que je la souhaite
En dépit de nos torts de femme et de poète,
Et je t’envoie, ô, ce sonnet fait aujourd’hui.[30]

While the statistics on characteristics such as metre, rhyming scheme and stanza composition indicate that Verlaine’s sonnets are overwhelmingly conventional, the experience or feel of them points to something quite different and unsettling – a loss of the defining anxiety and uncertainty associated with reading a sonnet, as described by Clive Scott, and a sense of lost potential and unmet expectations. The effect is essentially destructive as the irregularities or breeches in question are rarely, if ever, compensated by an expressive reward. From 1880 onwards, Verlaine relies on the increasingly uncertain and frayed contours of his sonnet to provide an illusion of the fixed form on the page, with perhaps a familiar rhyming scheme. These texts rarely have the feel of sonnets: they are unpredictable and amorphous, and illustrate – even if it was not the poet’s intention – the subversive and destructive potential of a poetics fusing ‘L’Indécis au Précis.’[31]


Vers laine (1)
Vers laine inversé

Nicolas Fève, Vers de laine : Où l’Indécis au Précis se joint, 2015, photograph.

‘Mon Rêve familier’, first published in Le Parnasse contemporain in April 1866, provides an all-too-rare example of his potential virtuosity as a sonneteer, demonstrating how ‘the Precise’ and the ‘Undefined’ can motivate interesting and positive subversions of the fixed form.

Je fais souvent ce rêve étrange et pénétrantam6-6 x 4-4-4
D’une femme inconnue, et que j’aime, et qui m’aimebf6-6
Et qui n’est, chaque fois, ni tout à fait la mêmebf6-6
Ni tout à fait une autre, et m’aime et me comprend.am6-6
Car elle me comprend, et mon cœur, transparentam6-6
Pour elle seule, hélas ! cesse d’être un problèmebf6-6
Pour elle seule, et les moiteurs de mon front blême,bf4-4-4
Elle seule les sait rafraîchir, en pleurant.am6-6
Est-elle brune, blonde ou rousse ? – Je l’ignore.cf6-6 x 4-4-4
Son nom ? Je me souviens qu’il est doux et sonorecf6-6
Comme ceux des aimés que la Vie exila.dm6-6
Son regard est pareil au regard des statues,ef6-6
Et, pour sa voix, lointaine, et calme, et grave, elle adm6-6
L’inflexion des voix chères qui se sont tues.[32]ef6-6 x 4-8


At first glance, there is much that is conventional in and about this sonnet: the use of the Alexandrine, the regular arrangement of masculine/feminine rhymes, and the traditional abba abba ccd ede rhyme scheme qualify it as ‘regular’, even according to Banville’s strict classification.[33] In the two quatrains, stanza and sentence-length harmoniously coincide, while the first tercet’s syntactic and typographical separation from the second, which tends to undermine the sestet’s formal unity, is compensated by the cross-space echo created by the   d // e/d/e rhyme arrangement. Also, the apparent subject matter, a woman loved or admired by the poet, firmly places the poem in the long tradition of Petrarchan love sonnets.

But all is not as it may initially appear. In the first instance, several deliberate strategies are employed by the poet to weaken the perception of the metric structure of the Alexandrine: in line 7, the mid-line metric stress falls on the semantically-poor definite article (‘les’); in line 8, a syntactic unit (‘sait // rafraîchir’) straddles the mid-line caesura, as is the case in the last line (‘voix // chères’). The structure of this line is further undermined by the use of enjambement which creates a final line whose metricity depends on a diaeresis being permitted in ‘L’inflex-i-on’. Furthermore, the d rhyme ‘exila / elle a’, linking lines 11 and 13, though original and interesting, is rather weak: it is based on just two common sounds, only one of which is a vowel and the ‘rime d’appel’ is separated from its ‘rime d’écho’ not just by a full twenty-three syllables, but in addition by the typographical space between the tercets, and the move into a new sentence.

The recognisability of the Alexandrine is also threatened by Verlaine’s use of the Romantic trimetre: though not in itself particularly destabilizing, as the 4-4-4 form had attained, especially since Hugo, de facto metric status as an accompanying metre, as used here by Verlaine, in a context where the exometric support available from the 6-6 form has been considerably lessened by the poet’s efforts at disguising the caesura, the trimetre’s equivalence to the 6-6 Alexandrine is more difficult to perceive. Hence, in line 9, for example, the syntactic and rhetorical division into two at the eighth syllable, combined with the use of a mid-line dash and a question mark weaken the sense of presence of the caesura. In making the Alexandrine a more uncertain and less easily recognizable form, Verlaine is undermining not just the ‘mètre royal’ of French verse, but also the only strictly regular metre for the French sonnet.

In the second instance, our experience of reading the poem constantly forces us to adapt and readjust our assumptions and expectations in terms of subject matter and genre convention. While the first quatrain seems to advance us towards being able to visualize or more clearly imagine the woman in the poet’s dream, this expectation is never quite met. The first word describing the woman (‘inconnue’, l. 2), initiates instead a process of semantic retraction and negation, denying her any real substance and us any clear sense of her presence. Any apparently positive or affirmative description of the woman (for example ‘que j’aime, et qui m’aime’) is counterbalanced by a negative (‘Et qui n’est… ni / Ni…’). Furthermore, an impression of semantic stasis or immobility is created by the repetitive nature of what is said (‘et que j’aime, et qui m’aime […] et m’aime et me comprend’) and the circular logic of the tautology straddling the first and second quatrains: ‘et m’aime et me comprend // Car elle me comprend’. The impression is reinforced, rather paradoxically, by the completely conventional abba abba rhyme scheme of the quatrains, containing as it does ‘regressive’ or backward-facing stanza-ending a rhymes and the ‘static’ or immobile bb rhymes.

While we learn more in the second quatrain about the woman’s capacity to console the poet and her uniqueness in this respect, semantic progression is virtually held in check by a series of interconnected and interrelating repetitions: ‘elle seule’ appears three times in the same number of lines, and the recurring presence of /s/ and /l/ sounds, especially in lines 5-7, and of the /m/ and /c/ sonorities from line 5, create multiple phonic links between adjacent and non-adjacent verse-lines. Added to this, the echoes created between the last syllables of the quatrain (‘en pleurant’) and those, not just in the original ‘rime d’appel’ (‘transparent’), but also at the end of the quatrain’s first hemistich (‘me comprend’), brings us back to the beginning, acoustically, reinforcing the sense of regression. Each main idea also appears at least twice in this quatrain: the woman’s capacity for effectively consoling the poet (‘comprend’, ‘rafraîchir’) and her status as being the only woman who can do this for him (‘Pour elle seule’, ‘elle seule’). The tears of the final line, both in elemental substance and metaphoric extension, re-present the ‘moiteurs’ of the previous line, as well as, figuratively repeating, the more direct ‘problèmes’ of the previous line.

At the volta, we confidently expect some crystallisation or vision of the dream woman, especially as it has been delayed or deferred in the quatrains, though invitingly anticipated by the indirect reference to the woman’s face through ‘en pleurant’ (l. 8). Our expectation is based not just on the conventions of the love sonnet, but also on the potential for synthesis and revelation associated more generally with the two tercets. As David Scott reminds us, ‘[T]he sestet […] is the most powerful and dynamic part of the poem, the main theatre of action, in which the poet comes into closest contact with his material, concentrating it and purifying it.’[34]

But here, the urgent questions – ‘Est-elle brune, blonde ou rousse ?’, ‘Son nom ?’ – must meet with evasive and unsatisfying responses, if the woman is to remain an ideal figure, free of any conventional or cultural associations that specification would inevitably create.

Thus, it is at the very point in the sonnet where precisions and details are most eagerly anticipated that the woman retreats and slips away; both present and absent, silent and audible, she transcends binary oppositions – ‘ni tout à fait la même / ni tout à fait une autre’, –  and the echo of her voice is the Derridean ‘trace’.[35] She is simultaneously defined by and emblematic of the poet’s ‘deconstruction’ of dialectical structures, including those of the sonnet, through the promotion of non-contradiction and sameness. Hence, the tercets, which confirm and compound the quatrains’ difficulty in identifying or specifying the woman, can be seen more as an extension or continuation of the quatrains rather than as their counterpoint. The concepts are emphasized too throughout the poem, by the multiple repetitions, or re-appearing ‘sames’ already noted, by the homophonies of ‘même’ (l. 3) and ‘m’aime’ (l. 2, 4), and by the promotion of a non-dialectical logic based on coordination through the insistent and rather unusual use of ‘et’ in the first quatrain.

The last line of the poem reflects and reproduces the ambivalence and uncertainties of the poem. It simultaneously fulfils and subverts the conventional expectations for the last line of a sonnet. On the one hand, as it falls without fanfare into the actual silence of its last word, ‘tues’, itself a dissipating echo of line 12’s ‘statues’, line 14 is exactly of the opposite of the type of dramatic ‘stroke’ prescribed for the last line of a sonnet by many practitioners, including Banville: ‘Le dernier vers d’un Sonnet doit contenir un trait – exquis, ou surprenant, ou excitant l’admiration par sa justesse et par sa force.’[36]

However, on the other hand, in this respect, and in the sense that it provides a type of ‘anti-denouement’ whose own unconventional characteristics – instability, deferral of expectation and anti-climactic silence –  re-present and synthetize those of the sonnet, the last line can be viewed as fulfilling an essentially conventional role: ‘[I]l doit suffire de lire le dernier vers d’un Sonnet; car, […] un Sonnet n’existe pas si la pensée n’en est pas violemment et ingénieusement résumée dans le dernier vers.’[37]

Verlaine sets out, in ‘Mon Rêve familier’, not to represent the woman, but to harness the sonnet’s potential for creating ambiguity and uncertainty in an attempt to render the sensory experience and viscerality of the oneiric world, its confusion and altered logic. So it is not substance or ideas or (logical) meaning which are rendered in this poem but, in their place, the suspension, postponement and withdrawal of meaning and a sense of the vagueness and fluidity of the dream or subconscious world. For Gilles Vanier: ‘Le texte dans son ensemble correspond à un lent retrait qui laisse pourtant paraître quelque chose : le retrait lui-même, véritable sujet du poème.’[38]

Verlaine’s relationship with the sonnet is ambiguous and uneven: after a promising start, it suffers from the same lack of creative energy and ability that characterizes nearly all of his poetic output from the 1880s onwards. While an initial healthy disrespect for the fixed – whether metre, form or convention – primed Verlaine to investigate within the space of the sonnet the potential for developing ambiguities and uncertainties and to explore the limits of measured and codified poetic expression, it also led him eventually, much more subversively, to ‘de-sonnetize’ the sonnet by compromising its organic structure, dispensing with its defining conventions, and creating prosaic texts with just the visual appearance of the fixed form. It is little wonder then the poet should ask himself in the forgettable and insipid ‘A Mlle Eveline’,[39]Mais où va donc mon Sonnet ?’: the irony of the question being posed at the conventional turning point of the sonnet summarizes in itself the contradictions and ambiguities that characterize Verlaine’s use of this fixed form.



Banville, Théodore de. Petit Traité de poésie française, reprinted in Œuvres, vol. VIII. Geneva: Slatkine Reprints, 1972.

Baudelaire, Charles. Correspondance générale, vol. II. Edited by Jacques Crépet. Paris: Editions Lambert, 1948.

Cornulier, Benoît de. Art poëtique. Lyon: Presses universitaires de Lyon, 1995.

Derrida, Jacques. Speech and Phenomena: and other essays on Husserl′s theory of signs. Translated by David Allison. Evanston: Northwestern University Press, 1973.

Scott, Clive. French Verse-Art: A Study. Cambridge: Cambridge University Press, 1980.

Scott, David. Sonnet Theory and Practice in Nineteenth-century France: Sonnets on the Sonnet. Hull: University of Hull Publications, 1977.

Vanier, Gilles. Verlaine ou l’enfance de l’art. Paris: Champ Vallon, 1993.

Verlaine, Paul. Œuvres poétiques complètes. Edited by Yves Gérard Le Dantec and Jacques Borel. Paris: Gallimard, Bibliothèque de la Pléiade, 1962/1989.


[1] Paul Verlaine, Œuvres poétiques complètes, eds. Yves Gérard Le Dantec and Jacques Borel (Paris: Gallimard, Bibliothèque de la Pléiade, 1962/1989), 77.

[2] Verlaine, Œuvres poétiques, 972.

[3] See my reference to FG’s ‘disguised’ sonnet, L’Allée.

[4] Verlaine, Œuvres poétiques, 549.

[5] David Scott, Sonnet Theory and Practice in Nineteenth-century France: Sonnets on the Sonnet (Hull: University of Hull Publications, 1977), 69-71.

[6] Benoît de Cornulier, Art poëtique (Lyon: Presses universitaires de Lyon, 1995), 77.

[7] Charles Baudelaire, in a letter to Alphonse de Calonne, dated 8 January 1859, in Correspondance générale, vol. II, 405, ed. Jacques Crépet (Paris: Editions Lambert, 1948), 256.

[8] Clive Scott, French Verse-Art: A Study (Cambridge: Cambridge University Press, 1980), 172.

[9] Théodore de Banville, Petit Traité de poésie française (1872), reprinted in Œuvres, vol. VIII (Geneva: Slatkine Reprints, 1972), 207.

[10] Banville, Petit Traité, 205.

[11] Clive Scott, French Verse-Art, 173.

[12] Verlaine, Œuvres poétiques, 995.

[13] Verlaine, Œuvres poétiques, 629.

[14] Verlaine, Œuvres poétiques, 323.

[15] Verlaine, Œuvres poétiques, 60.

[16] Verlaine, Œuvres poétiques, 569.

[17] Verlaine, Œuvres poétiques, 948.

[18] Verlaine, Œuvres poétiques, 108.

[19] For example, “Quatorzain pour tous” and “Quatorzain pour toutes”, Verlaine, Œuvres poétiques, 580-1.

[20] Verlaine, Œuvres poétiques, 76-7.

[21] Verlaine, Œuvres poétiques, 566.

[22] Verlaine, Œuvres poétiques, 588.

[23] Verlaine, Œuvres poétiques, 583.

[24] Verlaine, Œuvres poétiques, 326.

[25] Verlaine, Œuvres poétiques, 617.

[26] Verlaine, Œuvres poétiques, 611.

[27] Verlaine, Œuvres poétiques, 323.

[28] Verlaine, Œuvres poétiques, 558.

[29] See note 4.

[30] Verlaine, Œuvres poétiques, 954.

[31] Verlaine, Œuvres poétiques, 326.

[32] Verlaine, Œuvres poétiques, 72. In the notation, m indicates the presence of masculine rhyme and f, a feminine rhyme.

[33] Banville, Petit Traité, 198-204.

[34] David Scott, Sonnet Theory, 32.

[35] Jacques Derrida, Speech and Phenomena: and other essays on Husserl′s theory of signs, trans. David Allison (Evanston: Northwestern University Press, 1973), 156.

[36] Banville, Petit Traité, 206.

[37] Banville, Petit Traité, 206.

[38] Gilles Vanier, Verlaine ou l’enfance de l’art (Paris: Champ Vallon, 1993), 79.

[39]. Verlaine, Œuvres poétiques, 603.

Print Friendly, PDF & Email
‘Mais où va donc mon sonnet ?’: Verlaine and the Fixed Form of PetrarchAlex Bradley