Navigation – Plan du site
I. Traduire le rythme du poème
Réflexions théoriques et pratiques traduisantes

Distending the rhythmic knot

David Nowell Smith
p. 29-45

Résumés

Cet article examine les points d’intersection entre deux formes de traduction : l’inter-linguistique et l’intra-linguistique. On y propose de distinguer la structure prosodique d’une langue (accent, quantité, intonation, cadence) de la rythmicité élargie, qui comprend non seulement des rythmes non-linguistiques mais aussi la motilité structurante isolée par Benveniste et Heidegger dans le rhuthmos dans son acception pré-socratique ou encore le nœud rythmique de Mallarmé. Cette distinction ouvre sur la question de savoir comment un poème peut, à partir d’une exploration de son matériau prosodique, engendrer une telle rythmicité. Comme le fait Michel Deguy dans sa lecture de Saint-Augustin, on considérera cette relation comme une forme de distension : ces vecteurs d’intension et d’extension mis en œuvre par la métrique mais irréductibles à celle-ci. Dans une lecture de sept traductions des derniers vers du sonnet « L’Angoisse » de Mallarmé, on s’interrogera sur les façons de conserver ces vecteurs dans une nouvelle langue, non pas pour identifier des équivalences mais pour en faire apparaître de nouveaux dans l’étoffe prosodique de la langue cible.

Haut de page

Texte intégral

for Clive Scott

  • 1 . See Heidegger (1998 [1939]) and Benveniste (1966 [1951]). It might seem problematic to try to al (...)

1What happens if we see translation not only as between rhythms from one language to another, but as between strata of rhythm within language—or more precisely, between rhythmicity and prosody? This is not simply to open language up to ‘rhythms’ that we habitually consider not to inhere within language itself (musical pulses and syncopations; circadian rhythms; the speed and din of the city; the varying work rhythms—and their economic and psychological effects—of the tenured academic, the adjunct teacher, the university administrator, and the cleaning staff). Rather, I have in mind temporalities within language that are not strictly speaking prosodic. By such “rhythmicity” I mean at the most basic level a pattern of movement in time; but we can also trace it back to the Greek rhuthmos which, following the glosses of Benveniste and Heidegger, one can posit as a structuring motility irreducible to the binary opposition of flux and fixity1. To distinguish rhythmicity from prosody is to enlarge our understanding of linguistic movement; to translate rhythm is to hear this structuring motility surface in the prosodic texture of a phrase, an intonation contour, a syncope or rhyme. Stéphane Mallarmé once said, “toute âme est un nœud rythmique” (1992: 644); that is, the soul is bound together through its rhythmic countermovements, something Mallarmé both models on language’s own rhythmicity, and uses to explain our own attentiveness to the workings of prosody. And, just as Mallarmé plays between these two levels of rhythm—between the metaphysical coherence of the soul and the segmentation of speech—I wish to ask how we might grasp such an enlarged, multidimensional rhythm in the prosody of speech and its deployment in verse. How, in other words, might we hear in Mallarmé’s own verse the entangling and disentangling of what here he calls the ‘rhythmic knot’?

  • 2 . I develop this particular strand of Mallarmé’s poetics in “Mallarmé and the Ontologization of Po (...)

2Opened up by Mallarmé’s phrase is a kind of ‘ontological difference’, opposing the rhythm that inheres in the soul to ‘ontic’ prosodic shapes2. Subtending my interest in this enlarged rhythmicity is ultimately a concern with its ontological stakes. I am particularly swayed by two accounts, both of which can find support in the Mallarméan “nœud rythmique”: Henri Meschonnic’s claim that rhythm is in the first instance the principle of human subjectivation in language, and Heidegger’s gesturing towards an originary movedness of language which antecedes our speech and sets in motion the linguistic horizon of being-human itself. For each, rhythm shapes our perceptual relation to language, but exceeds the perceptual: both return to the Greek root rhuthmos to this end: movedness which is structure and formation rather than flux. The present reflection, however, will leave such ontological considerations largely to one side.

3However, we might also wish to distinguish rhythm from prosody in order to do justice to the myriad forms a poem’s rhythmicity takes. For instance, structures of recursivity which develop over an extended period of time, registered psychologically or phenomenologically, in terms of mnemonics or anticipation, as a dialectic between linear progression and antilinear self-reflexion, as the motivic structure of a long poem, or sequence, or poetic oeuvre; epistemic blockages and double-takes, whether by virtue of syntax or parataxis, of polysemy or polyphony, in each case demanding that we waver between glosses, between voicings; the peculiar temporalities of deixis, of apostrophe, of other shifts in address, discourse, or tense, in which the dynamics of hearing and reading are animated, deanimated and reanimated. An enlarged rhythmicity thus confronts us with that peculiar temporal matrix which is a poem, insists upon the performative dimensions not only of the poem’s address to us, but of our encounter with it.

4To distinguish rhythm from prosody in this way is also to ask what kind of sensorium language is: marked not simply by sonority but by the inhabitation of time, and more specifically its rending asunder of linear, isochronic time. Poetry, so shall I argue, distends; in such distension are rendered audible the dimensions and dynamics of rhythm irreducible to isochrony, but equally to any attempt to attribute them simply to a lived, subjective durée. I say ‘distension’: as opposed to ex-tension, this implies a vertical as well as a horizontal axis: not merely ‘stretching’ a pre-existing temporal dimension, but rending dimension open as such. To speak of translating between strata of rhythm would be to ask how such rending is registered, how effected, in verbal language. In what follows I shall suggest that we grasp this distension not only through extensibility, but also through an opposite movement, whereby poems set up vectors of intensity—following Pound’s famous etymological wordplay, Dichtung = Condensare (Pound, 2010 [1934]: 97). Taut and slack are metrical values, but they might also describe the irruption of an aporetic, anti-isochronic time, into the metre itself; they hold in abeyance verse technique and speculative philosophy.

5One of the most compelling attempts of recent decades to outline a speculative prosody can be found in Michel Deguy’s 1979 essay L’infini et sa diction, ou de la diérèse. Speaking of a celebrated passage in Baudelaire’s essay on Théophile Gautier:

C’est […] le caractère de la vraie poésie d’avoir le flot régulier, comme les grands fleuves qui s’approchent de la mer, leur mort et leur infini, et d’éviter la précipitation et la saccade. La poésie lyrique s’élance, mais toujours d’un mouvement élastique et ondulé. (1996: 126)

6Deguy asks: “quelle est cette élastique ondulation qui caractérise la ‘vraie poésie’?”(1986 [1979]: 27). I want to see how far we can think this elasticity before it snaps.

7What is particularly extraordinary in Deguy’s argument is the inference that the constraint of the poem’s measure engenders and releases dimensions and dynamics in excess of the measure itself. In the OED definition 1.b of the verb distend, we find a now obsolete 17th-century distinction: to stretch beyond measure. “Beyond measure” requires measurement as its point of reference, as that which frames and gives access to such a “beyond”. The poetic measure—incarnated for Deguy in the construction of feet, in the counting of the e muet, and most crucially in the breaking up of diphthongs by diæresis—would impose isochrony on spoken language. Not only does this draw vocal performance away from speech, rendering audible cadences, contours, inflexions, otherwise unheard; it also creates the possibility of a syncope in which the rhythms of speech, diverging from the metrical set, attain greater æsthetic power. This Deguy calls:

[L]e boitement pair/impair constitutif du rythme […] Le doublement de l’impair produit le pair et l’« alexandrin » retombe sur sa parité même s’il ne retombe pas sur ses pieds. La pause, ou blanc, qui disjoint, si prompte qu’en soit la syncope, le temps fort du temps faible ; puis disjoint le « pied » […] de sa réitération, insère le supplément non compté qui rend possible toute extension de la diction, chaque lecteur pouvant varier infinitésimalement ce laps […] (ibid.: 27-28)

8The purpose of the foot, as it were, is to lose its feet, to trip up on itself. In the fissures between verse and speech patterns, figured most completely in the <i> of “Ayant l’expans-i-on des choses infinies” from Baudelaire’s “Correspondances”, sounds the irruption of the infinite into the poem’s finition. This is not simply a logic of syncopation: but of Derridean supplementarity. The conventions of verse writing first make speech proper to itself: speech is thus fatally marked by the non-proper. And the supplement inheres at a further level: the syllabic unit which introduces the principle of counting remains itself strictly speaking uncountable, always dependent on each new timing. And so he continues:

La pause […] introduit la distensio animi, ou temps, dans la substance phonique de ce métronome qu’est une phrase rythmée, ou vers, et en fin de compte un poème, par où le temps passe, prend figure de mesure […] Dans la cadence rythmique […] l’infini prend une mesure. Le vers est ce défilé où l’infini prend finition et le fini s’infinitise. Le vers est cette chose ayant sa limitation métrée qui est travaillé par la distension, l’extensibilité – expans-i-on de chose infinie. (ibid.)

9The extensibility of language within metrical measurement constitutes a clash of two different temporalities: the measure is rent open by the poem’s rhythms, yet it is this same measure which had engendered the excess opened up; it is thus that the word, finite both through its referential function and through its sonorous husk, comes to gesture aporetically towards the infinite.

10Yet one phrase here should give us pause: “ce métronome qu’est une phrase rythmée”. It appears here that ‘rhythmed’ (English lacks the verb) is being linked with isochrony, and this points to a broader problem. Distension, argues Deguy, is only truly activated in reading: “La lecture peut distendre – les vocables, le phrasé ; elle est élongation, étirement, ralenti de la langue” (ibid.: 31), yet this happens insofar as “La lecture, en tant que diction, cherche à freiner ‘indéfiniment’ le flux.” We are thus returned to a binarism of flux and fixity, rhythm versus metre as it were. Underpinning every theory of prosody is a tacit metaphysics; Deguy’s work is striking for the way his metaphysics pulls against his prosody.

11Might Deguy’s insight into the poem’s distension lead us out of that same binarism to which his account of measure falls prey? Such is my suggestion: to see this let us follow the distentio animi back to its source.

Le poème interfère, il distend les éléments sonores et les pauses, il entrouvre mots, syllabes, phonèmes. Reprenant les termes d’Augustin, c’est comme si par là l’âme entrait au-dedans de sa propre distentio, l’âme qui n’a pas d’autre « substance » de sa temporalité que cette entente poétique de son élément langagier. (ibid.: 30)

12Against this model of linguistic sonority as the static élément sonore as abstracted from its discursive-prosodic context—the syllable, the phoneme—we find a countervailing movement in the invocation of Augustine’s famous discussion of temporal-linguistic distension in the Confessions. But the example of Augustine offers a very different rhythmic dynamics.

O Lord, since you are outside time in eternity, are you unaware of the things that I tell you? Or do you see in time the things that occur in it? If you see them, why do I lay this lengthy record before you? (St. Augustine, 1961 [397-98 AD]: 253)

13So starts Augustine’s meditation on time and eternity in the Confessions. Trying to grasp what lies outside of time, he immediately has recourse to markers of deixis—the opening vocative, the rhetorical questions which demand the interlocutor’s presence, which perform a mind in the act of wavering, the self-referential ‘this lengthy record’: at once a scene of utterance and a scene of writing. The Confessions, whose conversion narrative intersperses theological argument, sermons, prostrations before God, is a work of great temporal complexity; hence the discussion of the problem of eternity in its penultimate chapter. Hence also, in a work concerned with linguistic temporality—narrative, argumentation, address—the depiction of the problem of eternity as first and foremost a linguistic problem. In the beginning was the Word; but what kind of word was it? God’s speech is the setting-into-motion of time, yet as it sounds to human ears its motion must be temporally bounded—in which case, God exists in time after all.

But how did you speak? Did you speak as you did when your voice was heard in the clouds saying: This is my beloved son? At that time your voice sounded and then ceased. It was speech with a beginning and an end. Each syllable could be heard and then died away, the second following after the first and the third after the second, and so on in sequence until the last syllable followed all the rest and then gave place to silence. From this it is abundantly clear that your speech was expressed through the motion of some created thing, because it was motion subject to the laws of time, although it served your eternal will. (ibid.: 258)

14Herein lies the problem: if God speaks in time, then something that is of God, his speech, ceases to be, and hence God is no longer an absolute plenitude of being. “We cannot rightly say that time is, except by reason of its impending state of not being” (ibid.: 264). Hence we must distinguish between two strata of language, an ontological difference between the Word and words:

These words, which you had caused to sound in time, were reported by the bodily ear of the hearer of the mind, which has intelligence and inward hearing responsive to your eternal Word. The mind compared these words, which it heard sounding in time, with your Word, which is silent and eternal, and said, ‘God’s eternal Word is far, far different from these words which sound in time. They are far beneath me; in fact, they are not at all, because they die away and are lost. But the Word of my God is above me and endures for ever’. (ibid.: 258)

15Augustine’s solution raises more problems than it solves. For now we have a sounding word which refers not to what it says, but rather to another, more originary Word, which is silent and eternal. Once more, the logic of the supplement: the silent, eternal Word must sound in time in order to be recognised as silent and eternal, in order that God first “made himself audible to the bodily ears of men” (ibid.: 259). But this means that the sounding word does not merely refer; rather, it distends, it rends open a further dimension through its sounding. The Word that sets time in motion, resounds within the very time it set up, it attains a body just as it bodies forth man. And inversely—bodying forth, it exceeds the body.

16Wavering between Word and words, it is in language that earthly time can distend into eternity, and eternity crystallises within time—“l’infini prend finition et le fini s’infinitise,” as Deguy had it. What is most striking about Augustine’s argument is then not ultimately his claims—brilliant though they are—about divine language, nor his response to the paradox of what God was doing “before” he invented the world (answer: “although you are before time, it is not in time that you precede it” (ibid.: 263); to ask this question is therefore to confuse the divine with the human); rather, we are struck by the predicament of Augustine himself, trying to use his words and their peculiar temporality in order to address God from within time, to bring God’s eternal truth into timebound human understanding. As with Deguy’s poetic measure, the rhythmic dynamics of Augustine’s meditation asks how measured speech can engender its own excess, how it can rend itself open in time.

17“Why is this so, O Lord my God?” (ibid.: 259); “He is the Beginning, O God, in which you made heaven and earth” (ibid.: 260); “These are tentative theories, Father, not downright assertions. O God, be my Judge and my Guide” (ibid.: 266); “O Lord, my Hope, allow me to explore further. Do not let me grow confused and lose track of my purpose” (ibid.: 267); “My mind is burning to solve this intricate puzzle. O Lord my God, good Father, it is a problem at once so familiar and so mysterious” (ibid.: 270); “Be resolute, my soul. Onward with all your endeavour” (ibid.: 275). “O Lord my God, how deep are your mysteries!” (ibid.: 279) Augustine’s text is scored with deictic markers—with questions, citations, prayers, vocatives, immersing him in the moment of questioning, of reasoning, arguing, and—most importantly—of uttering. This introduces a countermovement to the negative theology of eternity he seeks to establish; the constant questioning of God is both a means of bringing God into time, and bringing ourselves out of time. But it does so, paradoxically, through a saturation of the present, rather than a sublimation of it: distension not by extensibility, by l’expansion des choses infinies, as Deguy reads Baudelaire; distension by intensity.

  • 3 . Augustine does in fact discuss metre as one means of measuring time, later in this book of the C (...)

18And there is one further fundamental difference between Augustine’s verbal performance and Deguy’s account of the diction of l’infini: whereas Deguy extracts speech from discursive time through the reification of the foot, of the measured syllable, Augustine traces the opposite movement, insisting on the present moment of discourse as the basis for temporal excess. No binarism here between rhythmic flux and metrical fixity: the sounding word is grasped not as a phonemic unit but as an utterance in its totality3. Deguy is caught between two forms of stasis: phoneme on the one hand, infini on the other. Augustine demonstrates, at times despite himself, that language is continually in motion; what it strives after is not a stable infinite, but its originary motility.

19So—what happens if we try to hear temporal distension of the kind that Augustine describes within the prosodic texture specific to a poem? Might we here trace the entanglement of the rhythmic knot? Thus far I have been talking of the translation between strata of language, between the finite poem and the infinite it figures and opens up, between sounding word and divine Word: translation as distension, not as transfer between ontic languages. But perhaps inter-linguistic translations might help us grasp the dynamics of such intra-linguistic, inter-rhythmic translation. To ask how to retain these vectors of intensity and extensibility in a new language, with a radically different prosodic support, is not only to make speculative equivalences; it is to reassert the specificity of each language’s prosodic support. Walter Benjamin once claimed that the aim of every translation is ‘to regain pure language fully formed in the linguistic flux’, a pure language ‘which no longer means or expresses anything but is, as expressionless and creative Word, that which is meant in all languages’ (Benjamin, 1970 [1923]: 80); Mallarmé is speaking of every soul—if the nœud rythmique is not pre-linguistic, then nor is it dependent on a particular prosody, even if each time it is within a specific prosody disentangled, reentangled, anew. Hearing the emergence of vectors of intensity and extensibility out of a singular verbal-discursive matter also entails attending to the residuum of unsubtractable prosodic noise in which each rhythmic knot is bound. So let us turn to an early instance of Mallarméan prosody, a particularly knotty hemistich from the final couplet of his 1866 sonnet “L’Angoisse”. Here is its final sestet:

Car le Vice, rongeant ma native noblesse,
M’a comme toi marqué de sa stérilité,
Mais tandis que ton sein de pierre est habité
Par un cœur que la dent d’aucun crime ne blesse,
Je fuis, pâle, défait, hanté par mon linceul,
Ayant peur de mourir lorsque je couche seul.

  • 4 . One of the pleasures of a conference such as the TRACT conference Traduire le rythme is when it (...)

20The word pâle becomes the point of prosodic condensation here, a result of the commas enclosing it, of the accretion of two accents without unstressed syllable between, of the circonflex accent drawing out the vowel length and incurring upon the e muet that follows4. Yet its prosodic density results not simply from phonic patterning, but also from the demands it makes on our interpretive activity as those reading, and voicing, the poem. Its syntactical rhythm is placed in syncopation with the question of how we scan it: the opening hemistich Je fuis, pâle, défait could plausibly be rendered as two 3-syllable feet, as a two-syllable foot followed by a single syllable foot and then a three syllable foot, or, treating the e muet of pâle as a coupe lyrique rather than coupe enjambante, as a hemistich of three feet of two syllables each, where the second of the feet (pâle) is inverted to give a falling rather than rising cadence. This third scansion is the most plausible in terms of speech rhythms, but, as we will see below, the hemistich’s broader rhythmic context complicates a reading purely in terms of ‘natural’ speech rhythm. For the time being, let us note simply that the question of scansion is at the same time a question of how to voice the line: not an academic exercise but our means of negotiating, in linguistic time, the life of the poem on the page and in sound.

21“Entre les feuillets et le regard règne un silence encore, condition et délice de la lecture,” says Mallarmé in “Mimique” (1992: 310). In this silence between page and eye wavers our voice, suddenly rent open, distended.

22How is this condensation negotiated in translation? Below are seven translations from the last hundred and ten years:

I wander palely, haunted by my shroud,
Fearing to die if I but sleep alone. (Symons, 1974 [1902]: 71)
I fly, pale, undone, and by my shroud haunted,
And fearing to die if I but sleep alone. (Fry, 1936: 57)
wasted and pale and haunted by my shroud
I flee, afraid to die if I sleep alone. (MacIntyre, 1957: 13)
Obsessed by my shroud, I flee, pale, undone,
Afraid of dying when I sleep alone. (Weinfield, 1994: 15)
I flee, pale, haunted by my shroud, uncouth,
afraid of dying when I sleep alone. (Blackmore, E.H & A.M., 2006: 17)
I flee, pale, defeated, haunted by my tomb,
Afraid of dying when I sleep alone. (Scott, 2008: 47)
I flee, unmade, pale, haunted by my shroud,
afraid of dying when I sleep alone. (Manson, 2012: 37)

23Before turning to the accumulation of je fuis, pâle, défait itself, let us look at some of the broader choices regarding the poem’s ambient intertexture that the translations make. Firstly, Mallarmé’s use of rhyme. Of the seven translations, four attempt to preserve, or better, recreate, rhyme in the English: Symons, Weinfield, the Blackmores and Scott. And yet, their rhyming diverges markedly from Mallarmé’s original. One of the most striking things about this sonnet, as well as “Renouveau” and “Le Sonneur”, written around the same time, is their employment of rhyming couplets to finish, a very “English” convention. In a way it is less glaring when he uses couplets in his later sonnets, whose heptasyllables and levity draw them out of the alexandrine sonnet tradition. Of the four translations that employ rhyme schemes, Symons and the Blackmores both adopt cross rhymes rather than couplets to Mallarmé’s own (one of the reasons that one cannot discern a rhyme scheme from the brief quotations given above), resulting paradoxically in a more “French” sonnet form, and the two couplets, in Scott and Weinfield, both make do with half-rhymes: tomb / alone, undone / alone, in contrast to the rich rhyme ceul / seul.

24Perhaps this seems a small gripe: it’s not as if this poem bears the exuberant assonance of, say, Verlaine. Yet what is striking in Mallarmé is the peculiarly rhythmic dynamic of rhyme. In his discussions of versification in the lectures on aesthetics, Hegel suggests that Romantic (which for him meant post-Hellenic) poetry is characterised by its increased subjective inwardness. This is facilitated by a transformation in the model of language, now understood as sign, with a corresponding shift in linguistic sound, from the plasticity of intonational phrase to the metrical onus on the ictus. The sonorous, exterior, objective dimension, now grasped as external to meaning, attains a mimetic function through convergences and divergences of sound and sense, syntactical and metrical segmentation. Rhyme constitutes ‘the one possible compensation for this loss’ of plasticity (Hegel, 1975 [1835 post.]: 1027). Hegel is, I think it is fair to say, not a fan: “in contrast to the delicate movements of rhythmical harmony, rhyme is a thumping sound that does not need so finely cultivated an ear as Greek versification necessitates” (ibid.: 1028).

25Mallarmé’s virtuosic rhyming is, for sure, predicated on a split of sound from sense, as is most clear in his heptasyllables where rhyme becomes a form of punctuation when the syntax has been so stretched—it both contains the syntax and yet by the same gesture gives a greater sense of elasticity. One can see the “Petit air: Indomptablement a dû”, or “À la nue accablante tu” as examples of this. Yet the rhyme’s deep implication in the poem’s syntactic thrust, as fetter, as counterweight, as punctuation, also undoes this split. Rhyme is not mere decoration, nor mere ‘verse form’; nor is it simply assimilated mimetically into the poem’s content. Attaining this multivalent syntactic power it comes to resemble the plasticity Hegel considered characteristic of Classical poetry, but also the rhuthmos of Heidegger and Benveniste (also incidentally reconstructed from pre-Socratic thought): the structuring motility of language. In Mallarmé’s poetic performance, end-rhyme becomes determinant not simply as an element in a soundworld but as the crux of a philosophical-linguistic praxis. The choice of half-rhyme in the translations of even so early a poem is, then, particularly striking: half-rhyme both gestures towards the reconciliation of sound and sense and withholds it, a rejoinder to Mallarmé’s reactivation of plasticity. Weinfield’s in particular, with its eye rhyme of done and lone, and the assonance between sleep and flee which arises from inverting the hemistiches of line 13, makes this deliciously pointed. The translation’s reflection lies as much in its prosodic divergence as in its transfer of the original into the target language.

26For Hegel, the significance of rhyme increased in languages, like French, with a less pronounced ictus: there is greater need for its emphasis to separate verse from prose. This means, by contrast, that the powerful use of the accent within a syllabic versification needs to be negotiated in the tonic-syllabic prosody of English. This is the problem posed by Je fuis, pâle, défait, and its equivocation between voicings. The most literal translation—I flee, pale, undone—is also the one closest to preserving its prosodic and syntactic density: if we lose an aspirated <e> after pâle, we compensate with an implied upbeat before pale. It is, then, somewhat curious that so many diverge from it; again, it is in divergence that we find the most powerful prosodic—but also rhythmic—reflection. Symons aims to keep the intonation contour of the pentameter line, and so adverbializes an adjective: both grammatically and metrically he imposes forward movement where Mallarmé’s concentration of accents and equivocating scansion would interrupt it. Fry inserts the word and, thereby defusing the momentary grammatical indeterminacy of Mallarmé’s hanté (is it adjective or past participle?), and then offers an uncharacteristic piece of poetic diction through the change in word order. This seems both superfluous and metrically fascinating: he is evidently desperate to avoid an easy cadence at the end of the line that haunted by my shroud would offer. MacIntyre adds and in place of the punctuation, and by changing the place of I flee makes the grammatical subordination far more pronounced, whilst himself also introducing some jarring poetic diction. These moments of diction and metricality in the free verse translations are fascinating: what kind of artistic choice do diction and metre become when no longer part of one’s basic repertoire? Why introduce them now? It is as though, having abjured the constraints of scansion, they generate a syncope in the shift between registers. As for the Blackmores’ translation, this is a warning of what can happen when, in a rhyming translation, the tail wags the dog.

27This brings me to what I consider the three most interesting recent translations, by Henry Weinfield, David Scott and Peter Manson. Why does Manson (whose rendering of défait as unmade is indicative of a single-minded literalism which is one of his translation’s great virtues and a constant source of bafflement) invert unmade and pale? Why does Scott opt for defeated rather than undone? Why does Weinfield switch the hemistiches? For Weinfield the answer seems clear: it places real prosodic force on pale, something facilitated by the choice of Obsessed for hanté and the forward thrust this activates. Yet, where Weinfield attempts to recreate, in a pentameter line, the vectors of intensity of Mallarmé’s original, both Scott and Manson force English into adopting a kind of alexandrine, replete with medial caesura—albeit a deeply accentual one: a six-beat line with pause in the middle. Manson’s choice of pale for the third beat forces a longer pause at the middle of the line, whereas defeated prolongs the cadence, and our breath. The two lines create wholly different vectors of prosodic movement, wholly different intonation contours; yet they exact the same basic metrical shape of the English verbal matter with which they are working.

  • 5 . The only divergence from this is MacIntyre’s I flee, afraid to die if I sleep alone, yet this co (...)
  • 6 . This might remind us of Giorgio Agamben’s claim in ‘The End of the Poem’, that the impossibility (...)

28The great divergence in fine translations by excellent translators reflects the extent to which the hemistich’s rhythmic dynamics arise out of its prosodic intertexture. This can be seen in the striking metrical fluency with which each translation endows the final line: Fearing to die if I but sleep alone; Afraid of dying when I sleep alone5. Symons’s Fearing to die if I but sleep alone, adopted by Fry, is particularly interesting, not only for its inversion of the first foot, creating greater momentum later in the line, but also for the archaic and semantically excessive but: its role as intensifier is subordinate to its metrical role: an unstressed syllable, yet one whose semantic role and its echo and inversion of to earlier in the line demands prosodic emphasis ([tə] becomes [ət]). One of the ways the English translations register the density of Mallarmé’s hemistich, then, is retroactively, through release, as though telling us what it is like to try to preserve the vectors of intensity and extensibility of these lines in English6.

29But they are also replaying the dynamics of the French, where rhythmic condensation is surrounded by a flot régulier on either side:

Par un cœur que la dent d’aucun crime ne blesse, 3 + 3 | 3 + 3
Je fuis, pâle, défait, hanté par mon linceul,[?] | 2 + 4
Ayant peur de mourir lorsque je couche seul. 3 + 3 | 4 + 2 (1 + 3 + 2)

  • 7 . Words in boldface are used liberally throughout the text to emphasize crucial points.
  • 8 . The rhythmic effects of figures of repetition such as alliteration and assonance lies beyond the (...)

30Above we saw that our speech rhythms gravitate towards scanning the opening hemistich as 2 + 2 + 2. Why, then, do I leave it unmarked? Because complicating our desire to give any single scansion is the broader prosodic context of the tercet. The opening hemistiches of lines 12 and 14 share the same distribution of accent: foot symmetry gives on to stanzaic symmetry, something reinforced by the rhyme between cœur and peur—indeed, breaking up the words into their constituent phonic matter, we see Par un cœur7 contracted into p-eur. When pâle shares the same syllabic position as each, and alliterates with peur in the following line, this metrical parallel will attunes our ears to the peculiar density of Je fuis, pâle, both as it highlights its syncopation from a 3 + 3 structure, and as it gestures towards the possibility of its being reconciled with this structure8. This is complicated further by the fact that this structure is only established retrospectively, when line 14 signals the presence of pattern. Line 13, it would seem, both contributes to the establishment of pattern, and refuses to fit the pattern itself.

31Would it then be possible to collapse Je fuis, pâle as a single foot? This would respond to the accentual pattern of lines 12 and 14, but reading fuis as un-accented would defuse the accretion of accents, ride roughshod over the coupe. Coupe and accent become thus come to sound as interference and disruption, despite being ‘metrically’ uncounted—this is, after all, a syllabic and not accentual metre. This then bears upon our aspiration of the e muet that follows—something which will recalls Deguy’s observation regarding the e muet: a graphic and phonic blank which obtains weight when counted metrically, but which remains without determinate quantity outside each individual voicing. One can compare the <e> of pâle to that of crime ne blesse or je couche seul in the lines directly preceding and following it. This e muet, obtaining phenomenal body out of blankness, immediately has this body smothered again. As noted, this is already suggested by the elongation of the circonflex, yet the disproportions between individual syllables brought to the fore here itself sits uneasily with any scansion which gives equal metrical weight to each syllable. It is in this infinitesimal aspiration that we will equivocate, however momentarily, between coupe enjambante and coupe lyrique. The intension-extension of lengthened <a> and aspirated <e> becomes the site of the line’s prosodic intensification.

32But what is particularly striking is that the more we wonder about the transposition of Mallarmé’s purportedly ‘syllabic’ versification into the ostensibly accentual prosodies of the seven translators (particularly the case, I suggest, in Scott and Manson, where the syllabic length of the verse line is undercut by implied offbeats and upbeats), the more we appreciate the dense workings of accent in Mallarmé’s original. Looking at the question of rhythmic intensity in terms of accent in English helps bring into focus the role played by the accent in French, where its various effects are both metrical—as it offers a scansion of feet which subsequently, to follow Deguy’s witticism, trip over themselves—and para-metrical. The translation (or transposition… or transfiguration…) of prosodic intensities and extensibilities tells us something about the original that had perhaps remained latent until interlinguistic, interprosodic dialogue was set in motion.

33In the most successful translations of these vectors, what we see translated is not the prosody, but rather the motility reactivated out of a different prosodic support. And indeed, a different body: to syncopate between scansions, is to equivocate between voicings—but how can a voicing both equivocate and yet voice? Only in the generation of a voicing does this become syncopation; and yet one cannot voice without dissolving the syncopation. The line’s temporal dislocation is both bodied forth and denied body. Mallarmé’s vectors of intensity and extensibility are not simply prosodic insofar as they engage what cannot be registered prosodically—herein lies their particular motility. Perhaps it is here that we should ultimately situate the distension of prosody into rhythmicity.

34I would not, then, wish to make some kind of analogy between inter-rhythmic translation and inter-linguistic translation; rather, let me finish by suggesting that translation between prosodies might cast light on the fraught place of linguistic specificity in the apparently universalising tendency of Mallarmé’s speculative claim. If the translator aims, in Benjamin’s phrase, to recreate in the new language the effect of the original on its own language, then the translation of prosody becomes a translation not of metres or even syncopations, but of vectors of intensity and extensibility through which in the target language the text sets up those measures that it will subsequently rend open, continually distending the rhythmic knot, with what Baudelaire might well have called “un mouvement élastique et ondulé”.

Haut de page

Bibliographie

Corpus Texts

Mallarmé, Stéphane, 1992, Œuvres complètes, H. Mondor and G. Jean-Aubry (éds), Paris, Gallimard, coll. Bibliothèque de la Pléiade.

—, 1936, Poems of Mallarmé, trans. Roger Fry, London, Chatto & Windus.

, 1957, Stéphane Mallarmé: Poems, trans. C. F. MacIntyre, Berkeley, University of California Press.

, 1974 [1902], Selected Writings, trans. Arthur Symons, Manchester, Carcanet.

, 1994, Stéphane Mallarmé: Collected Poems, trans. Henry Weinfield, Berkeley, University of California Press.

, 2006, Stéphane Mallarmé: Collected Poems, trans. E. H. Blackmore, and A.M. Blackmore, Oxford, Oxford University Press.

, 2008, Mallarmé Sonnets, trans. David Scott, Exeter, Shearsman.

, 2012, Stéphane Mallarmé: the Poems in Verse, trans. Peter Manson, Miami, Ohio, Miami University Press.

Works cited

Agamben, Giorgio, 1999 [1996], “The End of the Poem”, in The End of the Poem: Studies in Poetics, trans. Daniel Heller-Roazen, Stanford, Stanford University Press, p. 109-115.

St. Augustine, 1961 [397-98 AD], Confessions, trans. R. S. Pine-Coffin, Harmondsworth, Penguin.

Baudelaire, Charles, 1996, Œuvres complètes II, éd. Claude Pichois, Paris, Gallimard.

Benjamin, Walter, 1970 [1923], “The Task of the Translator”, trans. Harry Zohn, in Illuminations, ed. Hannah Arendt, New York, Fontana, p. 69-82.

Benveniste, Émile, 1966 [1951], “La notion de ‘rythme’ dans son expression linguistique”, in Problèmes de linguistique générale, Paris, Gallimard, p. 327-335.

Deguy, Michel, 1986 [1979], “L’infini et sa diction ou de la diérèse”, in Choses de la poésie et affaire culturelle, Paris, Hachette, p. 21-32.

Hegel, G. W. F., 1975 [1835 post.], Aesthetics: Lectures in Fine Art vol. II, trans. T. M. Knox, Oxford, Clarendon Press.

Heidegger, Martin, 1998 [1939], “On the Essence and Concept of Phusis in Aristotle’s Physics B, I”, trans. Thomas Sheehan, in Pathmarks, ed. William McNeill, Cambridge, Cambridge University Press, p. 183-230.

Nowell Smith, David, 2012, “Mallarmé and the Ontologization of Poetry”, in J. Acquisto (ed.), Thinking Poetry: Philosophical Approaches to Nineteenth-Century French Poetry, New York, Palgrave.

Pound, Ezra, 2010 [1934], ABC of Reading, New York, New Directions.

Haut de page

Notes

1 . See Heidegger (1998 [1939]) and Benveniste (1966 [1951]). It might seem problematic to try to align ‘structure’ and ‘motility’ in Benveniste’s reading of rhuthmos, as he argues so powerfully against any attempt to read rhuthmos as “couler” and “le mouvement régulier des flots” (Benveniste, 1966 [1951]: 327), and indeed would situate it first of all as a spatial configuration. There is not the space here to elaborate on Benveniste’s account at any length, yet I would like to note that what is equally at issue in his essay is the schematic separation of spatial and temporal modes of form or formation. If rhythm means “la modalité particulière à [l’]accomplissement” of a form (ibid.: 332), then this modality its itself at movement. Such is the lesson we should take from Heidegger in his Aristotle reading, in which he reads morphe (“form”) as a form of kinesis (“movedness”).

2 . I develop this particular strand of Mallarmé’s poetics in “Mallarmé and the Ontologization of Poetry” (Nowell Smith, 2012).

3 . Augustine does in fact discuss metre as one means of measuring time, later in this book of the Confessions, he restricts himself to two observations: (1) that long and short syllables are relative, rather than absolute terms; (2) that we measure them through a complex of future and past, anticipation and memory. For both, the central claim is that we cannot measure time objectively, but rather time exists in the mind. However, it is striking that his most powerful claims about the temporality of language lie not in this brief discussions of a quantitative metre, but rather in the textual performance of time through its deictics.

4 . One of the pleasures of a conference such as the TRACT conference Traduire le rythme is when it gives rise to moments of collective thinking, and in what follows I am indebted to many of the participants for helping me see further levels of complexity and nuance in these lines.

5 . The only divergence from this is MacIntyre’s I flee, afraid to die if I sleep alone, yet this contrasts with the metrical fluency of his previous line, as though he has inverted the prosodic relation between the two lines in Mallarmé’s original.

6 . This might remind us of Giorgio Agamben’s claim in ‘The End of the Poem’, that the impossibility of enjambment in a poem’s final line will lead every metrical poem to a crisis in which what is at stake is nothing less than its dissolution into prose (Agamben, 1999 [1996]:115). Here the sudden metrical impetus would seek to reinscribe ‘regularity’ at the very point that such regularity is shown to be unsustainable beyond the verse itself.

7 . Words in boldface are used liberally throughout the text to emphasize crucial points.

8 . The rhythmic effects of figures of repetition such as alliteration and assonance lies beyond the present account, but is an issue of enormous significance and one whose import has not to my knowledge received the systematic attention it demands.

Haut de page

Pour citer cet article

Référence papier

David Nowell Smith, « Distending the rhythmic knot », Palimpsestes, 27 | 2014, 29-45.

Référence électronique

David Nowell Smith, « Distending the rhythmic knot », Palimpsestes [En ligne], 27 | 2014, mis en ligne le 01 décembre 2016, consulté le 21 août 2017. URL : http://palimpsestes.revues.org/2017 ; DOI : 10.4000/palimpsestes.2017

Haut de page

Droits d’auteur

Tous droits réservés

Haut de page
  • Logo Presses Sorbonne Nouvelle
  • Revues.org