Navigation – Plan du site
II. Traduire le rythme de la prose
La participation du lecteur: récepteur et/ou créateur du rythme

Translation and the expansion of the rhythmic sense

Clive Scott
p. 219-237


Cet article s’intéresse principalement à la manière dont la traduction peut se faire l’instrument de l’intensification et de l’extension des rythmes sous-jacents au texte de départ, aussi bien que de ses structures formelles ; la traduction n’est ni la reproduction ni la représentation d’un texte, mais plutôt un enregistrement phénoménologique de l’expérience de sa lecture. L’argument porte sur deux manifestations particulières de cette proposition, à savoir : la proxémique rythmique, c’est-à-dire les proximités et les distances perceptuelles engendrées par le rythme ; et la « rythmicité », c’est-à-dire le rythme comme événement performatif, comme énergie multidimensionnelle, comme champ de force du paralangage, plutôt que comme propriété innée du texte. Les véhicules de cette enquête sont des traductions du « Dormeur du val » de Rimbaud et de « Into my heart an air that kills » de A.E. Housman.

Haut de page

Texte intégral

1I have, in recent years, been pursuing the argument that translation should attend more to the phenomenology of reading than it should to the interpretation of texts (2012a, 2012b). Translation is an exploration, palpation, and inscription into the source text, of readerly consciousness, and this involves not only finding a form adequate to that consciousness—a search which, for me, has shifted from free verse (Scott, 2000), in the wake of Yves Bonnefoy’s practice, to different kinds of tabular and multi-medial disposition—but also wresting translation from the monopoly of the monoglot reader (with its narrow constraints of fidelity and reliability) and retrieving it for a polyglot reader, able to read the source text. I want here to take that project a step further, using rhythm as my instrument. I set out from the assumption, as argued elsewhere (2012a: 31-54; 2012b: 111-32), that metre and rhythm spring from radically different principles, that metre is linguistic-textual while rhythm is paralinguistic-performative, that metre is mono-dimensional, quantitative, homogeneous and discontinuous, while rhythm is multi-dimensional, qualitative, heterogeneous and continuous. I want particularly to explore two translational expansions of metrico-rhythmic consciousness/perception, the first of which has to do with rhythmic proxemics, that is, with a particular activity of rhythm, the second with the experience of rhythmicity (as an extension of “rhythm”), that is, with the multi-dimensional field of rhythm as a whole. But, first, my founding assumptions require further development.

2In Relevés d’apprenti (1966), Pierre Boulez tirelessly pursues a serialism devoted to the development of polyrhythmy, the multiple activity of rhythm across timbre, duration, attack, dynamics, tempo, silence and so on. For Boulez, “tout a été remis en question” in order that “se crée une organisation nouvelle du monde sonore” (1966: 183-184). One might argue that this is where the worlds of music and language sharply diverge: language ceases to function if it transgresses certain limits of communicative intelligibility and any ambitions to expand language’s polyrhythmic potentialities must therefore be kept in check. But there are significant steps that can be taken in the enlargement of our understanding and exploitation of rhythm: 1. to cultivate ways of better perceiving its multi-dimensionality; it is too often assumed that rhythm, like metre, is essentially mono-dimensional, about accentuation / stress only; 2. by virtue of performance, (a) to aid that process of perception; (b) to encourage readers actively to explore the polyrhythmic potentialities of a text; and (c) to develop a scansion to enable that exploration to be described or projected. This is further to assume three rather disparate things: (i) performance ensures the reader’s re-inhabitation of polyrhythmy and defines the reader’s task not as the identification of the rhythmic possibilities of a text, but the rhythmic possibilities offered by the embodiment of text; (ii) translation is peculiarly called upon to perform the expansion of polyrhythmic perception; (iii) a paradox: it is the printedness of verse which has inhibited our perception of the multi-dimensional nature of rhythm, that has encouraged us to see rhythm as a variation of metre, operating in the same dimension as metre, but it is the page which can deliver rhythm into its multiplicity, through the following means: (a) by developing a visual paralanguage which acts like a transcription of vocal paralanguage; (b) by developing page layouts which suggest simultaneities of effect—translating the linear into the tabular; (c) by developing page layouts which take the voice conceptually beyond the limits of vocal normality and even of vocal capacity (as musical graphics / musikalische Graphik does for music), into a realm where duration, silences, dynamics can act out their full effects without our having to adopt the delivery modes of an Artaud, a Bernard Heidsieck, a Bob Cobbing, i.e. into a new kind of reading. This is, of necessity, to develop a translation which is itself a scansion, but a scansion re-imagined, a scansion able to take us, if need be, beyond the bounds of performative possibility, a scansion which may, on occasion, confine performance to the visual paralanguage on a page.

3As a first step in this small odyssey, I would like to enlist the aid of layout and visual paralanguage to extend our sensitivity to one particular form of rhythmic activity, which I am calling “rhythmic proxemics”. But if it is one of the tasks of translation to sharpen both rhythmic and rhythmo-structural apprehension, then in my particular example, a translation of Rimbaud’s “Le Dormeur du val” (2009: 112), I shall have to find ways not only of drawing out rhythmic proxemics, but also of giving new rhythmic energies to sonnet-structure:

  • 1 . An apostrophe after the number indicates a coupe lyrique (coupe after word-terminal counted e), (...)

C’est un trou de verdure où chante une rivière
Accrochant follement aux herbes des haillons
D’argent; où le soleil, de la montagne fière,
Luit : c’est un petit val qui mousse de rayons.


Un soldat jeune, bouche ouverte, tête nue,
Et la nuque baignant dans le frais cresson bleu,
Dort; il est étendu dans l’herbe, sous la nue,
Pâle dans son lit vert où la lumière pleut.


Les pieds dans les glaïeuls, il dort. Souriant comme
Sourirait un enfant malade, il fait un somme:
Nature, berce-le chaudement : il a froid.


Les parfums ne font pas frissonner sa narine;
Il dort dans le soleil, la main sur sa poitrine
Tranquille. Il a deux trous rouges au côté droit.


4First, a prise de position: I am not translating these alexandrines as sequences of language which fulfill the form of the alexandrine (12 syllables, medial caesura, tetrametric unless deviating into a trimetric alternative), but rather the reverse: I am translating these alexandrines as activations of certain kinds of rhythmic experience which can be re-engineered, deformed, amplified, in other metrico-rhythmic contexts. This is not to say that the alexandrine has no peculiar metrico-rhythmic ontology; it is to say that where that ontology is a given, where the reader is polyglot and thus already has the alexandrine as a readerly presupposition, then what is to be explored through translation is not the given, but the perceptual repercussions—isolations and emphases, intensities and forces, in the reading mentality—and to find a form adequate to those repercussions.

  • 2 . It should be observed that Rimbaud further heightens our sense of sudden shifts in consciousness (...)

5The rhythmic experience activated by the late 19th-century alexandrine is one of problematic junctures and the consequent disruption of standard sequences of measures. Rimbaud’s poem is a dramatization of the tension between the disjunctive and the conjunctive, juncture and rejet (enjambement over both caesura and line-ending), which is connected not merely with the staging of revelation—the real nature of the soldier’s sleep and the reason for the “depth” of his assimilation into the natural environment—but also with rhythmic proxemics, that is, different degrees of rhythmic/perceptual intensity, different sensory distances, which exploit what could be identified as an Impressionist optics/rhythmics: the play between the “colour” notation in close-up (sensation) and the blending of such notations at a distance (mood, impression). Rimbaud’s rejets are moments of sensory fixation, of magnification, of sudden focus. It is as if the ongoing description, in its pastoral ease, were subject to eruptions of awareness whose value cannot quite be determined2. In themselves they produce a charge of perceptual innocence which Rimbaud can then turn, by small twists, into perceptual knowledge. No doubt Rimbaud became familiar with these techniques partly through Verlaine’s poetic practice. Indeed, one might suspect that Rimbaud’s rejet-technique is in fact a parody of rejet-technique, that the real subject of this poem is not so much a casualty of the Franco-Prussian war, as a poetry whose dislocative effects are becoming too automatic, too transparent. But in the two lines which have trimetric possibilities, lines 5 and 14, acoustically affiliated by their repeated [u]’s around the caesura, we are asked, as readers, either to opt for a telegraphic, peremptory, matter-of-factness (trimètre):

Tranquille./Il a deux trous rouges/au côté droit. 2>6’>4

6or the sensory émerveillement of surprised proximity (tétramètre, caesura with enjambement):

Tranquille./Il a deux trous//rou/ges/au côté droit. 2>4>2’>4/2>4>1>5

  • 3 . I realize that, in providing this notation, I am in danger of undermining my own argument and of (...)

7It should be said immediately that English verse is bound to have difficulty with this kind of rhythmics, (a) because its longer lines have no fixed medial caesura—caesuras are syntax-determined; and (b) because the continuity of its metres over the line-ending, in regular verse, diminishes the intensity of syntactic/metrical discrepancy in enjambement. My solution to these potential difficulties may seem naïve in the extreme: the use of the slash, apparently to generate junctures artificially, so that, to begin with, my version may seem even more theatrical than Rimbaud’s. But I want to defend this as a possibly extendable practice, indeed to suggest that the unconventional use of punctuation and/or diacritical marks might be a realistic means of capturing the endless diversification of rhythmic experience (I supply a metrico-rhythmic notation alongside)3:

8Here I develop the following types of slash: 1. between two adjacent stresses (a) syncopated, with a glottal stop; (b) negotiated with a silent off-beat [(x)]. 2. between stressed + weak syllable, or between weak + stressed syllable, or between weak + weak syllable, (a) in phrasal verbs and adjectival phrases, between the verb/adjective and following preposition/adverb; (b) between roots and endings in adjectives/adverbs; (c) between determiners (articles, possessives, demonstratives) and following adjectives.

9The difference between the slash between stresses with a silent off-beat and the slash between stresses without a “mitigating” off-beat corresponds roughly to those enjambements where the rejet occurs after a comma—e.g. de la montagne fière,/Luit ; dans le frais cresson bleu,/Dort—and those where no comma buffers the transition—bouche//ouverte ; enfant//malade ; sa poitrine/Tranquille. With the silent off-beat, the shift from stress to stress is cushioned by a slight increase in the duration of the first syllable, or by a pause (the fresh˄blue˄cress, Two˄red), to allow the stress to play itself out. Without an intervening silent off-beat, the two adjacent stresses create a syncopation, necessitate a glottal stop, express the irresistible momentum of sense.

10Those instances where the slash involves unstressed material are unlikely to provoke any marked events in the vocal tract. They are intended to inject tonal or modal nuancing into enunciation. Radiant/with the mountain’s sun and bubbling/with shafts of light, for example, would, without the pointing of the slash, be flatly descriptive; with the slash placed after the preposition they would be no more than a tease; with the slash where it is, the adjective is invested more deeply, more instrumentally, more dynamically, with the prepositional phrase, as the voice remembers itself and gives more enunciatory relief to the construction, making the adjective more consciously chosen. Where the slash separates a root from an adverbial or adjectival ending—boisterous/ly, grass/y—the slash heightens the expressivity of the ending, its levity, its yielding fluency. The voice, as it were, confers the ending, tunes into and urges its particular expressive qualities.

11Rimbaud has, to all appearances, produced a sonnet that is something of a hybrid: an octave in changing rimes croisées (abab/cdcd), followed by a sestet in a more classical Petrarchan mould (eef/ggf). But it is difficult to identify in it any signs of a structural volta after the octave, other than that already indicated by the movement of the rhyme-scheme itself: the shift from the to-and-fro animation and progression of the quatrains’ rimes croisées to the immobilisations of the sestet’s couplets and the sealing enclosure of the froid/droit rhyme. If, then, this is a poem which does not do full justice to sonnet form, there are perhaps ways in which translation can indicate some of the energies bound up in the abrupt dynamic change from octave to sestet and in the variation of stanzaic equilibrium acted out between quatrains and tercets. For me, the answer lies both in the manipulation of typographic distribution, in the interaction of varied layouts, and in the morphing of rhythmic dominants.

12I have tried to mould the quatrains around the spine of the second margin. The first margin provides activating headwords, while the second margin explores the circumstance those headwords engender. The other, more random margins take the eye further afield, compulsively, digressively even. Abruptly, however, the tri-columnar arrangement of the first tercet, driven by currents of acoustic echo and variation, reveals the complicities and collusions of a tightening, more concerted orchestration of the soldier’s condition. The second tercet then turns to a serene linearity, a restored evenness of repose, in which the easy rhythmic flow of the first two lines suddenly becomes dammed up and turbulent in the stressing and stripped-down notation of the final line.

13Rhythmically, the octave is built on the insidious undercurrent of the antispast (x / / x), announced in A green gulley, driven underground throughout the rest of the first quatrain, to reappear in A young soldier and his neck steeped in, and in the incomplete forms of the antibacchic (/ / x: bare-headed, stretched out in) and the bacchic (x / /: like light rain). In the sestet, the antispast is “reversed” into the choriamb (/ x x /), gradually introduced through iamb and anapaest, and first appearing in taking a nap and immediately taken up in cradle him cos- and Nature he’s cold. In the second tercet, we hear it in (A)sleep in the sun and motionless chest. It reaches a kind of apotheosis in a double-stressed form red holes in his right side (/ / x x / /), perhaps to be seen as a conjoining of bacchic and antibacchic.

14The very process of reading/translating activates the reader, activates the resources of the reading body, and it is up to translation to capture this enrichment of bare text. To ask a polyglot reader what the source text means is not to necessitate the process of translation; it is merely to necessitate the metalanguage of literary criticism. To ask how a source text acts upon a reader, on the other hand, to ask what textual desires, what textual variations, associations, energies, it releases in the reader, does necessitate translation. This process of activation is what the target text is designed to incorporate. A printed source text is an inert thing, outside time and space (on a library shelf), until inhabited by a reader. What the reader brings is a singular life-experience, a body, a paralinguistic behaviour and capacity. Accordingly, our ambition should be a translation which is also a scansion, but a scansion in the broad sense of a project for paralinguistic engagement with a text, and that requires a radical flexibilisation of our typographical and paginal languages, comparable, as already intimated, to the new kinds of notation and musikalische Graphik experimented with by modern avant-garde composers. The slash, as used here, is intended as a modest constituent of that graphics, as something which engages with rhythmic proxemics as part of a new rhythmic language or perception of rhythmic effects.

15My second brief argumentative foray, into the rhythmicity of rhythm as a whole, relates to the translation of A.E. Housman’s “Into my heart an air that kills” (A Shropshire Lad XL (1896), in Housman, 1956: 70):

Into my heart an air that kills / x x / x / x /
From yon far country blows: x / x / x /
What are those blue remembered hills, / x x / x / x /
What spires, what farms are those? x / x / x /

That is the land of lost content, / x x / x / x /
I see it shining plain, x / x / x /
The happy highways where I went x / x / x / x /
And cannot come again. x / x / x /

16The business of translation, particularly in the case of the polyglot reader, is not that of textual transfer but of textual supplementation, in two senses: the translator supplements the source text with the responses of the reader, with the events of the readerly consciousness; and, relatedly, the translator supplements the text with, or restores to the text, what the process of printing removes, namely all those prosodic and paralinguistic characteristics that the text might have had in its composition. A first principle, then, is that literary translation should entail the translation of the linguistic towards the paralinguistic, the textual towards the performative. In the translation of verse, this policy supposes that translation resists the seductions of the through-read line of verse, the reading of the line as printed. The metrico-rhythmic implications of through-reading are little considered, but include the following: through-reading favours the establishment of the metrical integrity of the line (whether that integrity lies in a sequence of repeated feet or in a certain number of syllables); it tends to reduce effects of pausing, rupture, juncture, fragmentation, all of which would help to establish the multi-dimensionality of rhythm; it tends to confine pitch to a narrow range and repeatedly to confirm the overall intonational Gestalt of the stanza/rhyme-scheme; it tends to even out durations (both of measures and of lines); it tends to restore cohesiveness to syntax (where it is in the nature of poetic syntax to blur articulations and to generate autonomies in individual syntactic segments); and consequently it tends to obstruct the release of the materiality of the acoustic, to re-integrate the phonetic into the phonological.

  • 4 . The life of the ballad—as opposed to the range of fixed-form medieval ballades, employed by Vill (...)

17Housman’s poem of exile from childhood, or from a previous life, or, quite simply, from Shropshire, acts as a commemorative trigger (through rhyme and other kinds of acoustic recurrence), like a photograph, holding memory in amber, while, at the same time, condemning event to image, the existentially active to an irretrievable past. As the second stanza makes clear, the plainer this land is, the more irrecoverable it becomes. The poem is written in ballad4 metre (alternating iambic tetrameters and trimeters), in what one might think of as its lyric (abab) rather than its narrative (xaxa) form. In this particular instance, Housman uses a so-called reversed first foot in the tetrameters of lines 1, 3, and 5, to produce a small turbulence, an affective fluctuation, an expectancy perhaps, which is as if cancelled by the impassive, unyielding iambics of the trimetric “responses”. It is not surprising that the final tetrameter—line 7—should surrender to that iambic foregoneness, even though it speaks of happiness recalled.

18The poignancies of the poem lie in the expressive nuances of linguistic use. The plurals suggest both abundance and elusiveness, both rich multiplicity and imminent dispersal; there is a systematic unstressing of the first-person forms (my, I (twice)), so that the subject is patient rather than agent, unable to affirm himself; the sense of a “present” topography, located and perceived in real space (Shropshire), created by the demonstratives (yon, those, that) and the as if directly observed aerial perspective (blue), is undermined by the failure to assign identity, expressed in the blunt questions (What): the hills, spires and farms are remembered but not remembered, because the question is what not where.

19Each reader will find his/her own particular way through the phonetic material. My own oral/aural experience begins in the vertebral [a:] of the first stanza (heart, far, are, farms, are). The post-vocalic [m] of farm takes me back to remembered ([rimembɘd]), to the different acoustic values of pre-vocalic and post-vocalic [m] in the notion of memory (pre-vocalic = labial-projective, post-vocalic = labial-assimilative + more voiced, more deeply embedded), so that the post-vocalic [m] of come ([kʌm]) has more piquancy. The second stanza favours alliterative pairing (land / lost; happy / highways; where / went; cannot / come) which helps project a sense of self-fulfilling destiny, of pre-ordained collocation, as the poem falls back into the somnambulistic.

20What is the French translator to do with ballad metre? Marc Porée’s answer (Bensimon et al., 2005: 1091) is, ostensibly, alternating decasyllables and octosyllables, unrhymed. If Housman’s poem has line-initial moments of affective fluctuation (reversed first foot), the metrico-rhythmic dynamic of Porée’s version is more resourceful in its indeterminacies:

Il souffle dans mon cœur un vent qui tue, 2>4>2>2 (6//4)
Venu d’une contrée lointaine. 2>4>2
Ces collines bleues dont je me souviens, 3>2>5 (5//5)
Ces flèches, ces fermes, que sont-elles ? 3’>3’>3

Elles sont du pays du bonheur perdu, 3>3>3>2 (6//5)
Et je revois sa splendeur simple : 4>4
Les grand-routes où je passais, heureux 4’>4>/2 (4//6)
Et où je ne puis revenir. 5>3

  • 5 . But punctuation is not quite so quickly to be dismissed. Porée has converted Housman’s colon at (...)

21As the metrico-rhythmic tabulation reveals, line 4 is an enneasyllable and line 5 a hendecasyllable: it is as if, at this stanzaic crossing, at this question > answer bridge, the poet has been overcome both by a certain loss of articulatory control and by a blurring of formal outline (the impair). My scansion heightens this loss of articulatory control by reading the mute e’s of flèches and fermes not with the fluent, cohesive coupes enjambantes of previous e’s (Il souff/le dans […]; Ces colli/nes bleues […]), but with halting coupes lyriques, interruptive moments of puzzled self-engrossment. And I would want to drive home this reversal of fortune, from fluent memory to halting memory, by resisting liaisons and cultivating hiatuses in line 7 (Les grand-routes/où je passais,/heureux), while letting the perfectly regular octosyllable of line 8 flow on as it wants. But one should beware of falling into any easy developmental account: the decasyllable, even when “regular”, cannot make up its mind about its formal affiliations—the classical 4//6 of line 7 depends on a césure lyrique, and the classical alternative 6//4 of line 1, quickly gives way to a romantic/post-romantic 5//5 (which includes a ‘tolerant’ post-romantic syllabic treatment of bleues, cf. contrée). The whole poem, then is permeated by the problematic, the unstable, the equivocating: Housman’s restrained melancholy finds a much more rhythmically nervous avatar in Porée’s account, despite its being, to my ear, less plotted in its acoustics, and lexically and punctuationally sedate5.

22In my own version, I want to lead the text in two further, but related directions. First, I want to capture the expressive power of the demonstrative that / those. Housman does not insist: those occurs twice, that once; but these demonstratives are supported by yon (line 2). Porée uses ces three times (lines 3-4), without indicating the direction of the deixis (proximity/present (these) v. distance/past (those)). The deixis of the demonstrative can generate immediacy whatever its direction, but while this / these applies an incoming pressure, an intrusiveness, something to be dealt with, that / those injects a hint of the auratic, of existential investment which is an effect of standing away and making a choice. I want, then, a French version which will increase the intensity of the shift from the deixis of proximity (Into my heart) to that of distance, even at the expense of linguistic correctness, though not at the expense of a certain linguistic plausibility. But I have another, larger aspiration: I want to escape the metrical pressure of through-reading. I have seen Porée inject greater metrico-rhythmic uncertainty into Housman’s text, to good effect. I want to go further: employing the concepts of striated and smooth space deriving from Boulez, and from Deleuze and Guattari, I want not only to suggest, uncontroversially I would hope, that while metre belongs to striated space, to pulsed time, to the State, rhythm is a quality of smooth space, of amorphous time, of the nomadic. In developing a smooth-space version of the poem, in a tabular but non-margined layout, I am attempting to restore rhythm to itself and in so doing to liberate Housman more fully into a sub-discursive field of pulsions and erratic energies (indicated crudely in minimal typographic variations—capitals, italics, bold):

23In a sense this tabular layout is a diagram of a mentality, a mentality in disarray, or delivered to its own anarchies. The sober Victorianism of the source text has gradually yielded to another Housman, the other Housman perhaps, described thus by Bernard Brugière: “Face à l’ironie métaphysique, à l’‘iniquité’ du cosmos qui règle les vies individuelles, Housman, pessimiste anglican devenu athée, veut exalter cette joie sauvage et secrète des condamnés, cette ivresse des buveurs de bière, cette allégresse funèbre qui font de ses poèmes autant de pieds de nez au destin” (Bensimon et al., 2005: 1853). The page puts before us the rhythms of a flouted or inverted order, a release into the temperamentally ductile, into impulsive, deconstructive formal vagabondage. At the same time, this is the map of another Shropshire, not the Shropshire of spiritually replenishing landscapes, but a Shropshire that has lost its coordinates, its spatial relations, its settled topography, its very mappability, a Shropshire whose sites and dimensions are floating in free metamorphosis.

24Correspondingly, this layout attempts to undermine a single-minded reading mode, a blind pursuit of linear text until the job of reading is done. Single-minded reading wants to get to the point. But what if there is no getting to the point? In the interests of getting to the point, reading economizes on other awarenesses. At the other extreme is a reading mode without horizon or vanishing point, in which the text is never to be got through, a place of ever-changing positions, relations, motions, the place where text is less communication than an apparatus of self-discovery and self-displacement. The text becomes not its meanings but the qualities of its activity, its dynamics: tempi, durations, accentuations, timbres, and loudnesses, in short, its paralanguage. Language exists to project its paralanguage as its affects, as its capacity to have sense, as a field of forces, rather than vice versa. What then is vital is that the translator should be a stranger in his own language. This is not especially or primarily a “foreignized” language, but a language whose elements (lexemes, morphemes, phonemes, graphemes, typefaces) are maximized in their autonomies so that we experience language as expanding in all directions and in all dimensions (a word I hesitate to use, because in Deleuze and Guattari’s parlance, it is connected with striated space). At the same time, the language of the source text is not equal to, or not coincident with, the responses of the reader, and must accordingly be adapted, deformed, set on a path of variation.

25The avoidance of margins is not just a removal of spatial coordinates; it is a refusal of pulsed time, of measurable segments and breaks, of beginnings and endings, of words and units locked into position. In this kind of disposition, individual elements become vulnerable to slippage and displacement. The reader operates without combinational expectations, is unable to read off something which conforms, is constantly subject to minor re-orientations. One of Deleuze and Guattari’s definitions of smooth space runs as follows: “L’espace lisse est occupé par des événements ou heccéités, beaucoup plus que par des choses formées et perçues. C’est un espace d’affects, plus que de propriétés. C’est une perception haptique, plutôt qu’optique. Alors que dans le strié les formes organisent une matière, dans le lisse des matériaux signalent des forces ou leur servent de symptômes” (1980 : 598). Reading is no longer a recuperative process, the metrico-rhythmic assimilation of a text, where phonatory confirmation accompanies optical retrieval. Indeed, we might propose that this is not at all a shared or shareable text, whose properties are universally perceptible. We each experience it as a movement from point to point, in perpetual close-up, in perpetual encounter, as a process of readerly palpation. In such a text, rhythm can only become, in a condition of permanent futurity; it cannot become what it was ordained to become, in a process of self-fulfilment; it becomes as a direction of being, and of being read.

26What makes rhythm into an event rather than a property of text? Texts have no rhythm as they are, apart from what metre allows them, but this is a rhythm of a mono-dimensional and “syntactic” kind. Texts have no rhythm because rhythm is a negotiation of text with reader, as indeed sense-making is; rhythm comes into existence in the process of reading, as part of the whole-body involvement of the reader. That is why rhythm is multi-dimensional. But as an event, rhythm, even as it takes place, remains only a set of possibilities, possibilities which are projected into new readings and new readers; it never becomes a property of the text. If we take the syllabic values of my translation of the first two lines of Housman’s poem (ci-dedans (3) – profondément (4) – un air mortel (4) – meurtrier (3) – me parvient (3) – de cette contrée (5/4) – lointaine (2)), I might respond to these numbers in the standard way, as “nombres nombrés”, to use Deleuze and Guattari’s term (ibid.: 482-91), numbers that constitute measures, with a metrico-rhythmic objective, transforming Housman’s two lines into three: two heptasyllables followed by a strangely structured decasyllable (supposing the counting of the mute e of cette), or an enneasyllable. But tabular disposition turns us in another direction at the same time, towards smooth space, a space in which these numbers are not measures, that is to say (a) they are not justified by their combinations nor by any sums of syllables they may, in groupings, add up to; the idea of the line has disappeared; (b) they have no given rhythmic status, rather they generate rhythm as a direction in space; (c) they are not matter organized as form; they are materials that make forces manifest, forces operating through them in the rhythmicity of their articulation, as a series. This word rhythmicity expresses a capacity for rhythm, a making possible of rhythm which is the expression of forces. Rhythmicity is an energy which resists all striation, which tends towards uncontrollable multi-dimensionality, towards virtuality, towards continuous variation, even where consecutive numbers are the same. This is a movement not of iteration and reiteration, but of itineration: it is concerned with the discovery of variables rather than of constants, with following continuous singularities: “Il y a bien un rythme mesuré, cadencé qui renvoie à l’écoulement du fleuve entre ses rives ou à la forme d’un espace strié; mais il y a aussi un rythme sans mesure qui renvoie à la fluxion d’un flux, c’est-à-dire à la façon dont un fluide occupe un espace lisse” (ibid.: 450). “La fluxion d’un flux” is the forward-moving, self-shaping dynamic of a flow. Here we are among “nombres nombrants”, autonomous, distributional, without imperatives of combination. They affirm, simply, a syllabic constitution, a potentiality of accentuation, a capacity for affiliation.

27Rhythmicity, then, has two principal fields of meaning. First, rhythmicity extends rhythm in its narrower sense, of patterns of weak and strong syllables. Metre has very little interest in the qualities of stress/weakness, in the multiple variabilities of intensity, pitch, duration, “vocal touch”, attack, tempo. In rhythmic description these qualitative variables are crucial arguments for rhythm’s irreversibility, for rhythm’s tireless itineration, for the view, expressed by D.H. Lawrence, that: “There is no rhythm which returns upon itself, no serpent of eternity with its tail in its mouth” (1967: 87). Second, it is the total force-field of paralinguistic elements and their interactions, which create a text’s overall rhythmic texture. These elements might be isolated, and it might be argued that only a small number impinge on the reading consciousness as determining factors. But there is a danger that we economize on paralinguistic awareness, that we dull the diversity of our response by falling back on age-old and unquestioned priorities (weak and strong syllables). Even if it is impossible, analytically, to capture the play of rhythmic elements in their full complexity, we must continue to try to keep them in consciousness, to mark their presence in scansional translations. Rhythmicity captures this multi-dimensional complexity, this complexity beyond the reach of full critical apprehension, this complexity we can only experience in first-hand sensory contact, in a way that “rhythm” does not. Rhythm, as we have said, is not in the text; it is a negotiation between text and reader, between a text and a psycho-physiological enactment. Rhythm is the present of the performance of text, whether read aloud or inwardly phonated or enacted; but since it is not a given, but a produced, or producible, it constantly maintains a virtuality. It captures the reader’s self-insertion into the text, the dynamic of a self captured in the dynamic of a discourse, for ever changing, never the same. Rhythm is thus the dynamic of the production of sense, the sense of self-text. Since rhythm, unlike metre, cannot exist prior to its manifestation in a reader, since it is always an immanence, so it serves to guarantee and safeguard the orality and the historicity of the reading subject. The dangers of the word “rhythm”, as at present used, lie in its implication of something resolved, something chosen, the end of ambiguities, performative specificity. Translators, for example, often speak of trying to capture the rhythms of the source text. But rhythm while it may indeed become manifest, does not resolve itself. It may take on particular forms, but only on the understanding that it presupposes other forms, that those other forms accompany, are the polyphony of, the manifest form. Rhythmicity is this spray of latent rhythms alongside the ostensible rhythm, this multiplicity and variability of paralinguistic quality, and it is translation’s business to incorporate and communicate this rhythmicity. This is one of the reasons why I reject the notion of an optimal translation of a text, a fairest of fair copies: because translation deals with all that is unresolved in a text, because translation wishes to keep alive this principle of continuous variation; accordingly, the translation of a text must be always ongoing, tirelessly multiplying. Our own trajectory for translation, therefore, would be the translation out of metre into rhythm, and, beyond rhythm, towards rhythmicity.

28Why should translation do this? Because translation translates readers not texts. Because it is readers who are the agents of textual proliferation, of intercultural connections, not texts. Translation takes readers into account and develops forms of reading which are not about the consumption of texts, but about the distribution and radiation of texts into the spaces of linguistic apprehension. The translator’s task is to release the work from its pre-ordainment, from its own patterns of expectation, from its sense of having constantly to come to itself. This is not territorial space, not a space that the very act of translation reterritorialises; it is the constantly self-deterritorialising “territory” of the nomad: “Si le nomade peut être appelé le Déterritorialisé par excellence, c’est justement parce que la reterritorialisation ne se fait pas après comme chez le migrant, ni sur autre chose [régime de propriété, appareil d’État] comme chez le sédentaire […]. Pour le nomade, au contraire, c’est la déterritorialisation qui constitue le rapport à la terre, si bien qu’il se reterritorialise sur la déterritorialisation même.” (Deleuze and Guattari, 1980: 473)

29My final version, then, is designed to capture the rhythmicity, the paralinguistic values, released by the text:

  • 6 . I should emphasize that my scansion presupposes that every reader of the text will develop their (...)

30These values include different kinds of enunciatory momentum and obstruction (► ◄), different tempi (spacing), pauses and intervals of different duration and intensity (▐ ▀ ◊ √), different pitches and pitch-impulses (▲ ▼), different amplitudes (italic, bold, capitals), different degrees of resolution in visual images (░ ▒ ▓), different senses of spatial direction and pathway (╬ ╣ ╦), different experiences of expansion and enclosure (╝ ╗ ( )), different kinds of reflective digestion (: ! , ; - «). In other words, this visual paralanguage covers both vocal events and psychic and cognitive associations. It tells us more about the forces at work at different points in the text than about their specific operation; scansional translation rocks between record and proposition, accuracy and indeterminacy, actuality and virtuality6.

31Let us remind ourselves: we are translating for a polyglot reader; our function as translators is not to provide that reader with a more or less reliable version of the source text, but, as argued before, to activate, and inscribe in the target text, the responses provoked by a reading of the source text, and, at the same time, by re-imagining the source text in the target text, to project it into its possible futures. But such a translation is also designed to engage its readers in their own adventures with the source text; the target text becomes that instrument through which other (polyglot) readers undergo a confrontation with the source text. And in order that all these functions can be fulfilled, the target text must be left in a state of becoming, unfulfilled, a play of directions, captured at the heart of its own dynamic, its own unfolding performance, having thrown off the inertiality of pure text; in this sense, the reader of the target text does not so much receive it as join it, as participate in its ongoingness. Rhythmicity is designed to capture this restless and proliferating activity, this urgent animation that translation restores to the source text.

Haut de page


Bensimon, Paul, Brugière, Bernard, Piquet, François and Remy, Michel, 2005, Anthologie bilingue de la poésie anglaise, Paris, Gallimard.

Boulez, Pierre, 1966, Relevés d’apprenti, éd. Paule Thévenin, Paris, Seuil.

Deleuze, Gilles and Guattari, Félix, 1980, Mille plateaux : capitalisme et schizophrénie 2, Paris, Minuit.

Housman, A.E., 1956, Collected Poems, intro. John Sparrow, Harmondsworth, Penguin.

Hugo, Victor, 1964, Œuvres poétiques I : Avant l’exil 1802-1851, éd. Pierre Albouy, préf. Gaëton Picon, Paris, Gallimard.

Lawrence, D.H., 1967, Selected Literary Criticism, ed. Anthony Beal, London, Heinemann.

Rimbaud, Arthur, 2009, Œuvres complètes, éd. André Guyaux and Aurélia Cervoni, Paris, Gallimard.

Scott, Clive, 2000, Translating Baudelaire, Exeter, University of Exeter Press.

—, 2012a, Literary Translation and the Rediscovery of Reading, Cambridge, Cambridge University Press.

—, 2012b, Translating the Perception of Text: Literary Translation and Phenomenology, Oxford, Legenda.

Haut de page


1 . An apostrophe after the number indicates a coupe lyrique (coupe after word-terminal counted e), in place of a coupe enjambante (coupe before syllable containing word-terminal counted e).

2 . It should be observed that Rimbaud further heightens our sense of sudden shifts in consciousness or perceptual acuity by his rather idiosyncratic use of colons in lines 4, 10 and 11.

3 . I realize that, in providing this notation, I am in danger of undermining my own argument and of implying that the metrico-rhythmic basis of rhythm remains paramount. It certainly remains important, but (a) it is not to be understood as the restoration of a metricality, but as a demetrified metrics, and (b) it is to be urgently supplemented by other factors of rhythmicity, as the article will go on to show.

4 . The life of the ballad—as opposed to the range of fixed-form medieval ballades, employed by Villon among others—has, in France, been a tenuous one, and dates from Hugo’s Odes et Ballades (1826, 1828). The Dictionnaire de l’Académie accepted “ballade” in the sense of “ballad” only in 1838. Hugo’s ballads clearly bear the marks of the English and German pre-Romantic and Romantic ballad in their choice and treatment of subject: the folkish and legendary, scenes, dreams and tableaux, the exploitation of the gothic, the performative presence of a balladeer. But despite a certain favouring of a five-line (abaab) and six-line (aabccb) stanza, Hugo makes no attempt to establish a standard generical form; in fact, he uses the ballad, if anything, as a pretext for exercises in formal virtuosity. Furthermore, in his preface of 1826, he confirms his own rather cavalier attitude: “Au reste, il [l’auteur] n’attache pas à ces classifications plus d’importance qu’elles n’en méritent. Beaucoup de personnes, dont l’opinion est grave, ont dit que ses Odes n’étaient pas des odes; soit. Beaucoup d’autres diront sans doute, avec non moins de raison, que ses Ballades ne sont pas des ballades; passe encore. Qu’on leur donne tel autre titre qu’on voudra; l’auteur s’y souscrit d’avance” (1964: 279).

5 . But punctuation is not quite so quickly to be dismissed. Porée has converted Housman’s colon at the end of line 2 into a full-stop and thus removed the revelatory force, the sudden urgent focus, that Housman aspires to; Porée’s lines maintain a more uniform perceptual consciousness. In fact, Porée seems to have postponed the colon to what might have seemed a more apposite position, the end of line 6, where, too late and only momentarily, the splendeur simple is translated, transformed, into happy highways, before being nullified by the final line. Housman’s punctuation encourages us to read line 6 not as a threshold of reawakened vision, but as a parenthesis, an aside, before the land of lost content is appositionally defined as The happy highways. Punctuation has, of course, consequences both for intonation and for syllabic durations: the colon will produce suspensions of greater intensity, the voice being pitched upwards, towards what follows, and this may help the poem to break free from the presiding patterns of stanzaic intonation. But these latter will be the more confirmed, the more we through-read, and for that reason we may wish to realize Porée’s line 6 as: 4>3>1 and read the line-ending something like: splendeur/ SIM…/ples les grand-routes, that is, with a greater withholding and more subsequent acceleration.

6 . I should emphasize that my scansion presupposes that every reader of the text will develop their own paralinguistic sensitivities, and thus their own scansional languages. In this sense, a translation for the polyglot reader is never more than a suggested model, an exemplar, something to be infinitely adapted to different needs.

Haut de page

Pour citer cet article

Référence papier

Clive Scott, « Translation and the expansion of the rhythmic sense », Palimpsestes, 27 | 2014, 219-237.

Référence électronique

Clive Scott, « Translation and the expansion of the rhythmic sense », Palimpsestes [En ligne], 27 | 2014, mis en ligne le 01 décembre 2016, consulté le 22 avril 2018. URL : ; DOI : 10.4000/palimpsestes.2116

Haut de page


Clive Scott

Clive Scott est professeur émérite de littérature européenne à l’université East Anglia et Fellow de la British Academy. Parmi ses publications récentes, on compte : Literary Translation and the Rediscovery of Reading (Cambridge UP, 2012) et Translating the Perception of Text: Literary Translation and Phenomenology (Legenda, 2012).

Haut de page

Droits d’auteur

Tous droits réservés

Haut de page
  • Logo Presses Sorbonne Nouvelle
  • OpenEdition Journals