Navigation – Plan du site
I. Traduire le rythme du poème
Du mètre et du rythme en traduction

“Cries and cracks”: “animal” rhythms in Ezra Pound’s translations from Arnaut Daniel

Andrew Eastman
p. 63-79


Cet article propose une lecture de la chanson « L’aura amara » du troubadour Arnaut Daniel et de la traduction qu’en a proposé Ezra Pound en 1917, afin d’étudier ce que la pratique de la traduction chez Pound indique de sa conception du rythme. Selon Pound, l’intérêt principal des chansons de Daniel réside dans leur organisation sonore, plus spécifiquement dans l’onomatopée et l’imitation de chants d’oiseaux ; cette manière de voir s’accorde avec certaines remarques dans lesquelles Pound envisage le rythme comme une manifestation physiologique. Cependant, Pound voit aussi le rythme par le prisme de la traduction, suggérant par exemple qu’il faut chercher les rythmes du vers anglais dans les cadences de poèmes étrangers. L’article envisage ainsi la manière dont la traduction de Pound se construit à partir de l’organisation prosodique de la canso d’Arnaut pour déformer et spécifier les rythmes de l’anglais. On cherchera à montrer que le rythme chez Arnaut peut se décrire comme l’interaction entre la composition phonologique, l’accent, et la syntaxe ; et que la traduction de Pound développe plusieurs aspects de l’écriture de ce troubadour, le vers court, les enjambements et le traitement rythmique et prosodique du pronom mi, de manière à faire de sa traduction une recherche prosodique inséparable du développement du vers libre. Introduisant dans sa version de « L’aura amara » des rythmes peu « naturels », ambigus, Pound se sert du discours de l’amour courtois pour inventer un sujet qui est autre à lui-même, un sujet traduisant.

Haut de page

Texte intégral

  • 1 . Dante Alighieri, La Divina Commedia, testo critico stabilito da Giorgio Petrocchi, Turin, Einaud (...)
  • 2 . “Translating rhythm” might be understood to suggest that rhythm is one problem among others face (...)

1“I’m true, so dree / Fates” writes Ezra Pound in the third stanza (lines 42-43) of his translation of Arnaut Daniel’s canso or song, “L’aura amara”, a poem written in Provençal in the late XIIth century and praised by Dante in his De Vulgari Eloquio. Pound’s translation, completed in 1917, grew out of his long preoccupation with the troubadour whom Dante called “miglior fabbro del parlar materno1” (Purgatorio, canto 26, line 117), the finest craftsman in the mother tongue. The passage I have just quoted is, for many English-speakers of today and perhaps also of 1917, to all intents and purposes unintelligible, until, consulting the dictionary, we learn that dree is a Scots form meaning “to endure, to suffer” (“to dree one’s weird” is to suffer one’s fate). Even then, however, we are not sure whether to read dree / Fates as narrating the speaker’s condition, glossing “and so I suffer the Fates” (with dree a first-person indicative), or as a challenge issued to the fates—“so suffer, Fates” (with dree as imperative). The obscurity brings us no closer to the medieval text; but Pound’s translation is not preoccupied with lexical meaning merely. The quasi-alliteration linking true and dree, the very uncertainty of the syntax, the odd word’s lexical salience, and its position at the rhyme may well lead the reader to give special prominence to dree, though it functions, at least in one reading, as a transitive verb, and as such, is subject to de-accenting or demotion; dree / Fates then introduces a “stress-clash”, two juxtaposed prominences, a disruption of the regular alternation which linguists tell us governs rhythm in English. Pound’s lexical “choices” in “L’aura amara”, I would like to show, are inseparable from the organization of movement and energy in his language, from a continuum of speech sound and syntax that, following Henri Meschonnic, one would like to call “rhythm2”.

2Pound’s use of obscure words in his Arnaut Daniel translations was disliked by critics concerned with “exact” rendition of lexical meaning. Dudley Fitts argued that Arnaut wrote a “living language”, direct and natural, for which Pound’s archaisms could provide no equivalent (quoted in Venuti, 2008: 185); Mark Smith-Soto, in a generally favorable article from 1984, noted the “frequent obscurity and literal inaccuracy” of Pound’s translations, and singled out “L’aura amara”, which, he argued, could “barely be deciphered” (Smith-Soto, 1984: 20). Yet Pound’s interest in Daniel’s work stemmed from the perception that he had “tried to make almost a new language” (Pound, 1968 [1954]: 112) and Linda Paterson, in her Troubadours and Eloquence, seems to bear this out; she argues that the troubadors’ trobar ric, or “rich elaborate style” is best exemplified in Arnaut’s poetry, and that one of its finest examples is “L’aura amara” (Paterson, 1975: 202). Lawrence Venuti took a different approach, seeing in the mixture of archaic, dialectical, literary and colloquial diction of Pound’s versions an example of “foreignizing” translation, translation that calls attention to itself as translation; the Arnaut Daniel translations were Pound’s “most experimental” because in them Pound produced “his most heterogeneous discourses” (Venuti, 2008: 173). Heterogeneity, however, seems like a problematic criterion for the value of a translation; particularly for the translation of a poem in which, as Paterson argues, “[a]ll aspects of the style and theme emphasize compactness, intensity, and concentration” (Paterson, 1975: 203); though Venuti emphasizes “the play of the signifier” in Pound’s translations (Venuti, 2008: 177), his approach can tell us little about how Pound translates rhythm as continuum.

  • 3 . Charlotte Ward’s “Notes to the translations” give numerous examples of what she calls “sound tra (...)

3The obscurity of Pound’s translations derives, it would seem, from a close attention to the speech sounds of the Provençal text3. In order to give a better idea of what he is doing, I quote here “L’aura amara” and Pound’s version of it.

  • 4 . In Translations Pound prints the Provençal text from U. A. Canello’s edition (1885); according t (...)

L’aura amara
Fals bruoills brancutz
Quel doutz espaissa ab fuoills,

5 Els letz
Dels auzels ramencs
Ten balps e mutz,

10 E non-pars;
Per qu’eu m’esfortz
De far e dir
A mains per liei

15 Que m’a virat bas d’aut,
Don tem morir
Sils sfans no m’asoma.

The bitter air
Strips panoply
From trees
Where softer winds set leaves,

5 And glad
Now in brakes are coy,
Scarce peep the wee

10 And un-mates.
What gaud’s the work?
What good the glees?
What curse
I strive to shake!

15 Me hath she cast from high,
In fell disease
I lie, and deathly fearing.

Tant fo clara
Ma prima lutz

20 D’eslir
Lieis don crel cors los huoills,
Non pretz
Mans dos aigonencs;

25 D’autra s’ esdutz
Mos preiars,
Pero deportz
M’es adauzir

30 Volers,
Bos motz ses grei
De liei don tant m’azaut
Qu’al sieu servir
Sui del pe tro c’al coma.

So clear the flare
That first lit me

20 To seize
Her whom my soul believes;
If cad
Blabs, slanders, my joy

25 Counts little fee
And their hates.
I scorn their perk$ And preen, at ease.

30 Disburse
Can she, and wake
Such firm delights, that I
Am hers, froth, lees
Bigod! from toe to earring.

35 Amors, gara,
Sui ben vengutz
Tem far sim desacuoills
Tels detz

40 Pecs
Que t’es mieills quet trencs;
Qu’ieu soi fis drutz
E non vars,

45 Mal cors ferms fortz
Mi fai cobrir
Mainz vers;
C’ab tot lo nei
M’agr’ops us bais al chaut

50 Cor refrezir,
Que noi val autra goma.

35 Amor, look yare!
Know certainly
The keys:
How she thy suit receives;
Nor add

40 Piques,
’Twere folly to annoy.
I’m true, so dree
No debates

45 Shake me, nor jerk.
My verities
Turn terse,
And yet I ache;
Her lips, not snows that fly

50 Have potencies
To slake, to cool my searing.

Si m’ampara
Cill cuim trabutz

55 Si qu’es de pretz capduoills,
Dels quetz
C’ai dedinz a rencs,
L’er fort rendutz

60 Clars
Mos pensars;
Qu’eu fora mortz,
Mas fam sofrir

65 Queill prec quem brei,
C’aissom ten let e baut;
Que d’als jauzir
Nom val jois une poma.

Behold my prayer,
(Or company
Of these)

55 Seeks whom such height achieves;
Well clad
Her, and would not cloy.
Heart apertly

60 States
Thought. Hope waits
’Gainst death to irk:
False brevities
And worse!

65 To her I raik,
Sole her; all others’ dry
I count not worth the leering.

Doussa car’, a

70 Totz aips volgutz,
M’er per vos mainz orguoills,
Car etz

75 De totz mos fadencs,
Don ai mainz brutz
E gabars;
De vos nom tortz

80 Nim fai partir
C’anc non amei
Ren tan ab meins d’ ufaut,
Anz vos desir

85 Plus que Dieus cill de Doma.

Ah, fair face, where

70 Each quality
But frees
One pride-shaft more, that cleaves
Me; mad frieks
(O’ thy beck) destroy,

75 And mockery
Me, and rates.
Yet I not shirk
Thy velleities,

80 Averse
Me not, nor slake
Desire. God draws not nigh
To Dome, with pleas
Wherein’s so little veering.

Erat para,
Chans e condutz,
Al rei qui t’er escuoills;

90 Car pretz,
Sai, lai es doblencs,
E mantengutz

95 E manjars:
De joi lat portz,
Son anel mir,
Sil ders,
C’anc non estei

100 Jorn d’Aragon quel saut
Noi volgues ir,
Mas sai m’a’ n clamat Roma.

Faitz es l’acortz
Qu’el cor remir

105 Totz sers
Lieis cui domnei
Ses parsonier Arnaut;
Qu’en autr’ albir
N’es fort m’ententa soma.

85 Now chant prepare,
And melody
To please
The king, who’ll judge thy sheaves.
Worth, sad,

90 Sneaks
Here; double employ
Hath there. Get thee
Full, and cates,

95 Gifts, go! Nor lurk
Here till decrees
And ring thou take.
Straight t’ Arago I’d ply

100 Cross the wide seas
But ‘Rome’ disturbs my hearing.

At midnight mirk,
In secrecies
I nurse
105 My served make
In heart; nor try
My melodies
At other’s door nor mearing.

(Pound, 1970 [1953]: 160-1674)

  • 5 . Translations from Provençal into English are by me, adapted from the versions by Canello (1883) (...)

4When we read Pound’s “L’aura amara” against Arnaut Daniel’s poem, we “see” an aspect of what he is doing: I’m true, so dree / Fates (lines 42-43) translates, apparently, Qu’ieu soi fis drutz, / Cars, which might be rendered “For I am a sincere lover, / And precious5”. There is more involved, however, for fis is the subject case of fin, an essential ethical term of courtly love, in Provençal “ fin’amor”; while drutz, likewise in subject case, is glossed “amant, fidèle, ami privé” by Emil Lévy’s dictionary. Pound’s true, dree and Fates thus recombine the corresponding English phonemes of fis drutz; true and dree echo each other like kindred expressions, like cognates, and together serve to redefine the tension or contradiction between fidelity and suffering at the heart of “fin’amor”. We are led, then, to read Pound’s translation as a “registration” of the speech sounds of Arnaut’s poem; yet this attention to speech sound also makes possible an invention of ways of saying, ways of “signifiering” one would like to say, ways of evoking meaning by associating signifiers: the translation of fis drutz by I’m true, so dree functions as an interlingual pun, and thus responds to the paronomasias which saturate Arnaut’s writing, and which can be seen at work in the opening line of his song.

5Pound’s interest in Arnaut’s poems appears to reside in their “orality” as Meschonnic defines it (1985: 125) — in the fact that they make speech sound organization a primary means of signifying. Provençal poetry, and Arnaut’s work in particular, is for Pound the prime example of “melopoeia”, the process “wherein the words are charged, over and above their plain meaning, with some musical property, which directs the bearing and trend of that meaning” (Pound, 1968 [1954]: 25). Melopoeia is to be found in the complex rhyme scheme of “L’aura amara”, apparently unexampled in Provençal poetry: rhymes make up almost one third of the syllables in each 17-line verse (19 out of 62). This scheme, Paterson notes, is a variant of the rimas escarsas commonly adopted by Arnaut, in which words rhyme not within a stanza but from stanza to stanza; in “L’aura amara” there are also five stanza-internal rhymes. Pound’s version of “L’aura amara”, which follows Arnaut’s rhyme pattern almost exactly (the initial feminine rhyme is made masculine) is then a translational tour de force, an attempt to render what he calls the “system of echoes and blending” (ibid.: 115) which specifies Arnaut’s writing. Yet Pound’s comments on “L’aura amara” appear to focus on another use of speech sound, as imitation of sounds in nature, “onomatopoeia”: in ABC of Reading, “L’aura amara” becomes the prime example of Provençal melopoeia because Arnaut “made the birds sing IN HIS WORDS” (Pound, 1934: 53).

6Pound’s focus on “onomatopoeia” is consistent with his poetics, and with the conception of rhythm in his poetics, if we follow the analysis of Michael Golston in his recent book on modernist conceptions of rhythm, Rhythm and Race in Modernist Poetry and Science. Birdsong appears elsewhere in Pound’s work as a figure for nature in art; and the function and value of rhythm in the poem depend, argues Golston, on Pound’s sense of rhythm as “primal”. Pound theorized an “absolute rhythm” in poetry, which he explained, in 1913, as “a rhythm […] which corresponds exactly to the emotion or shade of emotion to be expressed” (Pound, 1968 [1954]: 9). Like contemporary champions of vers-libre, Pound sees rhythm as a mode of individuation in language: “A man’s rhythm”, he writes, “will be […] his own, uncounterfeited, uncounterfeitable” (ibid.); “something”, as Meredith Martin notes, in her study The Rise and Fall of Meter, “that begins and ends with each poem” (Martin, 2012: 181). Golston, however, reads the relation between rhythm and emotion as implying a relation between rhythm and the body. In his account, Pound sees rhythm in the poem as a “registration” of the body’s physiology, analogous to the registrations of rhythm made by early-XXth-century experimental phonetics (Golston, 2008: 98), which, as he points out, Pound held to be central for the development of modernist poetic practices of rhythm. Pound, notes Golston, describes poetry as a “centaur”, in which the “thinking word-arranging clarifying faculty” must “move and leap” with “the energizing, sentient, musical faculties” (ibid.: 108); and this image suggests, according to Golston, that rhythm engages with the “irrational beast” (ibid.: 109), with instinct and the unconscious, and thus with an identity which is at once personal and racial. For Pound, he claims, “[t]he rhythms of poetry index not merely the private emotions of the poet but the deeper, historically determined instincts that structure the collective unconscious of the race” (ibid.: 113). Pound’s rhythms, identified with the racialized body, are then “animal rhythms”—inscriptions of inherent nature.

7Yet Pound’s conception of rhythm cannot be fully understood without taking translation into account. Translation for Pound is an invention, in the translating language: “is a fine poet ever translated,” he asked, “until another his equal invents a new style in a later language?”—and this formulation implies the paradox that we can “really only know Ovid”, not in the original text, but through the work of his translator Golding (Pound, 1968 [1954]: 235). Rhythm is central to Pound’s translating projects. The concept of “absolute rhythm” was first formulated in the preface to his translations of the Italian medieval poet Guido Cavalcanti, published in 1910 (Pound, 1970 [1953]: 25). “Absolute rhythm” is infinite variety of rhythm, and implies a critique of the metrical tradition; what is at stake, for Pound, in the translation of medieval lyrics is the rhythmic variety of poems composed to music, the way poetic line and musical phrase interact; the sonnet, from this point of view, brought on a divorce of words and music, and Pound condemns “the monotony of the 14 even lines as compared to the constantly varying strophes of Ventadour or of Arnaut” (Pound, 1968 [1954]: 170): translating Arnaut is thus a problem of rhythm, inseparable from the problem of how to write now. Rhythm, Pound writes, in his advice to poets of 1913, is to be found in the foreign: the “candidate”, or apprentice poet, should “fill his mind with the finest cadences he can discover, preferably in a foreign language” (ibid.: 5); this, presumably, because “English verse” as he later noted in a discussion of Laurence Binyon’s translation of Dante, lacks rhythmic definition, is not “CUT” (1968 [1954]: 205). Yet, rhythm, as “melopoeia”, is “practically impossible to transfer or translate […] from one language to another” says Pound in “How to Read” (ibid.: 25). Translating poetry will then depend on bringing foreign rhythm into English by “foreignizing” English. The expression “[m]ade the birds sing IN HIS WORDS” might then be understood as a metaphor for translation itself, and for rhythm in translation as a form of otherness, not the emanation, index, or imprint of self, but the process of othering self.

8How, then, is rhythm to be heard, and translated, in “L’aura amara”? As Hugh Kenner noted, the “birdsong” Pound refers to is not “mellifluousness” (Kenner, 1971: 87), but rather a kind of harshness, “the chatter of birds” rendered through “the ‘-utz, -etz, -encs, and -ortz’ of the rhyme-scheme” (Pound, 1970 [1953]: 127); similarly, Pound translates the beginning of Arnaut’s “Doutz brais e critz”, which he likewise associates with birdsong (Pound, 1968 [1954]: 109) as “Sweet cries and cracks” (Pound, 1970 [1953]: 173). Bird cry, or chatter, would seem to be a way of talking about the general movement of “L’aura amara”, which depends, crucially, on the abruptness of its short lines. The poem is composed of a variety of short lines, disposed erratically: there are two lines of one syllable, three of two syllables, one of three syllables, six lines of four syllables, two of five and three of six. These may be seen in relation with the breaking off of birdsong in autumn (Els letz / Becs / Dels auzels ramencs / Ten balps e mutz, l.5-8; “and makes the joyous beaks of the birds in the branches stammer and go silent”) and with the speaker’s attempt to remedy it through his own song (Per qu’eu m’esfortz, l.11; “which is why I strive”). The poem dramatizes the effort or tension towards speech, the difficulty of speech itself. Pound’s introduction to his translation describes “L’aura amara” as “clear sound, with staccato” (Pound, 1968 [1954]: 114); his application of this musical term is perhaps explained in his discussion of Laurence Binyon’s translation of Dante, where he noted: “The common verse of Britain from 1890 to 1910 was a horrible agglomerate compost, not minted, most of it not even baked, all legato” (ibid.: 205); in contrast to which, Binyon’s translation has “vast numbers of mono-syllables” (ibid.). “Staccato”, then, would seem to involve the use of short, often monosyllabic tone groups, while “legato” would correspond to the use of longer groups where accents are subordinated to each other, allowing the alternation of prominent and non-prominent. This is to say that Pound’s translation is an experiment in the possibilities of the line, this being a central preoccupation in free verse.

  • 6 . Charles Hartman suggests that the line unit is a way of “increasing the density of accents”, in (...)
  • 7 . The importance of juxtaposed accents in Pound’s poetry is shown for example in Hall, 2003 and Jo (...)

9The shorter lines are conducive to denser accretions of lexical material: Linda Paterson, noting the effect of “concentration” in “L’aura amara” (Paterson, 1975: 205), remarks that “[s]triking sounds are closely packed together by the use of monosyllables” (ibid.: 204). Pound’s version of “L’aura amara” multiplies short tone groups, notably through the introduction of caesurae, mid-line pauses; whereas Arnaut’s poem, in its modern editions, contains 3, 5, or 8 of these, Pound’s translation has 35. The result is a series of monosyllabic tone groups which work against the traditionally alternating movement of the English metrical tradition, as in the translation of lines 22-24, where Arnaut’s Non pretz / Necs / Mas dos aigonencs (“I don’t consider secret messages worth even two wild roses”), is rendered with a sequence where almost every syllable takes an accent: If cád, / Blábs, / Slánders, m´y jóy”. Where they come into conflict with syntactic units, the short lines of “L’aura amara” add accents, both as a mark of line break and as a mark of line onset6; the result is that accents fall where they would not ordinarily. We can see something of this in the way Nor add / Piques, at lines 39-40, invites us to place an accent on add, which might normally be subordinated or de-accented, again giving rise to an accentual “clash”. Pound uses other means, such as phonemic echo, to produce similar effects, as where he writes My verities / Turn terse (l.46-47): the repetition of initial [t], the rich rhyme in Turn terse invite a reader to place an accent on the relatively empty word turn. Enjambment and caesura work together to enforce accent on a word which would not normally take it at l.56 ff., Well clad / Seeks / Her, here investing the “empty” feminine pronoun with a significant emotional charge; this treatment of her is then repeated through the juxtaposed accents of sole her in line 667.

10Enjambment in “L’aura amara” is not simply a way of multiplying accent; Pound, building on constructions in Arnaut’s poem, makes the scissoring of line and syntax a signifier of tension or attraction in a condition of separation. At l.20-21, Arnaut makes use of line end to separate the verb eslir (“elect”, “choose”) and its direct object lieis (“her”), signifiers which are at the same time identified through paronomasia; Pound’s translation of the passage, To seize / Her, introduces a verb denoting movement towards and contact with an object, lending significance to the enjambment. Arnaut’s sequence al sieu servir / Sui (lines 32-33), which might be rendered as “in her service / I am”, once again uses paronomasia to identify forms referring to the persons of the drama, beloved and speaker, Sui (“am”) suggesting that the speaker’s essence or existence is involved in that of the lady; Pound’s translation isolates the nominal form I at line end. At l.57-58, Dels quetz / Precs refers to silent prayers which the speaker hopes the lady will accept; in Pound’s translation, Seeks / Her, based perhaps on a whimsical association of quetz with “quest”, reverses the situation by portraying the prayer in search of the lady. Here again the conflict between syntax and line affects a verb implying direction towards an object, so that the syntactic tension effected by enjambment enters into relation with the emotional tension inherent in the situation. Arnaut’s use of enjambment is most striking in l.69-70, Doussa car’, a / totz aips volgutz (“sweet face, with / all desired qualities”) where the line break separates a preposition from its complement: here a functions both as the elided feminine ending of car, and, it seems, as an interjection, thus calling attention to the repetition of [a] and the rhyme with amara; Pound’s version, Ah, fair face, where supplies both exclamation and enjambment, while making the line break a signifier of the speaker’s disorientation.

11The poem’s and the translation’s dramatization of separation through enjambment works through the grammar of enunciation; rhythm and speech sound are inseparable from the organization of person and pronoun. In Pound’s translation, line-break is a way of isolating first-person pronouns; here me, a grammatical word, is made accentuable. Yet Pound’s treatment of me is attentive, I will argue, to the way Arnaut’s song works with the first-person singular pronouns and determiners mi, mos, ma. The song as a whole may be seen as a development, through phonemic echo, of its opening line, “L’aura amara”; its system of echoes links amara and first-person forms beginning with [m]. In the first stanza, the [ma] sequence in amara is linked to the first-person pronoun in m’asoma in l.17 (“puts an end to for me”), while [a], [au], [m], [r] in “L’aura amara” are echoed in Que m’a virat bas d’aut (“who threw me low from high”) in l.15: the bitterness of the season is made continuous, in Arnaut’s song, with the bitterness of the speaker’s experience. Arnaut’s treatment of mi and other first-person forms with initial [m] functions in relation with the use of short lines in his song; these forms, placed initially, are marked by onset: Ma prima lutz (l.19; “my first light”), Mos preiars (l.27; “my prayer”), M’es adauzir (l.29; “is to me to hear”), Mi fai cobrir (l.46; “makes me hide”), M’agr’ops us bais al chaut / Cor refrezir (l.49; “I would need a kiss to refresh my hot heart”), Mos pensars (l.61; “my thought”), Sofrir / M’er per vos mainz orguoills (l.72; “For you I am to suffer many affronts”). Positioned thus, we find them linked to other morphemes with initial [m], as in stanza 2:

Mos preiars,
Pero deportz
M’es adauzir
Bos motz ses grei
De liei don tant m’azaut (l.27-32)

12(“[For rarely is drawn out by another] my prayer, / yet is it a pleasure / to me to hear / the desires, / the good words without recriminations / of her who charms me so”). Here Mos (“my”) and M’es (“is to me”) are linked to motz (“words”) and m’azaut (“charms me”) through the repetition of the sequence <m-s> and <m-z> (Provençal <tz> is pronounced [ts]), and this series is concluded by Amors (“Love”), another echo of amara, at the beginning of stanza 3. A similar series appears in stanza 4, where Mos pensars (“my thought”) at l.61 is echoed by mortz (“dead”) at l.62 and Mas fam sofrir / L’espers (“but hope makes me put up with it”) at l.63-64. The poem’s intrication of amara, motz, mortz, and Amors weaves into its word skeins the first-person object pronoun and related forms.

13Pound’s translation builds on the system of echoes linking amara and mi in Arnaut’s poem. Here me is both marked and made rhythmically unstable, poses a problem of pronunciation, in that we may hesitate as to whether it is to be realized with a reduced vowel [i], or full vowel [i:]. In the first occurrence, Me she hath cast from high (l.18), which translates Que m’a virat bas d’aut, syntactic inversion and initial placement would tend to accent and lengthen Me, setting up an echo with the glide in high at line end. Pound is here translating the salience which the enclitic pronoun mi acquires through the repetition of <a> in Arnaut’s line. In That first lit me at l.18, the syntax determines a reduced vowel reading (lít me), while the position of the pronoun at the rhyme would lead us to read it as a full vowel (lít mé). At l.44-45, where we read No debates / Shake me, nor jerk, the context suggests a “contrastive accent” on meme as opposed to all others; and at l.72-73, cleaves / Me; the placement of Me at the beginning of the line, followed by a caesura, induces an accent and full vowel pronunciation. Here as well Pound seems to be translating the “rhythm” of Arnaut’s text; in Sofrir / M’er per vos mainz orguoills, M’er is made salient through the syntactic break which follows, and is linked by consonance with Sofrir. Me is again separated by enjambment from its governing verb and placed at the beginning of the line in And mockery / Baits / Me at l.77, and Averse / Me not at l.81. In this last case, the pronoun does not seem to require an accent; but the contextual pressure of the poem’s system of echoes suggests that all occurrences of me should be marked. Pound makes the rhythmics of me central to stanza 5, where for the first time the speaker addresses the lady (Doussa car’, a, l.69). The instability of me in the poem is a function of the speaker’s lack of autonomy; the rhythms of Pound’s translation invent a specific subject, a subject of translation, in which the unstable, dependent position of the speaker-lover shadows that of the translating subject.

14This treatment of me can be found in other translations of Arnaut Daniel. In a version of Arnaut’s “Chanson doill” (poem II) which was not included in Arnaut Daniel, Pound introduces feminine rhymes or quasi-rhymes in the sixth and ninth lines of each verse, rhyming greenly and seemly in the first, turn me and spurn me in the second, screen thee and burn me elsewhere: here the reader is led to de-accent the pronoun, pronounced with a reduced vowel. But in the sixth and last verse, he contravenes this pattern, writing:

  • 8 . We give here the stanza in Lavaud’s edition, and his translation (Lavaud’s italics mark the rhym (...)

Save ’neath Love’s thong I move no thing,
And my way brooks no measuring,
For right hath spring
In that Love’s heat
50 Was ne’er complete
As mine, since Adam. ’Tween me
And sly treason
No net is spun
Wherefore my joy grows greenly. (Translations 417-18)8

15The caesura in line 6 calls for an accent or onset on ’Tween which begins a new sentence, in accordance with the pattern of feminine rhymes running throughout the poem; but the contrast introduced by ’Tween opposing me and sly treason requires an accent on me: the line consequently cannot be read without juxtaposed accents on the two [i:] vowels. What seems interesting about the rhythm of this line is that ’Tween, at first sight an awkward, artificial-sounding aphaeresis (the dropping of a word-initial syllable), becomes, as juxtaposed (and, through accent, identified) with me, a signifier of internal, or internalized, contradiction, of the tension between true and dree suggested in Pound’s “L’aura amara”.

16Because they respond to the close connection between speech sound repetition (phonemic echo) and prominence in Arnaut’s poem, Pound’s translations invite us to pronounce English in ways that appear awkward, distorted, “unnatural”; they put into question the linguistic and aesthetic norm which, in Pound’s day as in ours, sees rhythm in English, and universally, as regular alternation. At the same time, exactly because they function as a continuum, as the multiple interaction of syntax, accent, and speech sound, Pound’s rhythms are unstable and ambiguous, most particularly where the “inscription” of the subject in language is concerned, in the grammar of enunciation: they suggest both that rhythm cannot be seen as a “registration” of physiology, and that “I” is not an essence or nature. The “me” or “I” they make is, then, a variable of the poem’s re-enunciation; the subject they define is an estrangement of self, and as such specifically a translating subject. It seems appropriate, then, that Pound closes his translation with the obscure word mearing, cited in its participle form: “mear”, absent from my Webster’s Collegiate Dictionary, is defined by the OED as “to abut upon, be bounded by”; the construction “nor try / My melodies / At other’s door not mearing” seems then to suggest that the speaker’s songs have no connection with other women; but also, that his melodies, and by extension, Pound’s translations, are an activity of “I-making”, and that this process is unbounded, “unmeared”.

Haut de page


Corpus texts

Daniel, Arnaut, 1883, La Vita e le opere del trovatore Arnaldo Daniello, ed. by U. A. Canello, Halle, Max Niemeyer.

—, 1910, Les poésies d’Arnaut Daniel, réédition critique d’après Canello, ed. by René Lavaud, Toulouse, Édouard Privat.

—, 1960, Canzoni, ed. by Giovanni Toja, Florence, G.C. Sansoni.

Pound, Ezra, 1970 [1953], Translations, London, Faber and Faber.

Works cited

Dessons, Gérard and Meschonnic, Henri, 1998, Traité du rythme. Des vers et des proses, Paris, Dunod.

Golston, Michael, 2008, Rhythm and Race in Modernist Poetry and Science, New York, Columbia University Press.

Hall, Donald, 2003, “Pound’s Sounds”, in Breakfast Served Any Time All Day: Essays on Poetry New and Selected, Ann Arbor, University of Michigan Press.

Hartman, Charles O., 1996 [1980], Free Verse: An Essay on Prosody, Evanston, Illinois, Northwestern University Press.

Jones, Chris, 2006, Strange Likeness, The Use of Old English in Twentieth-Century Poetry, Oxford, Oxford University Press.

Kenner, Hugh, 1971, The Pound Era, Berkeley, University of California Press.

Lévy, Emil, 1909, Petit dictionnaire provençal-français, Heidelberg, Carl Winters Universitätsbuchhandlung.

Makin, Peter, 1978, Provence and Pound, Berkeley / Los Angeles, University of California Press.

Martin, Meredith, 2012, The Rise and Fall of Meter, Princeton, Princeton University Press.

Meschonnic, Henri, 1977, Écrire Hugo, vol. I, Paris, Gallimard.

—, 1985, Les États de la poétique, Paris, Presses universitaires de France.

—, 2007, Éthique et poétique du traduire, Paris, Verdier.

Paterson, Linda, 1975, Troubadours and Eloquence, Oxford, Clarendon Press.

Pound, Ezra, 1934, ABC of Reading, New York, New Directions.

—, 1968 [1954], Literary Essays, New York, New Directions.

—, 2003, Poems and Translations, ed. by Richard Sieburth, New York, Library of America.

Smith-Soto, Mark I., 1984, “Ezra Pound’s Map of Arnaut Daniel”, Comparatist, 8, 14-20.

Venuti, Lawrence, 2008, The Translator’s Invisibility: A History of Translation, 2nd edition, London, Routledge.

Ward, Charlotte, 1991, Pound’s Translations of Arnaut Daniel, New York, Garland Press.

Haut de page


1 . Dante Alighieri, La Divina Commedia, testo critico stabilito da Giorgio Petrocchi, Turin, Einaudi, 1971, p. 248.

2 . “Translating rhythm” might be understood to suggest that rhythm is one problem among others faced by the translator, an aspect of language to be taken account of. Henri Meschonnic’s work suggests that, where literary texts are concerned, rhythm is what is to be translated, to the extent that rhythm is the enunciation of a subject, what specifies the language of a text. More particularly, Meschonnic argues that the rhythm of a text is better understood as the continuum of syntax, rhythm, and “prosody” (the text’s phonemic composition), meaning by continuum that each of these components interacts with and determines the others. In his discussion of Hugo’s Châtiments in Écrire Hugo, Meschonnic refers to the “grammaire prosodique et rythmique du poème” (Meschonnic, 1977: 225), suggesting thus that grammar, accentual and phrasal rhythm, and vocalic-consonantal composition determine each other mutually; in a talk given to the Institut des Textes et Manuscrits Modernes (I.T.E.M.) on March 3, 2002, entitled “S’inventer en écrivant”, Meschonnic noted: “du point de vue de mon expérience de pensée, il n’y a plus ni son ni sens dans le langage, il n’y a plus de mots, il n’y a plus de phrases, il y a du continu, un continu rythme-syntaxe-prosodie”. Essential to Meschonnic’s approach to rhythm and to translation is the sense that rhythm and “prosody” are not separate and separable “levels” of language; rather they are what transforms the givens of the language, syntax and lexicon, they are what a subject “does” to language and reader—and thus, likewise, what a translation must “do” (Meschonnic, 2008: 111).

3 . Charlotte Ward’s “Notes to the translations” give numerous examples of what she calls “sound translation” (Ward, 1991: 126).

4 . In Translations Pound prints the Provençal text from U. A. Canello’s edition (1885); according to the introduction to his translations in Literary Essays, he also consulted the edition by René Lavaud (Pound, 1968 [1954]: 115), which departs from Canello’s in significant ways. His translation mostly follows Canello’s, as where he translates nei (l.48) as snow; but one also finds traces of Lavaud’s readings, as where m’esfortz (l.11) is translated strive (l.14; here Canello translates m’è faticoso [Daniel, 1883: 127], Lavaud je m’efforce [Daniel, 1910: 49]). Pound’s translation is shorter than Arnaut’s text by one line; two lines, it seems, are combined in line 73, which should presumably have read Me; mad / Frieks if Pound had followed the lineation and rhyme scheme here as he does elsewhere; all editions of his translation, including the first publication in The New Age of December 26, 1918, read line 73 as here. Pound’s indentations at lines 11-12 of each stanza correspond to an important articulation of the rhyme pattern (paralleled by important syntactic articulations in Arnaut’s poem): lines 10 and 11 mark the division between the two parts of the stanza, frons and cauda; the pattern of the second part (lines 11-17) is then repeated in the coda. Charlotte Ward’s edition of Pound’s translations, based on the typescripts prepared for a 1917 publication which never occurred, gives not mearing at l.108, which I follow here; Literary Essays and Translations give nor mearing. Ward’s text otherwise differs from that in Translations in that it shows no periods at the ends of lines 17 and 41; no comma at the end of line 102; and places a comma instead of a period at the end of line 45 (Ward, 1991: 70-72).

5 . Translations from Provençal into English are by me, adapted from the versions by Canello (1883) and Lavaud (1910); I have also consulted Toja (1960).

6 . Charles Hartman suggests that the line unit is a way of “increasing the density of accents”, in particular because “the energy required to start again each new line tends to throw unwonted stress on one of its first words” (Hartman, 1996 [1980]: 56).

7 . The importance of juxtaposed accents in Pound’s poetry is shown for example in Hall, 2003 and Jones, 2006.

8 . We give here the stanza in Lavaud’s edition, and his translation (Lavaud’s italics mark the rhymes linking verses):

Ges no.m duoill d’amor don badaill,

Ni no sec mesura ni taill;

Sol m’o egaill

Que anc no vim

Del temps Caim

Amador meins acuoilla

Cor trichador

Ni bauzador:

Per che mos cors capduoilla.

Je ne me plains point de l’amour parce qu’il est inattentif et n’observe ni la bonne mesure ni la ‘taille’. Que seulement il me rende la réciproque de ceci, que jamais nous ne vîmes, depuis le temps de Caïn, un amant qui accueille en son cœur moins de mensonge ni de ruse: raison pour laquelle ma joie monte au faite.” (Daniel, 1910: 16-17)

Haut de page

Pour citer cet article

Référence papier

Andrew Eastman, « “Cries and cracks”: “animal” rhythms in Ezra Pound’s translations from Arnaut Daniel », Palimpsestes, 27 | 2014, 63-79.

Référence électronique

Andrew Eastman, « “Cries and cracks”: “animal” rhythms in Ezra Pound’s translations from Arnaut Daniel », Palimpsestes [En ligne], 27 | 2014, mis en ligne le 01 décembre 2016, consulté le 28 mars 2017. URL : ; DOI : 10.4000/palimpsestes.2038

Haut de page


Andrew Eastman

Andrew Eastman est maître de conférences au département d’études anglaises et nord-américaines de l’Université de Strasbourg, où il enseigne la littérature de langue anglaise et la traduction. Ses recherches portent sur les pratiques et conceptions du rythme dans la poésie moderniste de langue anglaise. Parmi ses publications récentes, on compte : « Ideologies of English: stress clash, “Anglo-Saxon”, and twentieth-century conceptions of rhythm » (Études britanniques contemporaines, 39, 2011) ; « Rhythm as continuum: grammar, speechsound and the invention of nature in Lorine Niedecker’s Paean to Place » (Thinking Verse, vol. 3, 2013). Andrew Eastman est membre du groupe Polart, Poétique et politique de l’art.

Haut de page

Droits d’auteur

Tous droits réservés

Haut de page
  • Logo Presses Sorbonne Nouvelle