The Routledge Handbook of Translation and Pragmatics

Marta Dahlgren

Introduction

This chapter deals with pragmatics in poetry translation, mainly in the evaluation of pragmatic elements in published poetry translations. The examples have been taken from editions by acknowledged publishing houses where the translator has been given prominence and from journals dedicated to literary translation. In most of these cases, the translators have been given the opportunity to explain the rationale for the translation, and the evaluation is based on the comparison between the translators’ expressed aims and the degree to which they have managed to fulfil them.

Section 2 of this chapter also deals with process-related issues, with a view to ascertaining the extent to which pragmatics is taken into account in translating literature. Translators, when attempting to communicate the same interpretation as the one intended in the original, often speak about “the spirit” or “the poetic essence” of a literary work. This has to do with authorial intention, but both in the criticism of an original, and in translation, authorial intention is extremely difficult to pin down, so the translator will have to concentrate on the text, and what can be learnt about the ST poet and the circumstances of creation. If the translators have made their intentions clear in forewords or otherwise, they should also accept that their work might be evaluated and criticised, just as authors of originals habitually accept – and even welcome – serious literary criticism. It is true that the translation of a specific literary work can be approached in many ways and that there is no single “correct” way of translating, but this does not mean that it is not legitimate to comment on the way it has been carried out. Contemporary translation criticism is not so much about discussing matters of taste, but about finding out whether the translators’ avowed aims have been achieved. The publications commented on here present the original and the version facing each other on a double page, supposedly in order to allow the reader to compare them. When the original is not present – and this has been the case when the translators are acknowledged poets – there is often an indication as to where the originals can be found.

When it is claimed that poetry in translation should create the same effect in the target reader and in the reader of the ST, this means that the translation should convey to the readers all the explicit information and all the implicit information that is present in the original. The reader should be able to perform a similar decoding of both texts and, as advocated in Relevance Theory, at a similar processing cost.

1 What is poetry?

It is a generally established opinion that poetry is different from other types of literature, including poetic prose. If this is true, it should be possible to detect elements that are exclusive to poetry, and which do not appear in any other type of discourse. The following traits are usually observed:

• poetry exhibits a typical division in lines, and some type of regular linguistic pattern, dependent on rhythm. Depending on the metre, poems exhibit a certain rhyming pattern

• in poems, ideas are “compressed”, and there is a lack of redundant expressions. Poems are difficult, if not impossible, to summarise

• poetry uses “marked language”, which is usually taken to mean that the density of tropes (metaphor, metonymy, oxymoron, hyperbole, etc.) is higher than in other genres

• poems contain emotive language

• connotation and association are vital aspects in poetry

When scrutinising these elements, it soon becomes evident that the only characteristic privative to poetry is a specific division in lines. Rhythm, which can be syllable-timed as in Romance languages, or stress-timed, as in Germanic languages, is important in all genres and even more so in poetry. A specific rhyming pattern, such as for example the one that is obligatory in a sonnet, used to be present, but contemporary poems rarely exhibit rhyme, at least not full end-rhymes. The lack of redundancy is a matter of degree, as poems can be quite explicative, and also extremely dense. Poems can tell a story, or be descriptive, in which case they can be summarised. Informative prose, including scientific articles and newspapers, contains tropes, especially similes (comparisons), metaphor and metonymy. It is actually difficult to utter anything totally devoid of metaphor, as Lakoff and Johnson (1980), and their followers, have shown. Recent research on cognition has made it clear that humans do not divide their activities into those related to reason and those related to emotion, and language is no exception. Important work has been published in cognitive semantics (Turner & Fauconnier, 1995; Ruiz de Mendoza, 1997) and on cognitive linguistics and translation (Tabakowska, 1993; Tsur, 2002). As for the pragmatic elements of connotation and association, they are present in all genres but are of great importance in poetry and they are instrumental in creating inference and implicature.

Even though it is difficult to pin down elements privative to poetry, poems are recognised as such practically at first sight, simply because they are edited as poems. They are clearly recognisable works of art. According to recent pragmatic accounts of literariness, aiming at separating what is habitually termed “literary discourse” from the pragmatic-cognitive aspect of this kind of discourse, poetry is “expressive discourse”, and this is so because of the existence of an expressive informative intention. Toolan (2017: 13) speaks about the poetic text as carrying a presumption of its own “imaginedness, or fictionality”. In my view, what is important is that both authors and readers recognise these traits. Longhitano (2014) claims that it is not just the presence or frequency of certain linguistic or stylistic devices that marks a text as poetic or literary, but the author’’s avowed intention of producing it. “Prima facie”, says Longhitano (2014: 188),

cognitive effects derived in expressive discourse interpretation seem to have three distinctive characteristics. First, they vary dramatically from one interpreter to another, and even for the same interpreter in different moments: the content of the propositional information derived is thus inherently vague and unpredictable, depending on the integration, in the context of interpretation, of personal, idiosyncratic information that is not represented as mutually accessible – not even in principle – and is subjectively relevant only for the interpreter.

This view of poetic discourse neatly explains why poetry often needs several readings and why what is commonly called “meaning” can be so elusive.

2 The translation of poetry: process-related issues

It is notoriously difficult to make a living from the translation of poetry, but professional poetry translators do exist. Their work is commissioned for publication by editors, publishers of translation journals, or organisers of poetry readings. Many of these translators have published their own verse in the target language, and therefore confer some of their own prestige to the target text (TT). However, most of the translated poetry on the market is self-published, appears on websites, or is part of academic publications on poetry translation. Venuti (2011) says that poetry translations are mainly issued by small and university presses, and they are therefore “marginal” and “ephemeral”, and adds that this is likely to encourage “experimental strategies” (127).

Jones (2011), in an empirical study of how translators tackle the difficult task of translating poetry, produces a list of elements quite similar to the one in section 2. From the study, which includes interviewing and presenting think-aloud protocols of professional translators, it becomes clear which elements cause most problems when translating. It is evident that the participants in the study consider “recreating semantics and other aspects of meaning” as a priority (184), but that they spend much time attempting to transfer “condensed, hermetic language, where the real-world reference is unclear” (186). Jones also finds that “[professional] translators only creatively transform when recreate-everything strategies prove unviable” (180). This includes exerting extreme caution when transferring rhythm and rhyme. Jones’s study is descriptive and not prescriptive, as so many previous studies have been (notably Lefevere, 1975; de Beaugrande, 1978), and clearly process-oriented, but it also gives a clue to what the product of this process might be like. Among the professionals studied by Jones, only one is a poet in the target language. In this case, the differences in approach to the task are not very noticeable. However, it often happens that poets take liberties with the source text (ST) in order to produce what they consider poems “in their own right”, as was the case in Pound’s translation of Chinese poetry, and in the Mexican Octavio Paz’s translation of Shakespeare’s sonnets into Spanish. There are also examples of poets-translating-poets who have included so much of their own idiosyncrasy that the result is not translation but a mixture of mimesis and re-creation. This has happened in the much-celebrated version of a selection of Emily Dickinson’s poems by the Spanish poet Nuria Amat (2004) and a recent selection by the Swedish poet Ann Jäderlund (2012), analysed below.

The translation of poetry, where the choice between the achievement of equivalence in meaning and the achievement of mimesis of form has to be made when approaching each and every poem, is particularly daunting. When emphasis is given to semantic meaning, there will have to be changes in form: rhymes can only be maintained at the cost of changes in ST meaning. An example of a translation determined by rhyme is Max Knight’s version of the German writer Christian Morgenstern’s (1871–1914) “nonsense” poem Das aesthetische Wiesel [The aesthetic weasel]:

Ein Wiesel

[A weasel]

Sass auf einem Kiesel

[sat on a pebble]

Inmitten Bachgeriesel

[in the middle of a gurgling brook]

Knight writes: “A weasel/perched on an easel/within a patch of teasel” (www.alb-neckar-schwarzwald.de). The animal is present, but the gurgling brook, in what in the original is a description of a natural occurrence, has disappeared. It is, however, possible to achieve a solution that is more true to nature, if the animal is changed as in, for example “A rook/sat on a rock/in a gurgling brook”. The second rhyme is partial, but there is alliteration to make up for this. However, as the poem goes on to indicate that the animal sits in the brook exclusively in order to rhyme, if the translation does not take this into account, the whole purpose is lost.

As for rhythm, it is strongly dependent on the natural stress-pattern in a language, and if such patterns differ in the translated language pair, it takes a proficient interpreter to preserve it. In the poems analysed in the experimental part of this chapter, the iambic metre in the English poems has posed considerable challenges in practically all the versions. Rhythm is fundamental in the ST, and an effort should be made to maintain the musicality (or lack of it) of the original in the TT. Rhythm determines pace in expressive discourse, and if a translated poem looks and sounds different from the original, this has more to do with stress patterns and pauses than with the choice of certain lexical equivalents. This is why both ST and TT poems should be read aloud before the final version is published. Rhythm and the pattern of sounds are part of prosody, a property of all genres (see Tsur, 1992).

It is not unusual for poetry translators to make clear at the outset, or in a postscript, what their aim is in translating, but the translators’ perceptions of their own achievements do not always correspond to what is offered in their versions. Many poetry translators claim to have translated what they call the “spirit of the poem”. The Spanish academic José Siles adds that this spirit can be distorted if the task of translation is undertaken “with passion” (Siles, 2006: 9). What is then “the spirit of the poem”, often also called “the poetic essence”, “the ineffable” or “what you read between the lines”? If poetry contains elements that cannot be accounted for, that is, some kind of free-floating extra-linguistic essence that does not depend on any element present in the ST, it would follow that poetry is un-translatable. But successful poetry translations do exist. The spirit may well be an extralinguistic element, but it is ultimately dependent on intralinguistic elements for its appearance. Pragmatics may not be mentioned at all, but as it deals precisely with the so-called “spirit”, it would be useful for translators to gain some knowledge of how it works.

The spirit of the poem is more inferred than spelt out. Rather than read between the lines, what we do when reading poetry is re-create the poet’s images in our minds. An image can be triggered by a single lexical element, but is dependent for its full re-creation on the surrounding discourse, which means that images are context-driven. This type of inference, based on imagery and emotions, has not been seriously studied within pragmatics, but rather in the field of cognitive linguistics, as already mentioned, and notably in publications on metaphor and metonymy. This chapter does not deal specifically with tropes, but will include a brief discussion on the relationship between imagery and inference.

3 Pragmatics and poetry: product-related issues

In her 2010 book Defining Pragmatics, Mira Ariel attempts a division between grammar (which includes syntax, semantics and phonetics/phonology) and pragmatics, based on elements related to code and elements related to inference (Ariel, 2010). I have argued elsewhere (Dahlgren, 2005b: 1082) that grammar includes pragmatics, but would modify this statement to say that syntax, semantics and phonetics all carry pragmatic effects. Suprasegmental elements, such as stress, accent and rhythm, can also be analysed as pragmatic. Code and inference work together in all genres, but this is especially evident in expressive discourse. Pragmatics is useful when analysing translated poetry, as will be argued below.

3.1 Relevance

The concept of relevance was originally meant to explain what occurs in the interaction between speakers and hearers, and presupposes the presence of “ostensive stimulus” on the part of the speaker, which is relevant enough to be worth processing by the hearer. In poetry, the existence of informative intention cannot be taken for granted (see Gutt, 1991; Lecercle, 1999), and there is always an extra processing cost. Relevance Theory postulates that the readers of a text, poetic or otherwise, are always in search for relevance. Some contextual effects may be more accessible, and are therefore given more attention than others, and are thus clues in the search for meaning. In translation, if such clues are eliminated, the processing cost increases. When reading poetry, in the ST or in the TT, it can happen that the reader gives up searching for syntactic or lexical consistency, and focuses on phonetic and phonological elements, mainly on rhythm and elements that make the poem “sound good”.

Gutt (1991), writing on translation and relevance, insists on this and argues that the difference between implicit information and information that is not expressed (simply absent) can depend on the speakers’ intention to convey it. As indicated above, when the audience has no access to the communicator’s intention, there is no way to tell one from the other. Also, an original poem may contain elements shared by the author and the readers that are not accessible to the TT readers, or are present at great processing cost. Processing, when translating into a different culture, can include inquiring about the poet’s circumstances and being conversant with the poet’s production. When a translator makes use of paraphrase, explication, de-poetising strategies and trivialisation, the poem might be easier to understand, but poetic effects caused by weak implicatures disappear. If, on the contrary, there are additions of what are habitually called “poetic” elements, often from the target language tradition, the processing cost might increase. Relevant translations, then, should be as close as possible to the ST on all levels: phonetic, semantic and pragmatic. Examples of translations exhibiting deviating syntax are rare in non-literary text types, but in poetic prose and poetry they can be found, and are often present in the translator’s attempt to mark the existence of deviating syntax in the original (see 5.4).

3.2 Denotation, connotation, association, inference and implicature

Denotation is a term used in pragmatics in connection with reference. It is related to what in the philosophy of language is called “referring expressions” or “propositions expressed” or “what is explicit”. “What is implicit” in a text is not the contrary of “what is explicit”, but consists of elements that can be inferred from what is explicit. The terms “inference” and “implicature” have been well defined in pragmatics, but they are not habitually used in literary studies. Contrariwise, connotations and associations are rarely mentioned in linguistics. However, Keith Allan, in a 2007 article in Journal of Pragmatics, offers a definition: “The connotations of a language expression are pragmatic effects that arise from encyclopedic knowledge about its denotation (or reference) and also from experiences, beliefs, and prejudices about the contexts in which the expression is typically used” (Allan, 2007: 1047). Geoffrey Leech speaks about literary texts as containing connotative elements used for deliberate effect. “Connotative meaning”, Leech says, is

this power of a word, sentence, etc. to conjure up the typical context of its occurrence. But this is not the whole explanation of “connotation” for this term is used not only of the associations which go with the use of the linguistic item itself, but also of the association of what it refers to. If, for instance, night, blood, ghost, thunder, are said to have “sinister connotations”, it is surely because this suggestive quality belongs to the things themselves [. . .] rather than just to the words. The sinister aura would be felt (no doubt more powerfully) in pictorial or auditory representations of these things, just as much as it is in the words denoting them. In my opinion, linguistics can say nothing about this latter kind of associativity, which is nevertheless of undeniable importance in poetry.

(Leech, 1969: 41)

Leech adds that connotations are “vague and indeterminate” and that “[t]his is the area of subjective interpretation par excellence; a person’s reaction to a word, emotive and otherwise, depends to a great extent on that person’s individual experience of the thing or quality referred to” (Leech, 1969: 216).

Dorothy Kenny, a translation theorist, lists different types of translation equivalence, and includes “connotative equivalence”, which implies that “source language and target language words produce the same or similar associations in the minds of native speakers of the two languages” (Kenny, 1998: 77–78). The existence of this type of equivalence would presuppose a stable, community-based way of assigning connotative meanings. This “connotative equivalence” is actually very problematic. If connotations are community-based and culturally determined it cannot be taken for granted that such connotations can be transferred into another community.

Leech and Kenny do not establish a clear difference between connotations and associations, but use them indistinctly, or rather claim that connotations are a certain type of association. In Allan’s (1991, 2007) view, it is difficult to separate out connotative meanings that depend on the prevailing “connotation” in a certain community, and individual affective meaning. For the purpose of analysis, it might be useful to separate them: any noun comes with a certain number of associations, which should be the same for all proficient speakers of a language, as they are culturally and socially determined. “Death” would then be associated with, for example, old age, illness and mourning. Connotations, on the other hand, are the elements of added meaning that cannot be taken for granted, not even within the same language. They are vague and indeterminate and dependent on subjective interpretation, a person’s reaction to a word, very often emotive. “Death” would then carry a set of connotative meanings triggered by personal experience, such as for example the emotions felt at a loved person’s deathbed, and not shared, not even by speakers of a specific language. From this it follows that translators have access to the ST authors’ associations, but rarely, if at all, to their connotations.

Both associations and connotations bear similitude to the pragmatic notion of inference. They coincide in that they are ultimately language-based, and in that they are, in Grice’s terms, “calculable”. This means that a proficient user of a language is capable of making out what the linguistic item refers to. Inference is also “defeasible”, which means that elements that appear later in the text can contribute to a change in meaning. Inferences can be weakened or strengthened by surrounding discourse. Connotation differs from inference mainly in its range: connotations are subjective and limitless and closely related to emotion. What for one person is a connotation-trigger might not trigger anything at all in another. It is therefore practically impossible for a hearer or a reader to have access to a speaker’s or a writer’s possible world of connotations unless what Sperber and Wilson (1995: 598) call “a mutually cognitive environment” is established, but this is rarely the case in poetry.

In speaking, inferences can be lost and retrieved, while in fiction the author has to create the appropriate surroundings and a natural dialogue. When we talk, and also when we write, it often happens that we do not spell things out: from what we say, our interlocutors draw certain conclusions, and they do not always draw the conclusions we intend them to draw. Inference is a pervasive element in all human discourse. However, in expressive discourse, and very especially in poems, it may not be possible for a reader to calculate meaning and it may not even have been the poet’s intention to make meaning clear. As inference is also defeasible, a certain notion can be established, only to be re-considered at a later stage and invalidated. Conversely, when inference is reinforced by the presence of subsequent expressions, an inference line is created, which can be referred to as “implicature”.

The circumstances of uttering, and what the people involved have in common (encyclopedic knowledge, circumstances, common environment) are important for pinning down the meaning. Sometimes the speaker does not want to be well understood and does what is called “flouting of conversational maxims” (set up by Grice, 1991: speak the truth, not to say too little or too much, to be relevant and to be orderly). What Grice calls “conversational implicatures” appear when these maxims are not adhered to. Poetry is a genre whose hallmark is the flouting of maxims: truth-related semantics does not apply, the poet often says too little, and information can come in a confusing order. As will be seen below, the maxim of relevance is subject to frequent flouting.

4 Practical examples

4.1 Relevance

The concept of relevance is most useful when considering the macrostructure of a poem, i.e., the poem as a whole. The relevance is then the overall impact of the poem, which can also be considered its “meaning”. The notion of relevance can also be invoked in order to understand what has happened in translation when inference triggers have been changed or left out and the creation of images has been impeded. Lakoff’s concept of image schema (IS) has been used in the literary analysis of poems in the original language. The literary work is seen as containing a master image, on to which several mappings are performed (Freeman, 2000). Such mappings are similar to inference triggers: it is not only certain words that trigger inference, but also the images called up by certain expressions, and especially by metonymy and metaphor. If the image cannot be retrieved, the interpretation of the poem will be seriously hampered.

A case in point is the poem by Emily Dickinson “Because I could not stop for death/He kindly stopped for me” (Franklin, 1998: 492, no. 479) where “Death”, travelling in a carriage with the poetic persona, takes on characteristics not commonly related to the noun. In the original, Death is personified as a kind suitor who, accompanied by a chaperone (Immortality) comes to take the persona for a drive. In Spanish, nouns are gendered, and Muerte is feminine, which has serious consequences for interpretation. Nuria Amat, who presents her versions as Amor infiel. Emily Dickinson por Nuria Amat [Unfaithful love. Emily Dickinson by Nuria Amat] has translated “death” with hombre-muerte [man-death] which calls up the medieval image of death as “the grim reaper”, and is not compatible with the view of Death as a delicate suitor.

George Monteiro (2008: 106), evaluating a Portuguese version by Krähenbühl published in 1956 where the first line reads:

Fazer convite á Morte eu non podía

[Make an invitation to Death (FEM) I could not]

observed that the fact that the noun morte is feminine “causes havoc” in the translation, and goes as far as to imply that the translation invites a homosexual metaphor. Another Portuguese translation by Paolo Vizioli (cited in Monteiro, 2008: 107) changes A Morte into O morrer, which Monteiro considers felicitous:

Não podendo esperar pelo morrer,

[Not being able to wait for dying]

De me esperar teve a bondade:

[to wait for me it (he, she) was good enough]

Levava a carruagem a nós dois

[the carriage carried the two of us (PL. MALE)]

E mais a Inmortalidade.

[and also Immortality].

However, this infinitive, which can be back-translated as “the (act of) dying” is difficult to turn into a personification, and there is a change in associations. Also, the syntax in this stanza is problematic, as the subject appears in the second line, incorporated into the verb “teve” which includes the third person singular pronoun (“it?”,”he?”, “she?”). The pronoun does not refer back to any noun. The carriage holds at least one male person (dois [two] is marked for male and plural, and includes at least one male). This produces the uncomfortable impression that “o morrer” is the one who “waits” both for “Death” and for “me”. From the above, it cannot be inferred who drives the carriage.

In view of the complications created by the associations related to “death” in this poem, a simple paraphrase, such as the one offered by the Spanish academic Margarita Ardanaz (Dickinson, 2000: 257) might be preferable:

Porque a la Muerte yo esperar no pude –

[Because for Death (FEM) I could not wait –

Ella por mí esperó amablemente –

[She for me waited kindly] –

La carroza albergaba a Nosotros tan sólo –

[The carriage held Us only] –

Y a la Inmortalidad.

[And Immortality]

Practically all the Spanish translations I have accessed use la Muerte, and stress the gender consigning the pronoun, ella [she], something that is not necessary in Spanish (the line could have read “por mí esperó amablemente”). Even more problematic is the translation of “stop for”, in the first line with the meaning “pause for”, in Spanish dejar mis tareas a causa de . . . , and in the second line, meaning “come to pick sbd up [venir a recoger a alg.], impossible to accommodate within the rhythmic pattern. Ardanaz, in her foreword, says that a good poetry translator should aim at producing poetry. Therefore, preserving the ST rhythm has been important to her, but not if that means sacrificing meaning. In her own words: “. . . ese otro texto resultante tenga entidad por sí mismo y su forma poética responda a los códigos de la respectiva lengua [this other resulting text should have an entity of its own and that its poetic form should respond to the codes of the TT]” (42), and “Hemos procurado /. . ./ mantener el ritmo del texto en castellano, pero siendo siempre más fiel a su palabra que a ninguna otra consideración [We have tried to maintain the rhythm of the ST in Castillian, but always being, above all, faithful to its meaning] (43).

Relevance can be applied to the analysis of a poem on a macrostructure level, but it is even more common to find it on the microstructure level, in phrases where the translator’s choice can be explained applying Sperber and Wilson’s tenets. Two examples will be given from a recent collection of translations from Galician into English published in 2016 with the title Six Galician Poets (Palacios, 2016). The translator is an Irish-born poet, Keith Payne, whose solutions are often daring, and the semantic equivalence can, at times, be questioned. Payne (in a presentation at the University of Vigo of the volume on 10 March 2017) acknowledged that the translations had been made in close contact with the authors, and with the help of the bilingual editor, which might be why the poems as “wholes” are always pragmatically relevant.

In the poem beginning Todos te pretendían by Xosé María Álvarez Cáccamo, from his work O lume branco [The white fire], the first six lines read as follows (the back translation is as literal as I can manage):

Todos te pretendían porque viñeras acompañada dun rumor e

[They all pretended you because you arrived accompanied by a rumor and]

chegabas de cidades non domésticas

[you came from cities non-domestic]

e dos teus labios esenciais cantabas a louvanza da fronteira

[and from your essential lips you sang a praise of the frontier]

con entoación excéntrica. Por iso todos

[with eccentric entonation. Therefore they all]

soñaron posuír a ciencia do teu corpo que imaxinaban sabia

[dreamt of possessing the science of your body which they imagined wise]

en dor pero proveedora de mortal exaltación. Só eu, . . .

[in suffering but a provider of mortal exaltation. Only me, . . .]

(Palacios 2016: 26–27)

Keith Payne interprets:

They all wanted to be with you when you blew in with the

good word from those faraway cities

and from your indispensable lips flew eulogies for the frontier

in a most singular pitch. And so

they all dreamt of possessing the science of your body they

imagined wise in suffering and a trader in grave praise. But I, . . .

(Palacios, 2016: 26–27)

If the translation is read without looking at the original on the opposite page, the phrase “trader in grave praise” will be interpreted in the pragmatically most relevant way, that is, reading “grave” as “serious”, and “grave praise” as “serious praise”, and no connection will be suspected with anything deadly or related to a grave (a tomb) unless mortal is spotted and gives a clue to a different interpretation. A less free version would make it clear that the person who arrives is the one who provides mortal exaltation, i.e., deadly excitement, to those who wish to possess him/her, but this person is not likely to hand out any deadly praise.

In the poem beginning Occidente mosca e sono [Occident, fly /NOUN/ and sleep /NOUN/] – translated by Payne as West louse and sleep, in order to make clear that “fly” and “sleep” are nouns – from the poetry collection Exodus by Daniel Salgado, the last lines read:

Que non se mire máis a si propia

[That it should not look any more at itself]

esta beira

[this shore]

escura

[dark]

dos lugares,

[of the places]

esta carta,

[this letter]

retirada,

[retired]

zona fea,

[ugly area]

ruído de luz.

[noise of /from light].

(Palacios, 2016: 158)

Payne translates:

Don’t let it look at itself any more

this empty

dark

place,

this deserted

letter,

ugly stretch,

light noise.

The translation “light noise” will be understood as “slight noise” and, unless the reader knows Galician or Spanish, “light” will not be taken as luz, i.e., as a source of illumination. Even though “light noise” can be interpreted as having two different meanings, the one consigned in the original, creating a strong image of noise made by light, will not be accessed by the reader without additional processing cost.

4.2 Inference

Depriving the reader of the possibility to infer detracts seriously from the quality of a translation. If the inference trigger is altered, the meaning of a poem will also be altered, or part of the discourse can clash with the interpretation or the image that has just been created. Siles (2006), in his admirable anthology of Anglo-American verse, which includes poems from Chaucer to Dylan Thomas, apart from mentioning the role played by “passion” in translation, says he has respected the number of lines in the poems and furthermore, he has preserved rhyme. His translation of Donne’s “A Hymne to God the Father” is an example of what occurs when the original has suffered elimination of fundamental elements, beginning with the iambic rhythm, which has become dactylic, and inference-related elements of meaning.

III

I have a sinne of feare, that when I have spunne

My last thread, I shall perish on the shore

which becomes:

Yo estOY en peCAdo por MIEdo a exhaLAR

[I am a sinner for fear of exhaling]

Mi ÚLtimo susPIro de rePENte

[My last breath suddenly]

The rhythm in the Spanish version depends on the natural syllabic stress pattern, which here has been marked by the upper case in the translation. There is no mention of the spinning of the thread, from which the original draws the inferred reference to Greek mythology, nor to the shore, which is the dividing line between sea and land, and in this context associated to the journey across the river Lethe. My own proposal (Dahlgren, 2007: 200) is:

Y PEco al teMER que CUANdo LLEgue

[And I sin, fearing that when I arrive]

Al CAbo del HIlo de mi VIda, MUEra en la oRIlla

[at the end of the thread of my life, I’ll die on the shore]

If it is important to preserve all the elements of significance (or of signifying) in a poem, it is also important not to say too much. While it is jarring in a translation to create an inference that is not warranted by the original, or to omit an important clue, it is equally important not to spell out what in the ST is dependent on inference.

In the translation of Auden’s “Oh, what is that sound?” Siles (2006) does exactly this. The title in Spanish, Ay ¿qué es ese tan tan?, includes the answer to the question and no tension is created between the title and the first line, because the onomatopoeia tantán carries the inference that the sound is created by drums. The explanation that the drumming comes from tambores [drums], repeated twice, sounds excessive in Spanish. However, these elements have allowed Siles to create an adequate rhythm, based on iambs and dactyls.

Oh what is that sound which so thrills the ear

Down in the valley drumming, drumming

Only the scarlet soldiers, dear;

The soldiers coming.

¿Ay, qué es ese tantán que retumba en el oído

[Oh what is that drumming sound that echoes in my ear]

Ese tantán de tambores, tambores que sube del valle?

[That drumming sound of drums, drums, that comes up from the valley?]

Los soldados escarlata, cariño

[The scarlet soldiers, dear]

Los soldados que vienen.

[The soldiers that are coming]

4.3 Connotations, associations and inference triggers

An example of how the atmosphere of a setting can change through the selection of lexical elements is the poem Ofelia, written in Galego by Xohana Torres (2004) and translated by Celia de Fréine and by Carys Evans-Corrales. Both versions were published in a selection of Galician poetry for the journal Metamorphoses:

Ofelia

O bóreas sempre sopra polo norte

entre as follas escuras dos alerces

nas almeas onde a néboa se axita.

(Xohana Torres, 2004: 265–267)

Ophelia (1)

The Boreas blows from the north

soughing through dark larch leaves

in the battlements where fog lurks.

(Celia de Fréine, in O’Donnell and Palacios, 2010: 60–61)

Ophelia (2)

Boreas always blows from the north

among the dark larches

on the battlements where the fog swirls.

(Carys Evans-Corrales, 2014: 49–51)

Both translations transmit the feeling of cold and darkness, but Ophelia 1 adds an element of the ominous through the verb “lurk”, associated with some threatening evil, while Ophelia 2 adds no such association, translating the Galician verb axitarse [agitate itself] with the similar “swirl”.

The preservation of source text indeterminacy is one of the most difficult matters in poetry translation. In Emily Dickinson’s poetry, what is generally called ambiguity, but which is rather a deliberate lack of definition, is one of the hallmarks. What some translators seem to have done is to add ambiguity in places where the original has none, to make up for the failure in preservation of ambiguity elsewhere in the poem. The complete range of ideas in the original, including ambiguities, should be preserved in translation whenever this is possible. However, pragmatically, it is just as inadequate to create an ambiguity where the source text has none (Dahlgren, 1998: 26–27).

A consistent change in lexical elements that produce connotations in the original, when added up, causes the elimination of inference triggers, and the final result is very often a translation that comes close to nonsense. In order to show how repeated lexical mistranslations can change the interpretation of a poem, a quotation from a critique by Fiona Macintosh of the translation, based on Johnson’s 1960 edition, by the Argentinean poet Silvina Ocampo can be illustrative:

Apparent mistranslations seem due to false friends, though make some change; for example Poem 249, Might I but moor – Tonight – in Thee (EDJ, 114) becomes ¡Ah! ¡si pudiera morar – esta noche – en ti! (Poemas, 55) where “morar”, though looking cognate, means to stay or dwell and loses the nautical allusion entirely, which would require the verb “amarrar” in Spanish. The verb “morar” gives the Spanish version a quasi-mystical connotation, perhaps calling to mind for Spanish readers such poems as Santa Teresa de Jesús’ “Castillo interior o las moradas”.

(Macintosh, 2005: 29–30)

Macintosh (2005: 29) also mentions Dickinson’s poem “Like Eyes that looked on Wastes”, where the choice of vocabulary gives rise to a succession of changes in associations, thus creating a change in the line of inference-triggers. My own analysis (Dahlgren, 2005a: 84) goes a step further: when misinterpretations come in series, they add significantly to the impossibility of making sense of a poem. Ocampo excels in this kind of lexical mistranslation, as the choices of basuras [trash] for “wastes”, quieta soledad [unmoving solitude] for “steady wilderness”, miseria [poverty] for “misery” produce a poem that is practically incomprehensible. The ST is Franklin, 1998: 664, poem no. 693.

Like Eyes that looked on Wastes –

Incredulous of Ought

But Blank – and steady Wilderness –

Diversified by Night –

Just Infinites of Nought –

As far as it could see –

So looked the face I looked upon –

So looked itself – on Me.

I offered it no Help –

Because the Cause was Mine –

The Misery as Compact

As hopeless – as divine –

Neither – would be absolved –

Neither would be a Queen

Without the Other – Therefore –

We perish – tho’ We reign

Silvina Ocampo offers the following:

Como ojos que miran las basuras

[Like eyes that look at trash]

*Incrédulos de todo –

[incredulous of all (totally unbelieving?)]

Salvo del vacío – y quieta soledad –

[but of the void – and unmoving solitude]

Diversificada por la noche

[diversified at/by night]

*Sólo infinitos de la nada –

[Only infinites of the nothing (infinites of void?)]

tan lejos como podía ver –

[as far as it (he, she) could see –]

así era la cara que yo miré –

[so was the face that I looked at]

así miró ella misma – a la mía –

[so looked she herself – on mine]

No le ofrecí ninguna ayuda –

[I offered it no help]

porque la causa era mía –

[because the cause was mine]

la miseria densa tan compacta

[the dense poverty as compact]

tan desesperanzada – como divina

[as hopeless — as divine]

ninguna – * se absolvería

[none would absolve herself]

ninguna sería una reina

[none would be a queen]

sin la otra – de modo que –

[without the other — therefore]

aunque reinemos – pereceremos

[even though we reign SUBJ – we shall perish]

(Dickinson, 1985 [1997]: 118)

What this poem describes is the despair produced by introspection: the narrator looks into her own soul and finds nothing there. It is a disquieting poem, and Ocampo’s translation produces much the same feeling, only for a different reason: the reader is incapable of making sense of it. The connotations of basura, nada, and miseria situate this poem in some kind of squalid slum area, and the reader infers from this that the hopelessness has to do with the difficulty of getting out of it (becoming a queen, for example, even though the subjunctive form in Spanish indicates that this is not probable). This inference is reinforced by the fact that misery is qualified as “divine”, therefore inescapable. This translation is a clear example of the disastrous effect of ignoring inference triggers. As for “foreignising” elements, there are three instances of expressions that are ungrammatical in Spanish: *incredulos de todo, which, in correct Spanish, should be “incrédulos del todo” “totally unbelieving”; *infinitos de la nada, which is a word-for-word rendering of the original that makes no sense at all in Spanish; and *se absolvería is ungrammatical, since in neither the legal nor the religious sense is it possible for people to absolve themselves.

Silvina Ocampo did not write any introduction to her translation, but the volume has a (very short) foreword by Jorge Luis Borges, indicating that “Casi siempre, en este volumen, tenemos las palabras originales en el mismo orden” [Nearly always, in this volume, we have the original words in the same order]. This may be the reason why so many of the translations have become incomprehensible in Spanish.

4.4 Pragmatics and syntax

For a translation to be accurate (i.e., faithful to the ST) and appropriate (acceptable in the TT, see Toury, 1995) it stands to reason that the syntax will have to be adapted to the TT code. In poetry translation, it is possible, though not frequent, to find examples of deviant syntax in the TT. One instance is the work of the Swedish poet Ann Jäderlund, who in 2012 published a selection of poems by Emily Dickinson with the title Gång på gång är skogarna rosa. Jäderlund, in a postscript, indicates that she has tried to transfer Emily Dickinson’s poems into Swedish “så ordagrant jag förmådde. Ner till minsta syntaktiska rörelse [As word-for-word as I was capable. Down to the minutest syntactic movement] (125). It has been “viktigare för mig att bevara det grammatiska och semantiska [sic] i en dikt, än att försöka upprätthålla dess mer formella/dekorativa drag [more important for me to keep what is grammatical and semantic in a poem, than to try to maintain its more formal/decorative traits]” (126). What Jäderlund means by “grammatical” is unclear, but it might be the same as syntax. Jäderlund sometimes does just this, following the English word order even when the TT calls for a different one.

In her version of “Safe in their Alabaster Chambers” (Franklin, no. 124), the line

Worlds scoop their Arches

has been rendered:

Världar öser deras Bågar [Worlds pour their Arches]

where deras is not reflexive and refers back to a plural noun in the first line of the stanza, namely “years”. This is a purely syntactic mistake – the correct pronoun is sina – and has no pragmatic overtone. However, in the poem “After great pain” (Franklin, nº 372), the last lines, in the original:

As Freezing persons, recollect the Snow –

First – Chill – then Stupor – then the letting go –

which Jäderlund translates as:

Likt människor som Fryser, erinrar sig

Snön –

Först Kyla – sen Dvala – sen

det släppta taget –

the last line is ungrammatical in Swedish, as the verb phrase släppa taget [let go] cannot be construed as a noun phrase. Readers (without access to the original on an opposing page) will infer that the original contains a grammatically deviant expression, which has been accounted for in the translation.

Jäderlund also offers her views on connotations.“Hon gör det tydligt att ordens innebörder inte är något fast – eller ens för stunden givet. Lyser in mot deras konnotationer. [She (Dickinson) makes it evident that the meanings of words are not fixed – or cannot even be taken for granted for the moment. Shine/s a light on their connotations]” (126). This is a misunderstanding of how double meanings work: it is impossible to process both – or several – meanings simultaneously. It is done successively, and even though the translator understands that two or more interpretations are possible, s/he will more often than not have to opt for one solution. Similarly, as associations are connected to lexical items, they depend on the word that triggers them.

The author of one of the most acclaimed Spanish translations of Emily Dickinson, Margarita Ardanaz (Dickinson, 2000) also presents deviant syntax on some occasions, but there is always a reason for it, as in a line from “She bore it till the simple veins” (Franklin nº 81):

Whose but her shy – immortal face

Of whom we’re whispering here?

for which Ardanaz has chosen to present an exact word-for-word translation “down to the minutest syntactical movement”, as it were.

¿Quién sino de ella tímida – inmortal cara

De quien hablamos en voz baja ahora?

(Dickinson, 2000: 89)

There is a clear relevance-related difficulty in processing, which causes the inference that there is a similar difficulty in the original.

Concluding remarks

Grammar includes phonetics, syntax, semantics (lexicon) and pragmatics, and some of these elements are impossible to transfer from one language into another. Phonetics and syntax can be imitated, but in such a case, the result will be an inadequate TT. A proper semantic translation implies sense-for-sense transfer. Pragmatics is ignored at great peril. Suprasegmentals, such as rhythm, prosody and metre, are among the most neglected elements in the translation of contemporary verse, but it can be argued that in the translation of “poetry into poetry”, they cannot be ignored. In some of the poems analysed above, even poets translating their favourite fellow poets, while purporting to transfer “the spirit” of the author, or the “essence” of the poems, make the translations extremely difficult to process. A non-specialist reader, and this category includes most of the literary critics who write about translated poetry, generally explains this away as “normal” in poetry, or as a result of the great passion with which the translator has undertaken the task, and therefore reinforce the stereotype of poetry as impenetrable and incomprehensible.