Poetry Translating as Expert Action
rhetorical structures. But if the Toen wij translators had gone on to translate the
next poem in Kouwenaar’s collection, this would almost certainly have been hard-
ly less challenging. This is because poetry’s challenges, unlike those of scientific
articles, tend to be text-specific rather than topic-specific: thus the new poem
would have presented new problems, for which few if any solutions would have
been transferable from Toen wij.
Turning now to the formal challenges of translating poetry, many if not most
of its intrinsic-form features are also found in other genres. Reactivated idioms
and rhymes, for instance, also occur in jokes and advertising copy. One feature
seemingly specific to poetry and the closely-related sub-genre of poetic prose,
however, is that of condensed, hermetic language where the real-world reference
is unclear. As with Toen wij and Krik, this forces translators to spend extra time
exploring the possible real-world and text-world referents of such language. More-
over, a distinguishing characteristic of poetry may well be the way that multiple
features interact: for example, the rich lexis plus fixed rhythm in a Line of Krik, and
the allusive ‘world deleted by darkness’ imagery plus rhyme-scheme in the whole
poem. Though poetry is not unique in combining such features (advertising slo-
gans, for instance, may also do so), poetry is arguably the only genre where this
extends over texts longer than ten or twenty words.
When these technical challenges are combined with the high symbolic capital
of translated poetry, this gives rise to two process characteristics that are more po-
etry-specific than those mentioned earlier. One, shared with no other genre I can
think of, is the exceptionally long time per word, often unpaid, which translators
are prepared to spend on translating poetry. Another is the relatively high use of
text helpers, and high frequency of linguist + target-writer partnerships, in order to
assure target-text quality. Interestingly, the genres that most closely resemble poetry
in this second respect are those where target-text quality is of crucial commercial
importance: the localization of automobile documentation, advertising campaigns
and computer software (Göpferich 2005; Adab and Valdés 2004; Schäler 2009).
Poetry translators, however, do enjoy more working autonomy than transla-
tors in many other genres – including the other teamwork-dominated genres just
mentioned, where translators typically have little autonomy. Firstly, they have rela-
tively high control over textual outcomes. Though a poetry translator may use text
helpers, these are recruited by the translator, and they advise rather than decree.
This book’s studies gave no evidence of more powerful players, such as source
poets or editors, negotiating target-poem solutions with the translator, never mind
insisting on them. Hence published reports of the latter – as with Russian poet
Joseph Brodsky, who “found it hard, or impossible, to accept his translator’s notion
of what was tolerable in English” (Weissbort 1989a; cf. Weissbort 2004; Kline 1989)- probably describe exceptions rather than the rule.