Poetry Translating as Expert Action
and operations, from draft to micro-sequence, with a similar goal at each level – as
predicted by Activity Theory, though there are more levels than the three pro-
posed by Leont’ev (Axel 1997; Engeström and Miettinen 1999). Alternating regu-
larly between zoom-in work (alternative-solution versions, slow runs-through)
and wide-angle work (pure-line versions, quick runs-through) is almost certainly
the most effective way of solving problems of detail whilst keeping the whole poem
in mind. Prioritising lexis is inevitable if translators are to understand and rewrite
a poem rich in associative or polysemous words, and in idioms with no obvious
receptor-language equivalent (a particularly time-consuming challenge even for
prose translators – Tirkkonen-Condit 2002a: 115). And spending much time on
macrostructural meaning is inevitable with a text world that is ambiguous and
multi-layered, but also evocative and emotionally rich.
Texts and ways of tackling them, however, are also socially designed, modified
or constrained. Learning from others how to read poetry, and later to translate it,
also means internalizing the conventional action patterns of the second-order lit-
erary or translating networks to which one’s teachers belong. Very few non-trans-
lator readers, for example, would spend as much effort as these translators on
identifying the role of word-sound, reactivated idioms and multiple meanings,
and on analysing surface semantics, extended images and authorial intent. Such
effort is specific to specialist readers, such as translators or students of literature: in
other words, part of a learned habitus that defines a social group. Similarly, there
is no intrinsic reason why these translators should all aim to write a viable recep-
tor-language poem which reflects as many of the source poem’s meaning poten-
tials as possible. They could have aimed to produce a phrase-for-phrase semantic
gloss, say, or a receptor-language poem that had much less overlap with the source.
In choosing the same translating approach, therefore, they were expressing a
strong measure of social agreement.
Moreover, the long but similar time they spent on the overall task implies a
shared cost-benefit norm: an unspoken agreement that seeking this type of rele-
vant similarity implies putting in a high but finite amount of work. This is linked
to the socio-cultural economics of poetry translation mentioned earlier. Because
of poetry’s cultural value and the popularly-acknowledged difficulty of translating,
the substantial efforts needed to ensure a good-quality target poem can be re-
warded by substantial gains in symbolic capital for the source culture or poet, and
often for the translator.
As Engeström and Miettinen point out, “people construct their institutions
and activities [...] by means of material and discursive, object-oriented actions”
(1999: 10). Thus effective cognitive strategies and processes are shared between
translators through discourse and joint action – in workshop discussions and book
introductions, via co-translating and feedback from peer text-helpers. This,