Poetry Translating as Expert Action
most translators’ reliance on literal first versions as a task-processing norm might
well be linked to a relation norm of semantic loyalty.7.2.3.2 Creativity
In these studies, which define creativity as applying novel but appropriate solu-
tions to problems, two types of creative solution emerged. Creative adjustments,
which adjust meanings within the same semantic field, are used largely to improve
the target version’s stylistic quality. As they do not breach the semantic-recreation
norm, they are used by all the translators studied here, and thus may be seen as a
key strategy in the experienced poetry translator’s toolkit.
Creative transformations, by contrast, are one of two techniques used specifi-
cally at recreation-impossible points. They prioritize poetic or stylistic correspond-
ence, but abandon semantic correspondence by shifting meanings into another
semantic field. The alternative, non-creative technique is to abandon poetic or sty-
listic correspondence but keep source semantics unchanged. As norms give no
guidelines at such points, some translators lean towards the former technique and
some towards the latter, depending on the relative importance of semantics, in-
trinsic form and style in their personal correspondence hierarchies.
The constitutive convention of communicative loyalty, however, gives strong
pressure against using creative transformations that cannot be construed as reliably
representing the source poem. Hence translators usually only creatively transform
when recreate-everything strategies prove unviable; and they only accept semantic
shifts that seem loyal to the source poet’s underlying intent or the source text world.
This contests the widely-held view that recreating rhyme and rhythm, which often
stimulates creative transformations, risks “mis-translations”, or “betrayal” and “fal-
sification” of source meaning (Lefevere 1975: 51–59; see also Bly 1983: 44–45): the
translators’ concern with deep-level semantic loyalty means they are constantly
guarding against the risk of betrayal or falsification. Instead, when exact lexical
equivalents fail to convey the required formal and stylistic features, they choose
looser “counterparts” which perform functions “that in many and appropriate ways
are closely akin” to those in the source poem (Holmes 1988: 54) – which Holmes
regards not as an aberration, but as a defining feature of poetry translation.
Of course, others might feel that a translator’s creative transformations have
betrayed or falsified the source – though the same holds for surface-semantics-
only solutions, as Chapter 3’s reviewers confirm. There may well be a consensus
among relevant players that some poetry-translation decisions are better than oth-
ers: as a translator, for instance, I would find it unacceptable to omit a verse from a
poem (as alleged by the reviewer of David Harsent’s “Bosnian Romeo and Juliet”
translation: p. 73). Establishing such a consensus is beyond the scope of this book;
until it is established, however, the limits of acceptability must remain uncertain.