Poetry Translating as Expert Action
which give valuable points of comparison. The questions, which were sent to inter-
viewees a few days beforehand, read as follows^1 :
Selecting works for translation: What strategies do you use here? Do you select? On
which criteria? And/or are you selected?
Translating Strategies/Phases: What are the phases of tackling a ‘typical’ translation
job, from opening the source text to sending off to the publisher? Do you have a
separate source-text reading/analysis phase? Do you first translate orally/mentally
or go straight to writing? What do you write in your first version? How do you
organise it? What strategies do you use here? Do you have a revision phase? If
so, what do you do at each revision? Do other people (e.g. native speakers of the
source language) have input into your target text? If so, when and how?
Translating Strategies/Knotty Problems: What are the knottiest problems you have
to solve as a poetry translator? How do you tackle tricky word-meaning issues?
How do you tackle word-sound issues? How do you tackle the issue of traditional
poetic form (rhyme, rhythm, etc.)? What about non-contemporary source texts?
Philosophy and beliefs: Do you have any overt principles/beliefs which you feel
guide your translation strategies (e.g. don’t worry about rhyme & rhythm, or try
to reproduce it exactly)? What are they?
Placing works with publishers: What strategies do you use for placing works with
publishers? What problems have you encountered, and how have these problems
been solved?
Other factors: Is source language/culture a variable? How? Is target language/dia-
lect a variable (if relevant)? Deadlines – how do these affect your strategies? Are
there any other factors we haven’t mentioned?All interviews were one-to-one, and lasted between 1½ and 2 hours. Four were
held face-to-face, and the fifth by telephone. The telephone interview was of aver-
age length: hence the medium did not affect the amount of data generated. Inter-
viewees gave wide-ranging and sometimes lengthy answers, which often extended
well beyond their initial question. Extra, improvised questions probed more deep-
ly into issues raised in individual answers.
All interviews were audio-recorded. Recordings were transcribed by other
transcribers. I checked the transcripts, and then analysed them qualitatively for
underlying trends. This analysis was repeated over a year later, thus increasing
its reliability.
All research procedures raise questions of validity: whether the data they gen-
erate can be trusted. One question with these interviews is how far they can gener-
ate generalizable conclusions, given the five translators’ profiles. Firstly, most of
their source languages are Indo-European, particularly Germanic. However, each
of the four interviewees with multiple source languages reported the same strategies- Omitting factual questions about translator background.