Chapter 5. Five translators translate
nesting sub-processes, with remarkably few differences between how individual
translators manage their work at each level.
This study’s translators reported two to five intensive ‘working’ drafts (not
dissimilar to the three to four drafts reported in Chapter 4), interspersed or fol-
lowed by one or more quick ‘tidy-up’ drafts. There was some variation in how
they distributed their efforts across drafts. Otherwise, each draft involved one or
more runs through the poem. Translators switched regularly between three run-
through types, each with a different purpose. Quick reading runs-through serve
to put the source or target poem’s macrostructural meaning into working memo-
ry, and/or to identify translation or revision challenges; slow reading + writing
runs-through ensure that the poem is translated and revised systematically and
thoroughly; and quick reading + writing runs-through quality-check the target
version. Shih (2006) reports that non-literary translators use runs-through for
similar purposes.
In most runs-through, translators referred both to the source and to the
emerging target poem. This confirms Chapter 2’s claim that source-poem reading
and target-poem writing are linked into one process of lectécriture (Peraldi 1978:
109–110), but stresses that lectécriture also involves target-poem reading. For
quality-checking, however, translators often used two ‘quick reading + writing’
runs-through in succession: one comparing source vs. target poem, and the other
evaluating only the target poem.
During reading + writing runs-through, translators generate or revise a se-
quence of receptor-language versions, in three phases:- A hand-written prose literal, typically with receptor-language synonyms and
notes. Stylistic or intrinsic-poetic features, if recorded, are mentioned in notes
rather than incorporated into the target text. - Several versions, on-screen and/or hand-written, incorporating stylistic or
intrinsic-poetic features into the poem text, and tackling other unsolved prob-
lems. Alternative-solution versions alternate with pure-line versions. - A final, on-screen, pure-line version.
This reflects the metacognitive strategy of managing a complex activity by separat-
ing the overall object into a sequence of sub-goals, and tackling each in a separate
action. Phases 1 and 2, for example, focus separately on Holmes’s two aims, build-
ing an original-matched base and then revising it into a receptor-language poem.
This strategy is mentioned in reports across the ages. Thus a contemporary
commentator wrote that, when the 9th-century-CE king and translator Alfred of
Wessex “had learned the book [of Boethius] and translated it from Latin into