Bilingual publishing system performance is measurable using various technical
and non-technical parameters: degree of simultaneity of source and target
language publication, accurate matching of published source and target
texts after post-editing, system throughput volumes, source and target
text compliance with local regulatory requirements, scope for author contributions
in either language. A theoretical benchmark model based on integrated
in-house systems that were operational in financial research translation
in the late 1990s and early 2000s (French, Spanish and German source languages
paired with English) is described.
The model’s salient feature is its two-way (author contributions
in either language) dual-platform architecture, entailing mirror-image
authoring, editing, DTP and management structures on source- and target-language
platforms. The model’s preferred physical location for in-house
target-language translators is the source-text platform. This heavy fixed-cost
infrastructure not only permits high scores on all the aforementioned
parameters but comes closest to meeting aspirations (nearly always unrealistic)
for fully simultaneous bilingual publishing.
The model runs counter to today’s trend of English-only publishing.
It provides a springboard for local source-language authoring as opposed
to English-language authoring by non-native English speakers. It makes
local content that would otherwise appear only in (post-edited) English
available to source-language readers. These also benefit from translated
English source-language content otherwise confined to English-readers.
Witherington has spent his whole working life in translation,
recently becoming a freelancer and writer on translation after 17 years
heading up the French to English equity research translation team at BNP
Paribas in London. He worked as an in-house agency translator in Paris
during the 1980s and began his career as Russian to English translator
at the BBC Monitoring Service. email@example.com