Relevance Theory and Translation (Gutt)
Ernst-August Gutt applied Sperber and Wilson's relevance theory to translation. The translator's job, on Gutt's account, is to produce a text that lets the target reader arrive at the intended interpretation with reasonable effort.
Direct vs indirect translation
- Direct translation: aims to preserve all the communicative clues of the source so the target reader can recover the original interpretation. Closer to formal equivalence.
- Indirect translation: aims to convey the relevant interpretation without preserving every clue. Closer to dynamic equivalence and to summary.
Why this matters for AI
Relevance theory asks whether the target reader can recover the right interpretation with reasonable cognitive effort. Chunked AI translation can break this when context that would have made one chunk relevant lives only in a chunk the model never sees. Glossary plus framework plus register profile compensate by carrying that context across chunks.
Further reading: dynamic vs formal equivalence, why AI translation loses consistency.