House's Translation Quality Assessment Model
Juliane House proposed a systematic way to evaluate translation quality by comparing the register profile of the source with that of the target. Her model rests on a distinction between two translation types.
Overt vs covert
- Overt translation: the source is so tied to its original cultural moment that the translation cannot pretend to be an original (political speeches, historical texts, source-cited scholarship). The reader knows it is a translation.
- Covert translation: the source has no such tie and the translation functions as if originally written in the target language (instruction manuals, product copy, most journalism).
The TQA procedure
Build a register profile of the source (Field, Tenor, Mode in Halliday's sense), build one of the target, compare them. Mismatches are categorised as covertly erroneous (register breach) or overtly erroneous (denotative error). The model is descriptive, not prescriptive.
Further reading: register foundations, worked register examples.