CAT'S Cradle: A Cognitive-Parametric Model of Computer Assisted Translation

摘要

Establishing inter-lingual textual equivalency, requires cognitive-linguistic competence as a metalinguistic bridge between different codes. This CAT model offers a 9 + step language-independent protocol to correlate Multi-Word-Complexes in different languages as a product of a parsing algorithm ('Point of View') which actualizes one cognitive 'aspect' of a referent shared with many operants. Every parsing choice is assigned a unique alpha-numeric code and each word-form obtains a sequence of codes as its unique parameter. Each sequence is a digital analogue of all cognitive-informational 'aspects' of a referent, operant, modifier, quantifier actualized by all instances of parsing. A protocol of matching word-form codes identifies parametric coordinates of a Multi-Word-Complex employed as a "unit of translation". Correlation is established when Multi-Word-Complex components with the same code parameters are located in the target language. Mapping code progressions in the two texts establishes and verifies cognitive equivalency. The paper offers theoretical basis and practical solutions of implementing cognitive-parametric design for CAT.