A Comparative Analysis of the Cognitive Processes in Machine Translation and Human Translation

Abstract

The rapid development of artificial intelligence technology makes machine translation show its great advantages in translating multilingual information; yet there are no systematic explanations about the essential difference between them with regard to their cognitive levels. In this paper, we conduct in-depth analysis of the fundamental differences between the two at the level of cognitive mechanisms according to the framework of cognitive science. We find out that machine translation is based on computational models and data driven foundation showing the characteristics of formalized processing and probabilistic optimization while human translation relies on the complex cognitive system of the biological brain possessing advanced cognitive ability like contextual understanding, dynamic knowledge management and creative decision-making. Establishing a systematic comparative analysis framework, we analyze the essential differences from three aspects: language understanding, knowledge application and decision-making processing and propose a new model of human-machine collaborative translation based on cognitive complementarity. Our research will not only enrich our theoretical understanding of the cognitive process of translation, it provides an important basis for building an efficient human-machine collaborative translation system.