A Cognitive Approach to Translation Shifts: using an LDC Arabic-English parallel corpus

Abstract

Translation has for long been unfairly presumed to be an uninteresting photocopy of the original text. However, with the advent of the discipline of Descriptive Translation Studies (DTS), the fact that losses, gains, and changes are a necessity and not a crime has become more established, opening new vistas in the field of Translation Studies. Translation has come to be viewed as a decision-making process. It is not only an act of communication and textual operation, but it is also the result of cognitive processing carried out by translators. It presupposes choices, alternatives, decisions, strategies, aims and goals. “Shift Analysis” is no longer directed at exposing faults committed by the translator. It is rather a means of getting a glimpse of what is inside the “black box”, i.e. the translator’s mind. This, in turn, has led to a shift from focusing on translation-as-product to translation-as-process manifested through its product, yielding insights into the decisions made in the act of translating.

Cognitive linguistics assumes that a linguistic structure is a bipolar form-meaning pair. Construal operations are central to cognitive semantics; as they refer to the ability to conceive and portray the same situation in different ways. They are based on the main four already-established cognitive abilities, fundamental not only to linguistics, but also to various forms of human cognition; namely, Attention/Salience, Judgement/Comparison, Perspective/Situatedness, and Constitution/Gestalt. Since translation is a process of dynamic meaning construction, it necessarily involves the activation of raw conceptual content and its construal. Exploring the reasons behind the shifts in the translation process would mean looking into the cognitive processes involved in this dynamic process.

This descriptive-cognitive approach to translation shifts has utilized another discipline that has grown recently out of the field of Corpus Linguistics; namely, Corpus Translation Studies (CTS). This corpus approach to translation studies is a powerful tool for perceiving differences and variations. Its point of strength is flexibility as well as adaptability.

This investigation is carried out in order to identify what construal operations are behind the shifts that take place in translation from Arabic into English. The corpus drawn upon in this study is one of the very few Arabic-English Parallel Corpora available for research. It is the Arabic-English Parallel News – Part 1 (LDC catalog no.: LDC2004E08) provided by the Linguistic Data Consortium (LDC). The alignment of the files, as well as the statistics, has been carried out by MultiTrans 4.4, one of the softwares developed by MultiCorpora. MultiTrans 4.4, originally a Translation Memory (TM) tool which helps translators perform more efficiently, includes a new alignment process which is completely automated, extremely fast and at high quality levels.

The dissertation falls into two parts; the theoretical setting (Part I) and the practical analysis (Part II). Part I is divided into two chapters. Chapter one introduces Cognitive Semantics, with special reference to Construal Operations. Chapter two provides a bird’s-eye view of the development of the notion of “translation shifts” from prescriptivism to descriptivism. Part II is the practical part where the analysis of the corpus under study takes place within the framework of Croft andCruse’s (2004) taxonomy of “Construal Operations”, with a detailed analysis of the construal operations that have taken place during the translation process of the instances chosen from the corpus under study as well as conclusions drawn from the analyses and recommendations for further research.