Machine translation tools: current use and perceptions by French translators

Technological developments have always changed human behavior and the nature of jobs. The profession of translator is no exception to the rule. This thesis addresses the topic of Machine Translation (MT). While human translators have so far always found a way to adapt to technological developments, the most recent evolution of MT might bring changes. The objective of this thesis is to provide the commissioner (myself, a professional translator) and everybody interested in the issue with information on how French translators currently use and perceive Machine Translation (MT) tools. This thesis tried to answer the following main question and two sub-questions: how do French translators currently use and perceive machine translation tools? What is their use of such tools and what do they think about them in terms of effectiveness, quality and reliability? According to them, do machine translation tools have more advantages than disadvantages? Our research was based on the following three initial assumptions: only a small minority of professional translators use MT systems voluntarily (1); translators presume that output data from MT are often of low quality and little use (2) and translators believe that it is better to translate all the segments of a translation themselves, instead of doing post-editing (3). The data was collected following a mixed methodology involving quantitative research (online survey) and qualitative research (semi-structured interviews) targeting French professional translators. The theoretical section explored the technical and historical context relating to MT and the Technology Acceptance Model (TAM). The finding showed that French translators do not utilize MT tools and perceive them in the same way and that perceptions of MT have an incidence on perceived usefulness, perceived ease of use and actual use. Out of this research came also the finding that there is a deficit of knowledge about MT tools and their functioning and also a need to clarify what role should be given to translators and machines in relation to MT. Even though human-machine interactions still have their best days ahead of them in the translation field, nothing, at this point, replaces human translation in terms of quality of writing