Ethical Issues in Machine Translation
R. Vasilescu
Abstract:ABSTRACT. Globalization and the new technological developments in the line of machine translation have had a strong impact on the translators' work, causing a rethinking of the relationship of the translator to the clients and the others, in general. While considerable attention has been paid to the phenomenon of machine translation and the new computer translation tools, less attention has been paid to the ethical issues arising from the positioning of the translators in an ever more challenging environment. The existing codes of ethics related to translation and translators approach aspects such as accountability, confidentiality, quality of the end product, but there is no mention of MT products. On the other hand, ethics of the profession is not included in most of the university curricula, in spite of the growing awareness of the importance of ethics in all areas of society.Keywords: machine translation; code of ethics; translator's accountability; confidentiality; translation quality1. IntroductionSome say that globalization is accountable for and has triggered radical changes in various fields and areas of human and social lives. It is true in most of the cases, as it is for the translation process. Translations volume has increased considerably with globalization and the break of space barriers side by side with the break of time barriers brought about by the boom of information technology. This increase of the volume of translations has raised the question of the increase of the rate of translating and of the means to do that. Interestingly enough, this is not a concern of contemporary, globalized society, vanguardist predecessors of ours have thought ever since the 17th century of facilitating the translators' effort and saving translating time by the use of machines. But the first patents for 'translating machines' were applied only in the early 20th century (mid. 1930s). By 1950s - the beginning of a turning point period in many directions - machine translation was already experimented in many US universities with tremendous echo in the media. A new branch of linguistics appeared: computational linguistics. Research in this area was driven and supported in the beginnings by the needs of the Russian and American military institutions, which accounts for the limited 250-word bilingual (Russian and English) translation system that opened the ways of machine translation. MT served well the need for information gathering, quality was not imperative. Bom out of an imperious need of having important texts in foreign languages decoded in very short periods of time, machine translation originated in the premise that a rough translation and a basic understanding of a text was good enough. Current global communication and the demand for texts translated in no time still rely on this premise, and the argument is that no human translation has ever been perfect either! Research on improving translation by means of machines has never stopped as organizations' needs are quantitative rather than qualitative, they have larger translation projects, so speed is of essence for (technical) documentation, reports, contracts, even personal correspondence, any type of messages which require no stylistic quality. MT has gradually gained ground in spite of the conclusions of the ALPAC (Automatic Language Processing Advisory Committee) Report 1966 that machine translation was more expensive and less accurate than human translation. MT research continued to be supported due to globalization and the extended need for translations and it passed from the research of limited language pairs in the '60s to low-cost translation systems in the 70s, to mainframe technology in the 80s (Systran, Logos). Research after the mid-1970s had - according to W. John Hutchins (Hutchins: 1995), three main strands:* The development of advanced transfer systems building upon experience with earlier interlingua systems;* The development of new kinds of interlingua systems;* The investigation of techniques and approaches from Artificial Intelligence. …
Sociology,Linguistics,Computer Science