Approaching machine translation from translation studies – a perspective on commonalities, potentials, differences

Approaching Machine Translation from Translation Studies – a perspective on commonalities, potentials, differences Oliver ˇCulo Faculty of Translation Studies, Linguistics, and Cultural Studies University of Mainz 76726 Germersheim, Germany culo@uni-mainz.de Abstract The exchange between Translation Studies (TS) and Machine Translation (MT) has been relatively rare. However, given recent developments in both fields like increased importance of post-editing and reintegra- tion of linguistic and translational knowl- edge into hybrid systems, it seems desir- able to intensify the exchange. This paper aims to contribute to bridging the gap be- tween the two fields. I give a brief account of the changing perspective of TS schol- ars on the field of translation as a whole, including MT, leading to a more open con- cept of translation. I also point out some potential for knowledge transfer from TS to MT, the idea here centring around the adoption of text-centric notions from TS both for the further development of MT systems and the study of post-editing phe- nomena. The paper concludes by suggest- ing further steps to be taken in order to fa- cilitate an intensified future exchange. 1 Introduction Translation Studies (TS) and Machine Translation (MT) share core goals, the most prominent among them being the study and accomplishment of trans- lation between two languages. Still, exchange has been remarkably rare between the two disciplines in the past decades. Despite possible reasons for misunderstandings and scepticism, some of them being discussed in section 2, this paper intends to show that intensi- fied exchange between the two fields is possible c⃝ 2014 The authors. This article is licensed under a Creative Commons 3.0 licence, no derivative works, attribution, CC- BY-ND. and even desirable, especially in the light of recent developments. On the one hand, paradigms in MT have been shifting into a direction in which lin- guistic and translational knowledge is being rein- tegrated in various ways in hybrid architectures, cf. e.g. (Eisele et al., 2008), by means of adding mor- phological or word order information , e.g. (Koehn and Hoang, 2007; Collins et al., 2005), adding syn- tactic information, e.g. (Quirk et al., 2005; Ding and Palmer, 2005), or adapting models to domains, e.g. (Koehn and Schroeder, 2007; Bertoldi and Federico, 2009). On the other hand, TS has seen a rise of empirical, often corpus-based research in various areas, e.g. (Hansen-Schirra et al., 2012; Oakes and Ji, 2012; Rojo and Ibarretxe-Antunano, 2013), which in method and communication style certainly is more accessible to researchers from MT. Last but not least, post-editing – where hu- mans and the machine meet – is of growing impor- tance in the translator’s world. This paper thus addresses some points with re- gard to fostering exchange between TS and MT. Section 2 gives an account of some emerging views towards a more open concept of the phe- nomenon called translation. Section 3 makes sug- gestions how MT could benefit from adopting text-centred notions prominent in TS. Section 4 presents some initial findings from post-editing studies, indicating a case of …

Content truncated to 500 words