Contrasting post-editing and human translation along the dimension of term and cognate variation
published: Nov. 7, 2016, recorded: October 2016, views: 1274
Report a problem or upload filesIf you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.
Post-editing is a rather recent mode of translation production. In the most basic definition, post-editing is the
correction of machine translation output, but some definitions include aspects such as quality criteria or the fact
that post-editors should be trained translators (see e.g. O’Brien 2011).
Post-editing has been studied from various angles, mainly with respect to the questions of efficiency or quality. In some cases, the process of post-editing has been contrasted to the process of human translation (see O’Brien et al. 2014 for a collection of studies). While differences have been shown to exist between the two processes, e.g. on the level of macro processes (see e.g. Carl et al. 2011), little is known about how the postedited products differ from human translation (with first insight provided e.g. in Lapshinova-Koltunski 2013; Lapshinova-Koltunski 2015) and which effects this might have on communication.
In this talk, I will first contrast post-editing and human translation along the dimension of term translation within the domain of Languages for Specific Purposes. In the study presented, terminological variation in translations from English into German was measured for both modes of translation. The findings reveal levels of variation on the terminological level in the post-edited texts close, but not identical, to those of the machine translation outcomes. They thus indicate a shining through of the machine translations in the post-editing products, motivating further research into the properties of post-edited texts within corpus-based translation studies. Also, I will present a brief pilot study on how cognates are used differently in machine and human translations: While the machine translation system used was heavily using cognates, in human translation (motivated) variation in the (non-)use of cognates could be observed.
Of course, both findings are very much dependent on the characteristics of the machine translation used. However, the point here is not to make generalizable statements on lexical properties of machine translated texts, but to identify dimensions along which human translations, post-edits and machine translations may differ and may be contrasted. On the basis of these findings, I will discuss in what way machine translation can have an impact on the product of translation and whether it might become a driving force for language change.
Download slides: jtdh2016_culo_human_translation_01.pdf (1.2 MB)
Link this pageWould you like to put a link to this lecture on your homepage?
Go ahead! Copy the HTML snippet !
Write your own review or comment: