Why go the extra mile? How different degrees of post-editing affect perceptions of texts, senders and products among end users

Abstract

In recent decades, post-editing has received its fair share of attention in the industry as well as in academic circles. What has attracted by far the most attention is the question of quality: together, machine translation and post-editing defy long-standing and commonplace notions of quality. In this paper, we try to observe quality from the vantage point of end users, who are believed to have the final say on a text's fitness for purpose. We will report on an experiment in which end users were asked to pass judgment on manipulated machine translations with different degrees of post-editing. Our findings demonstrate that the additional effort associated with higher degrees of post-editing does not necessarily lead to more positive judgments about text quality. The evidence suggests that text quality is context-dependent and is, therefore, subject to a somewhat opaque process of constant (re)negotiation.