"How can I say this in fewer words?" Exploring how a cognitive theory of conceptual metonymy can be useful for understanding reduction in subtitling

Historically, metonymy has been conceptualized as a feature of rhetoric, where we use the name of a person, object, event, etc., to stand for another, e.g., ‘brain’ for ‘intellect’ or the name ‘Picasso’ to refer to the artist’s work. However, with the emergence of metaphor and metonymy research within cognitive linguistics (Lakoff and Johnson 1980; Kövecses and Radden 1999; Barcelona 2003), metonymy is now regarded as a cognitive process that allows language users to communicate complex ideas economically and efficiently. This paper examines how this theorization of metonymy can shed light on the interface between two previously self-standing research areas, namely audiovisual translation and cognitive linguistics and gauges to what extent the principles of conceptual metonymy can be applied to the study of reduction in subtitling. Previous studies on reduction in subtitles have been both qualitative (examining which parts of speech are more likely to be reduced or omitted in subtitles) and quantitative (where the focus is on how much of the source text is omitted in the subtitles or how many written subtitles can viewers process within a given time). By contrast, this paper will offer new insights into how reductions in subtitles can be produced efficiently with the help of conceptual metonymy theory. The starting point for the discussion will be Radden and Kövecses’ seminal essay ‘Towards a theory of metonymy’. According to the authors, cognitive principles such as human experience, perceptual selectivity and cultural preferences allow for the use of metonymy. Among the outlined cognitive principles there are several that are particularly relevant to the study of translations, such as “typical over non-typical”, “central over peripheral”, “visible over non-visible”. In addition to this, Radden and Kövecses build on Lakoff’s (1987) theory that all knowledge of objects, events, and processes, both in the abstract and concrete realm, can be organised into idealised cognitive models (ICMs), and that we can use one part of an ICM to describe other parts of that same ICM. By using examples from the Norwegian subtitles of Sherlock (2010-2017), I will illustrate how, in certain contexts, a cognitive theory of conceptual metonymy can inform a meaningful reduction of text in subtitles, before proceeding to argue that this knowledge can be useful for subtitlers.