Metonymy beyond reference: how nominal metonymy carries speaker’s attitude

Metonymy, traditionally characterized as referential, also conveys evaluative meaning: by selecting a particular vehicle, speakers foreground one aspect of the referent, index their perspective, and implicitly signal an attitudinal stance. Despite extensive theoretical discussion of these mechanisms (Carston, 2010; Barcelona, 2011; Littlemore, 2015; 2022; Schumacher, 2019), little controlled empirical work has examined how metonymy contributes to attitudinal meaning or which components of metonymic reference give rise to such effects.

This study presents a series of three experiments with a total of 178 native English-speaking neurotypical adults, investigating how nominal person-related metonymy modulates comprehenders’ interpretation of speaker attitude. Our experiments each examine:

1. whether metonymic form — speaking metonymically vs. literally about the same attribute — affects perceived speaker attitude, and whether such effects arise across both negatively and positively valenced attributes;
2. whether the attitudinal contribution of metonymy (found in i)) presupposes the presence of lexical valence in the vehicle, or instead persists independently of it, by systematically comparing negative, neutral, and positive attributes;

iii) whether attitude interpreation in metonymy is further modulated when the process of speaker’s vehicle selection is made explicit.

All dialogue contexts used as experimental stimuli were independently normed to ensure attitudinal neutrality, so that any evaluative interpretation could be attributed to metonymy itself rather than to contextual bias. In addition, norming was conducted for balanced sets of positively and negatively valenced attributes as well as a large pool of neutral attributes used across the experiments.

Link to session