Back to list

Detail of contribution

Auteur: Anna OGARKOVA

Co-Auteur(s): Cristina SORIANO, Swiss Centre for Affective Sciences, University of Geneva, Switzerland

Metaphor in emotion concepts: experimental testing of corpora evidence

Abstract/Résumé: Few topics have received more elaboration within Conceptual Metaphor Theory than emotions. However, emotion metaphor research faces serious challenges when issues of its broader utility and interdisciplinary integration come into question. Among several others, concerns have been voiced on the methods for emotion metaphor extraction and analysis, the validity of a predominantly onomasiological focus of relevant studies (e.g., Stefanowitsch 2006), and the cognitive status of language-based observations (e.g., Crawford 2009). The present research takes the latter issue seriously and aims to explore experimentally the status of our previously reported corpora-based findings about the metaphoric representation of various ANGER concepts in three languages: English, Russian, and Spanish (e.g. Ogarkova & Soriano in press). Two studies—one seeking support to corpora evidence through the insight provided by the speakers themselves, and the other testing those findings with language-independent experimental methods—are reported. In the first study, we test the relative saliency of certain source domains in the speakers’ representation of the emotion target. We hypothesize that statistically significant differences in corpus appearance should be reflected in the native speakers’ preference for one or another domain in the characterization of an emotion. This will be captured through forced-choice and graded acceptability rating tests (Gries 2002, Arppe & Jarvikivi 2007). The underlying expectation in the paradigm is that the most frequent patterns of use in corpora should be rated as most natural/acceptable by language speakers. Similarly, we expect the most exploited conceptual metaphors to also be perceived as the most natural in connection to an emotion category. In the second study, we test the linguistic findings on saliency through a priming experiment in a memory task (e.g., Crawford et al. 2006). We hypothesize that subjects primed to think of an emotion and then presented with a list of images, may find it easier at a later time to remember the images associated to source domains central for the metaphorical representation of that emotion.