Defining taste in English informant categorization

The sense of taste has been classified in the lower ranks of perception by most Western philosophers since the very beginning of the philosophical tradition in Ancient Greece. One of the reasons for this undervaluation is its instinctual and intimate nature. Furthermore, the vocabulary of taste does not seem to be as variegated and rich as that pertaining to other senses. The present paper aims at exploring the domain of taste descriptors in English by unravelling the family resemblances that structure the conceptual category. The free-sorting task is the experimental procedure that I adopted to investigate this domain, in keeping with a usage-based approach in a Cognitive Linguistics perspective. Firstly, I collected a list of potential taste- descriptors by searching for synonyms of the five basic English taste terms (i.e. sweet, bitter, sour, salty, umami) and two general taste terms (i.e. yummy, yucky). Secondly, I conducted a pre-test in which a pool of native speakers was asked to evaluate those lexemes, by simply saying whether they could be used to describe a taste or not. Lastly, I carried out a sorting task in which native speakers were asked to group the terms in any number of categories. Participants were then asked to name each category by choosing one of the members, thus priming the formation of a metonymic model, in keeping with Lakoff (1987). Results of the analyses suggest patterns of categorization among the different participants in the test, and show asymmetries in the strength of the association among members, i.e. some of the items in the domain collocate more frequently with other items. I conducted the analyses with the software GraphColl, which enables a visual representation of the collocation patterns, thus allowing an immediate interpretation of the data.