Abstract
The aim of the paper is to defend the idea that, from the semantic point of view, predicates of personal taste behave in communication like ordinary indexical expressions (pronouns, demonstratives, etc.). It means that they express different semantic contents relative to relevantly different contexts of utterances. This is a consequence of the claim that “tasty” (which is our paradigm example of a predicate of personal taste) and “tasty for (someone)” (where “someone” is a placeholder for an agent) express the same, or very similar, semantic contents relative to the same context of utterance. Now this idea can be challenged by pointing to certain communicative phenomena, such as disagreements about matters of personal taste. It is argued, however, that there is an explanation of taste disagreements that is compatible with the indexical nature of predicates of personal taste. Moreover, it is shown that an explanation along these lines is rather natural because it seems to be necessitated by communication-based evidence.