Abstract

There is a long-standing dispute among scholars concerning Berkeley’s supposed commitment to an emotivist theory of meaning as the very first (and an early modern) instance of non-cognitivism. According to this position, the domains of religious and moral language do not refer to facts about the world, but rather express the emotional attitudes of religious or moral language users. Some scholars involved in the dispute argue for taking Berkeley to be an emotivist (non-cognitivist), while others hold that we should not do so. This paper puts forward an interpretation that lends support to the non-cognitivist reading of his stance, but in expressivist rather than emotivist terms. It argues that the label “expressivism” does more justice to the textual evidence concerning his understanding of moral language, as what is distinctive where this philosopher is concerned is his interest in explaining the nature of our practice of employing moral language (i.e. how we come to formulate moral statements as expressions of our non-referential attitudes, and the meta-level considerations pertaining to morality associated with this), rather than whether morality is just a matter of our emotions or feelings (i.e. such first-order considerations about morality as whether moral rightness and wrongness correspond merely to our emotional states).