Collective Plural Identity

ABSTRACT I argue against a near‐universally accepted extensionality principle concerning the necessary and sufficient conditions for plural identity, that is, the relation of identity that holds between pluralities of things. The extensionality principle says that pluralities xx and yy are identical iff they are same‐membered (i.e., iff any z that is one of xx is also one of yy , and vice versa). I argue that the apparent intuitive case for this principle confuses facts about the nature of plural identity with facts about certain kinds of plural predication, and that other attempts to establish the extensionality principle fail. I then show that the lack of motivation for the extensionality principle translates to a lack of motivation for another widely accepted claim concerning the alleged modal rigidity of pluralities. What will emerge from my discussion is the plausibility of an account of what (for reasons that will become clear) I will call collective plural identity, according to which the relationship between pluralities and their members is not as intimate as has been supposed: Pluralities are neither individuated by the members (for there can be distinct, same‐membered pluralities) nor incapable of changing their members from world to world.