“Simple generics” with bare plural subjects (e.g., dogs bark) predicate of a kind a property that the kind “inherits” from its individual members. But what does that inheritance amount to if it is not, like most dogs bark, based on how many individuals have the property. My conclusion: there is no determinate account of which (fundamentally individual-level) properties can be truly predicated of a kind: generics are not quantificational, and language users’ interests guide judgments on their truth-conditions. At the same time, even “canonical” predications of ordinary predicates of ordinary individuals are not so straightforward as they might appear. Generic claims about social groups show the indeterminacy of truth conditions for simple generics and the relation to stereotypes and sometimes conflicting interests.
