Generics and ways of being normal

This paper is concerned with the semantics of bare plural I-generics such as 'Tigers are striped', 'Chickens lay eggs', and 'Kangaroos live in Australia'. In a series of recent papers, Bernhard Nickel has developed a comprehensive view of a certain class of bare plural I-generics, which he calls characterizing sentences (Nickel in Linguist Philos 31(6):629-648, 2009. doi:10.1007/s10988-008-9049-7 10.1007/s10988-008-9049-7" ; Linguist Philos 33(6):479-512, 2010a. doi:10.1007/s10988-011-9087-4"10.1007/s10988-011-9087-4" TargetType; Philos Impr 10(6):1-25, 2010b). Nickel's ambitious proposal includes a detailed account of their truth-conditions, an account of certain pragmatic phenomena that they give rise to, a metaphysical picture of their truth-makers in terms of mechanisms, and an epistemological story connecting characterizing sentences to such concepts as induction and explanation. This paper offers an extended critique of the central truth-conditional component of Nickel's proposal. In a nutshell, his account has it that 'Tigers are striped' is true iff, for tigers, there is a way of being normal with respect to fur-pattern such that all tigers that are normal that way are striped. I begin by explaining what characterizing sentences are and distinguish several readings that are available for sentences with bare plurals in subject position. I then introduce Nickel's account and discuss some of its predictions which, in my view, seem highly problematic. Moreover, I argue that Nickel's principle of Homogeneity does not go together well with his proposed truth-conditions, and that his truth-conditional account violates a plausible principle about the logic of generics, a principle I call generic non-contradiction.