Suarez for helpful discussions and valuable comments. 1 See Post (1963) and Adams (1979). As is well-known in the metaphysics literature, this is incompatible with the first alternative because PII cannot take primitive properties of self-identity and numerical distinctness into account, as this would trivialize it as a criterion of individuation.2 objects. In this paper, I analyse Saunders' argument in favour of this conclusion in detail. 2 In section 1, I introduce the 'generalist' perspective endorsed by Saunders and consider its possible justification and philosophical basis; and then look at the Quinean notion of weak discernibility he appeals to. Next, I move on to a consideration of quantum mechanics. First, in section 2, I summarize the established understanding of the way things stand as regards quantum particles and PII; then, in section 3, I evaluate the way in which Saunders intends to employ the concept of weak discernibility with a view to providing an alternative picture for fermions. There, I expand on the criticisms formulated by Hawley (2006) and Dieks and Veerstegh (2008) by explaining in detail what I take to be the basic problem, not entirely made explicit in these papers: that the properties invoked by Saunders cannot be pointed to as 'individuators' of otherwise indiscernible (and thus numerically identical) entities because their ontological status remains underdetermined by the evidence and the established interpretation of the theory. In addition to this, in section 4 I contend that, even if he is granted his conclusions as regards fermions, Saunders does not deal adequately with bosons, and cannot do so because he subscribes to PII and the generalist picture. Section 5 contains a critical examination of the widespread claim (or at least implicit assumption) that the generalist picture should be regarded as obviously compelling by the modern-day empiricist.
The generalist picture, weak discernibility, and Black's universeIn his (2003), Saunders endorses what, following O' Leary-Hawthorne and Cover (1996), he calls the 'generalist picture'. This view is a distinctive and uncompromising form of realism, a commitment to the adequacy of purely descriptive concepts (Saunders 2003, pp. 289-290). 3
Matteo Morganti2 Saunders' ideas have been elaborated in more detail and applied more generally by Muller and Saunders (2008). Since the considerations made in this paper apply to the more general framework as well, however, I will only consider Saunders' original paper here. 3 This is intimately connected to the 'semantic universalism' developed by Van Fraassen ((1977-8), (1991). That is, the thesis that all factual descriptions can be given completely in terms of general propositions making no reference to individuals. See also Hintikka's remark that "[e]ach possible world contains a number of individuals with certain properties and with certain relations to each other. We have to use these properties and relations to decide which member (if any) of a given possible world is identical wit...