![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
Abstract
Proponents of the subset account of property realization commonly make the assumption that the summing of properties entails the summing of their forward-looking causal features. This paper seeks to establish that this assumption is false. Moreover, it aims to demonstrate that without this assumption the fact that the subset account captures an entailment relation—which it must if it is to be of any use to non-reductive physicalism—becomes questionable.
Notes
1 See, for example, Shoemaker [2013]. Other central proponents of this strategy include Clapp [Citation2001], Watkins [Citation2002] and Wilson [Citation1999].
2 For a discussion of a version of the entailment requirement and its importance in evaluating theories of realization, see Morris [Citation2010]. Morris argues that the primary physicalist constraint on a theory of realization is that it should non-trivially imply that instances of physically realized properties are necessitated, in a modally strong sense, by how things are physically.
3 Note that Shoemaker [2013: 42] has recently retracted a previous version of the subset account precisely because it failed to meet the entailment requirement. That version added the requirement that for X to be s-realized by Y the backward-looking causal features of Y must be a proper subset of those of X, where these concern what sorts of states of affairs can cause its instantiation [Shoemaker Citation2003b, 2007]. McLaughlin [Citation2007: 160] demonstrates that, given this extra claim, the subset account fails to meet the entailment requirement. This paper establishes that, for very different reasons, even without this extra claim, the subset account fails to meet the entailment requirement.
4 An anonymous referee has pointed out that the subset account can easily be made to satisfy the entailment requirement by reformulating it as follows:
Where X and Y are properties instantiated by the same object, X is s-realized* by Y just in case the forward-looking causal features of X are a proper subset of those of Y, and having Y entails having X.
It is trivially true that if Y s-realizes* X, having Y entails having X. However, there are at least two problems with this re-formulation. First, what now seems to be doing all of the work is the claim that X is s-realized* by Y just in case having Y entails having X. It is unclear what further purpose is served by the claim that the forward-looking causal features of X must be a proper subset of those of Y. (For further development of this kind of point, see McLaughlin [Citation2007] and Morris [Citation2010: 399].) Secondly, if Y s-realizes* X, the fact that having Y entails having X is not a consequence of the fact that the forward-looking causal features of X are a proper subset of those of Y. It is instead an additional, brute claim of the subset account. Hence, it leaves the question of why the realization relation is an entailment relation unanswered, when it is precisely this fact that cries out for explanation. (See Morris [ibid.: 399–400] for further defence of a point along these lines.)
5 Shoemaker is certainly not alone in this reasoning. See, for example, Wilson [Citation2009: 165].Note that Shoemaker wishes to deny that every conjunctive property s-realizes its conjuncts. He therefore restricts the class of conjunctive properties that count as s-realizers. According to Shoemaker, the forward-looking causal features of a conjunctive property always contain as a proper subset those of each of its conjuncts. However, Shoemaker adds the requirement that a conjunctive property counts as an s-realizer of one of its conjuncts only if there is an asymmetrical relation between the conjuncts, such that the instantiation of one of the conjuncts narrows the way determinable powers bestowed by the other conjunct (the one that is realized) can be exercised, but not vice versa [2007: 28]. A discussion of this restriction is not of relevance to this paper. My issue is with Shoemaker's assumption that the forward-looking causal features of a conjunctive property always contain as a proper subset those of each of its conjuncts. I reject this assumption. I therefore consider that, even without Shoemaker's restriction, not all conjunctive properties s-realize their conjuncts. Furthermore, if one grants Shoemaker's restriction, this does not affect my argument. As will become clear, it is concerned with establishing that some conjunctive properties are s-realized by their conjuncts. Shoemaker's restriction limits which conjunctive properties s-realize their conjuncts. It does not limit which conjunctive properties are s-realized by their conjuncts. Nor is it designed to do so, for Shoemaker assumes that there are no such cases.
6 This reply was made by Shoemaker in private correspondence.
7 I’m very grateful to E. J. Lowe for helpful discussion on this issue.
8 I’m grateful to a referee for this suggestion.
9 §4 presents a more complex version of the argument that abandons this assumption.
10 See, for example, Armstrong [1980: 30] and Armstrong [1997: 123].
11 Alternatively, one could simply observe that the circuit in Four-light circuit does not have D—there is instead a gap between the bulbs—and appeal to whatever (complex) property of the circuit it is that makes this true. This would work just as well. However, what this property might be is a notoriously contentious issue and one that I don't wish to get into here. For a discussion of the ontology of gaps, see Casati and Varzi [Citation1995].
12 The combination of properties each conjunctive property picks out must, of course, be a combination that it is possible for the circuit to have. However, for this to be a serious objection to my argument, some specific example must be provided to demonstrate that an impossible combination of properties would need to be invoked.
13 Shoemaker used to maintain this theory. (See, for example, Shoemaker [2003a]). However, he considers the subset account to be entirely independent from it [2001: 80].
14 I would like to thank two anonymous referees for their helpful comments and criticisms. I would also like to thank James Clarke, Jonathan Lowe, Sydney Shoemaker and Nick Zangwill for helpful comments on earlier drafts of this paper.