Natural Kinds and the Problem of Complex Essences

Abstract

Natural-kind essentialism faces an important but neglected difficulty: the problem of complex essences (PCE). This is the question of how to account for the unity of an instantiated kind-essence when that essence consists of multiple distinct properties, some of which lack an inherent necessary connection between them. My central goal here is to propose an essentialism-friendly solution to this problem. Along the way I also employ some points from that solution to argue for the necessary truth of essentialism (necessary, that is, in all possible worlds in which there are material objects), and to support the essentialist ontology of laws over and against a major rival.

Notes

¹Recent advocates of kind-essentialism include Elder 2004; 2007, Ellis 2000; 2001; 2002; 2005, Fine 1994, Harré2005, Lowe 2006; 2008, Macdonald 2005, Oderberg 2001; 2007, Wiggins 2001, and Wilkerson 1995; 1998. For essentialist accounts of law, see for instance Balashov 2002, Elder 1994, Freddoso 1986, and Ellis's listed contributions. On the resurgence of biological essentialism consult Devitt 2008, Elder 2008, Walsh 2006, Webster and Goodwin 1996, and, again, Oderberg 2007 and Wilkerson 1998.

²I use an elementary particle and its basic properties here, and (mostly) throughout, not because I consider it improper to speak of essences in higher-level kinds, but for reasons of space—a treatment of chemical and biological kinds would involve us in complications that are best left aside for the present. Moreover, I believe PCE is most evident and acute in the case of fundamental kinds. Focusing on these also allows me to remain neutral on the question of whether the essences of higher-level kinds are reducible to the essences of their constituent microparticles, a subject of disagreement among essentialists.

³Just what kind of necessity is involved here is controversial. Some would identify it as conceptual, others as metaphysical, and still others would wish to equate the two. For my purposes it suffices to say that the relevant necessity—regardless of its precise characterization—is robust, applying in all possible worlds containing solid objects.

⁴Though comparable problems have received some treatment in the substance ontology literature. See Denkel 1997, Loux 1978: 153–66; 2002: 123–35], Robb 2005, and Simons 1994. Something like PCE also shows up in certain discussions of historical essentialists such as Aristotle, Aquinas, and Suárez. See for example Butler 2007, Charles 2001: chaps. 11–13], Kim 2008, Koslicki 2006, Kronen 1991, and Te Velde 1995: 245–52].

⁵Ellis 2001: 69, 73–4] objects to such a use of ‘property’. An electron has a number of essential properties, but one cannot speak correctly of the ‘property of being an electron’.

⁶By ‘appropriate’ here I just mean a question that does not make a category mistake.

⁷This might be more evident in a formalization. Let E stand for ‘electron’, V for ‘achieves the specified velocity upon receiving the specified force’, and A for ‘attracts protons’:

1. (x) (Ex ⊃□ (Vx ∨∼ Vx))

2. (x) (Ex ⊃□ (Ax ∨∼ Ax))

3. ∴ (x) ((Ex ⊃□ (Vx ∨∼ Vx) & (Ax ∨∼ Ax)))

Note that since (Vx ∨∼ Vx) is a theorem, the necessity operator preceding it is strictly redundant, but added for clarity's sake.

⁸Reference to the previous formalization might help. It has just been posited that ‘V’ and ‘A’ are the true answers for the disjunctions given in the previous footnote. So: (x) (Ex ⊃ (Vx & Ax)). Now, since rest mass is the truthmaker for V, let ‘R’ stand for ‘rest mass’. Negative charge is the truthmaker for A, so let ‘N’ stand for ‘negative charge’. We then get: (x) (Ex ⊃ (Rx & Nx)).

⁹This point could be used to argue against the particular brand of bare-particularism advocated by Sider 2006; likewise, against the idea that an object could exist while instantiating only one property, an idea entertained by Schaffer 2003. For reasons of space I will not pursue these arguments here.

¹⁰My thanks to an anonymous referee for drawing my attention to Lowe's treatment of this issue.

¹¹At least, for the sort of laws we are worried about here; for Lowe, distinctively causal laws are a bit different, involving two or more kinds being related causally in some fashion, such as the kind ‘salt’ dissolving in ‘water’.

¹²To be fair, Loux's view that the kind-essence is unanalysable rests on a separate underlying argument, one distinct from our worries about PCE. He takes the basic unity here to be required as part of a solution to the problem of individuation, a problem that goes to the heart of why he favours a primitive substance ontology over and against a bare substratum or bundle theory. See Loux 1978: 153–66; 2002: 123–35]. I lack the space to go into this here. However, I will note that his idea that the kind-essences of objects are self-individuating upon instantiation is not without difficulties; see Macdonald 2005: 129–30]. In addition, it conflicts with much traditional essentialist thinking on the problem of individuation, especially that advocated by the Thomistic school. See Clarke 2001, chap. 6] and Oderberg 2002.

¹³My thanks to an anonymous referee for pressing me on this important point.

¹⁴See Armstrong 1983: 83; 1989a: 94; 1989b: 52; 1997: 86–7], LaBossiere 1994: 370], Martin 1980: 6], and Moreland 1998: 258–61]. Scholastic authors say much the same with respect to prime matter and substantial form: prime matter cannot exist on its own, but must always be informed. See for instance Clarke 2001: 146–7].

¹⁵Some might see me as here criticizing one of the explanatory roles granted to substantial form within Aristotelian/Thomistic substance ontology. However, I would argue that my proposal here is consistent with the understanding of substantial form actually present in Aristotle and Aquinas (or, at least, the latter). Yet this is a question best left for another day.

¹⁶Lange 2002: 224–40] would not approve of this phrase, but the general point being made here is independent of its exact propriety.

¹⁷My thanks to an anonymous referee for raising the following concern.

¹⁸For example: (1) Why must there be any intrinsic truthmakers? Cannot external laws fulfil the requirements here? (2) Would my solution to PCE work in an ontology lacking fundamental kinds? (3) The solution appears to take ‘object’ as a given in ontology. But what of certain bundle theorists who are eliminativists with respect to objects? Is my solution consistent with their ontology?

¹⁹Versions of this paper were presented at the 2008 meeting of the Canadian Society for the History and Philosophy of Science, the 2009 meeting of the American Philosophical Association, Pacific Division, the 2009 Metaphysics of Science Conference at the University of Nottingham (travel to which was made possible by a generous research grant from the Killam Trusts), and colloquia at the universities of Calgary and Gdansk. I would like to thank all those in attendance for their suggestions, and especially my APA commentator, Tarun Menon. My sincere thanks also to Alice Drewery, Marc Ereshefsky, Noa Latham, Jack MacIntosh, Duncan Maclean, and two anonymous referees for their many helpful comments on earlier drafts.