The Natures of Types and Tokens: On the Metaphysical Commitments of Non-Reductive Physicalism

This point is intimately tied to debates that clearly transcend the boundaries of this paper, such as the debate about ways, such as encoding vs exemplifying, in which an object might be relevantly related to distinct properties (Zalta 1983, 1988; for alternative interpretations of related issues, see Parsons 1980; Mally 1912). Differences between descriptions such as “Being Carbromal (instantiating the type) makes this pill dormitive”, and “Being Carbromal (having this specific Carbromal-token) makes this pill dormitive” are not exhausted by a type-token ambiguity in the position of “being Carbromal”. If, by applying the predicate “_makes a pill dormitive”, we describe part of the respective entities’ natures, and if for each aspect of a type’s nature we get a corresponding description of its tokens’ natures, then we should assume that there is a systematic correspondence between the properties constituting the type’s nature and the properties constituting its tokens’ natures – a correspondence that seemingly leads to an ambiguity in the predicate “_makes this pill dormitive” as well, for otherwise, we would end up with the result that types share their natures, or at least part of their natures, with their tokens.

Note that not every property that belongs to a type’s nature is repeated at the token level – such as being a type (which is, probably, part of the nature of every type). Rather, the “natures” we are talking about here are parts of the nature of a type that distinguish one type from other types.

If this is the correct interpretation, detailed reference to dispositions and categoricity is irrelevant, and it is not the case that, as Schneider puts it, the “devil is in the details of the metaphysics of property natures” (Schneider 2012a, 720).

Note that on purely modal interpretations of “determination”, we would get strictly analogous results.

This is mainly done to avoid complex reformulations for different sorts of cases, where the relata of the determination relation are substances, or events.

An instance of [Schema] that reflects [Principle] would be this: Necessarily, if the proposition that token t has nature Nt is determined by the proposition that type T as nature N_T, and if the proposition that token t* has nature N_t* is determined by the proposition that type T* has nature N_T*, and T ≠ T* then t ≠ t*.

The view as discussed here is relatively close to what Schneider discusses under the heading of a neo-Aristotelian conception of a substance in her (2013, section 4). It should be noted, however, that any weaker version is sufficient for the argument to go through, as long as it includes the assumption that a particularized property of an object is essentially a property of that object, or, even more liberally, as long as it includes the assumption that any token has some feature which is such that (i) it has it essentially, and (ii) it distinguishes it from other tokens of the same type.

An alternative reading would be this: a has F-ness contingently qua being a token of T iff a has F-ness and a is a token of T and it is not the case that (a has F-ness essentially because a is a token of T); but this is not the reading intended by Schneider.

Being an allotrope of carbon is here regarded as a first-order property (an argument for similar cases can be found in van Riel 2012). If you regard it as a second-order property, replace it with the first-order property of being an occurrence of an allotrope of carbon, or construct an analogue, using, say, causing light-red experiences under ideal conditions as an essential property of being light red and a contingent property of redness).

To keep things simple, the quantification should be restricted to essential properties of tokens that do derive from some type, as opposed to those deriving from, say, adhering to a particular substance.

I will not dwell upon her brief discussion of “neo-Aristotelian” accounts; she seems to assume that on this view, the mind just is not identical with its physical or physiological counterpart. I do not see how this is supposed to follow, if we assume that a substance just is an independent entity in a specific sense as, for example, Schnieder (2006) would have it.

These two alternatives are discussed and (correctly) dismissed in Schneider (2013, section 3).

Though this version of NRP may, upon reflection, turn out to clash with other metaphysical theses, such as those which give rise to the problem of mental causation.