Kinds of Kind

Kinds of Kind

Kinds are of many kinds. Sortals are of many sorts. We should not assimilate one to another. Some kinds are obscurely defined. My question concerns mental kinds—what kind of kind are they? First, consider chemical and physical kinds, such as heat, water, and light. The modern theory of these is well known: they are empirically discovered natural kinds with a hidden real essence; the terms denoting them are indexically introduced (see Kripke and Putnam). We can identify these kinds with molecular motion, H2O, and streams of photons, respectively. They are what heat, water, and light really are—objectively, essentially. They are mind-independent and naturally bounded. They are structural, compositional, and corpuscular. For short we could call them “structural kinds”. They contrast with what can be called “phenomenal kinds”: kinds that are defined entirely by appearance not hidden structure. Color kinds would be an example (or color-appearance kinds if you are partial to physicalism about colors). Sensations are clearly phenomenal kinds. Accordingly, sensations are not defined by a hidden real essence that can come apart from sensory appearance. The connection between sensations and brain states is not like the connection between heat and molecular motion; the connection is (apparently) contingent. Therefore, sensation terms are not semantically like terms for structural kinds: appearance cannot come apart from reality in their case. The principle of unity is not a hidden structure but a phenomenal appearance; there is no indexical pointing to a hidden essence. So, it is wrong to assimilate sensation kinds to chemical and physical kinds—semantically, conceptually, and ontologically. They are both natural kinds not artificial kinds, but they are unified differently; their individuation follows different rules.

Does this cover mental kinds, concepts, and words? Is mental individuation entirely phenomenal? That seems doubtful, but what other individuation conditions might there be? Two ideas spring to mind: function and causality. Mental states have a function in the life of the organism; they are biological. Two mental states belong to the same kind if and only if they have the same function. We determine what kind of mental state a state is by ascertaining its biological function. Connectedly, mental states have causal powers—these also help to fix the mental kind. It is the same with organ kinds: they have a function and a causal role (as well as a physiological structure). Biological species kinds operate according to their own rules: here evolutionary origin matters, as do phenotype and genotype. They don’t have phenomenal criteria of identity, though we do need to take account of gross anatomy (as well as internal physiology and genetic endowment). Interbreeding is often supposed a necessary and sufficient condition of species identity. Species don’t really have a function, but types of behavior figure in determining a species. It is clear that they are not simple structural kinds like water, heat, and light. Their classificatory principles are more complex, more varied. Thus, we can say that we have three basic kinds of kind: structural-compositional, phenomenal-functional-causal, and phenotypic-genetic-originative. I haven’t talked about mathematical or ethical or aesthetic kinds, which have their own rules; I am limiting myself to natural kinds in a narrower sense. There is obviously considerable variety even here; it would be wrong to take all kinds to obey the same individuating principles. It used to be that philosophers assimilated chemical and biological kinds to phenomenal kinds, stressing nominal essence; it would be equally wrong to assimilate mental kinds to chemical and biological kinds. Taxonomy is not a homogeneous science. Real essences come in different forms.

In the case of mental kinds, we might feel a sense of incompleteness: have we really got to the heart of the matter? First, how are the three elements related? Is there anything deeper that unifies the phenomenal, functional, and causal? Second, is there a missing ingredient? Is there something about mental kinds that we just don’t know and which helps fix their identity? A panpsychist will presumably think so, given that macro mental kinds depend on micro mental kinds: but we have no knowledge of those micro mental kinds. Also, anyone of mysterian tendencies will wonder if there is some hidden property of the mind or brain that contributes to mental individuation; it may not play the decisive role played by hidden real essence for structural kinds, but it could be a factor in the overall package of individuating conditions. It will no doubt be closely connected to the phenomenal, functional, and causal, but it will require a different conceptual articulation. Thus, we may need to recognize a mysterious element in the determination of mental kinds, possibly having to do with intentionality and logical transitions (many mental states have a logical role as well as a causal role). The main point I want to make is that kinds are of several kinds and we shouldn’t take one kind to be paradigmatic. In particular, not all natural kinds are structural kinds, like chemical and physical kinds.

This has implications for cross-kind connections, especially causal and nomological connections. Suppose a kind from one group causally interacts with a kind from another—as it might be, mind and matter. That will involve one kind of kind interacting with another kind of kind, e.g., tasting water. There might be physical laws relating water (H2O) to physical processes in the body, but it wouldn’t follow that there are (or must be) psychophysical laws, since gustatory kinds are individuated differently from chemical kinds. The same chemical kind may taste differently to different organisms, and different chemical kinds may taste the same. The different kinds are likely to be irreducible to each other, given their different conditions of individuation. How could you reduce mental kinds, individuated as described above, to physical kinds individuated as theyare? Wouldn’t you lose the whole essence? Mental kinds are fixed phenomenally, functionally, causally, and possibly mysteriously; so how can they be identical to physical kinds that are defined by corpuscular structure and composition? The kinds of nature just don’t line up this way. To put it differently, a language comprised only of terms for structural kinds is impoverished compared to a language that includes terms for mental kinds (also biological kinds). You can’t reduce kinds to kinds of different kinds.[1]

[1] This paper goes back to an old paper of mine, “Mental States, Natural Kinds, and Psychophysical Laws” (1978). The background includes Locke, Davidson, Kripke, Putnam, Fodor, Nagel, and others. I now think that talk of natural kinds as defined by a hidden real essence is misleading; rather, there are varieties of natural kinds not all of which are so definable. Biological and mental kinds are natural kinds (not man-made), but they are defined by other criteria. This is the root of irreducibility to physical facts and psychophysical irregularity (lack of “strict laws”).