Evolutionary Naturalism and the Normativity of the Mental
I examine the following plausible and oft-defended line of thought:
(1) Rationality is a normative phenomenon.
(2) Facts about intentionality essentially depend on facts about rationality, so intentionality is a normative phenomenon too.
(3) Normative phenomena resist naturalistic treatment.
So, (4) the normative aspect of intentionality precludes a fully naturalistic account of the mind.
I believe the conclusion is mistaken, and I attempt to show this by means of a clarification and assessment of claims (2) and (3).
Unlike many of those who resist the above line of thought, I think (2) can be adequately defended. Its defense depends, in my view, on sufficiently appreciating the dynamic character of intentionality. That is, as worldly conditions and conceptual repertoires change, so too does the intricate causal network in which intentional states figure. A purely causal-dispositionalist theory, however, is unable to say anything about these changes and why they take place; it must simply be updated to take account of them. It is here that the essential role of rationality considerations comes in: the development of an intentional state's causal profile tracks the development of its rational profile. In other words, the dispositionalist would have to await a verdict, so to speak, on a given connection's rational status before knowing whether to include it in a state's defining causal repertoire. (An analogy: a purely dispositionalist approach is akin to defining a biological species in terms of its current structural features, when a species is in fact a dynamic entity whose characteristic structural features change over time).
If (2) is defensible, then the problem must lie in claim (3). That said, I do not attempt to overturn (3) by offering, as others have, a naturalistic account of normativity or rationality as such (e.g. expressivism); in my view, the prospects for a naturalistic account of intentionality are independent of this more general issue. I focus, instead, on two key ways in which (3) is underspecified: first, it needs to be said precisely what is required for a "naturalistic" treatment of some phenomenon, and second, it needs to be said precisely how normativity comes into play in a given instance. Once these details are properly filled in, I claim, (3) is false for the particular case of intentionality.
On the first point, I argue that the standards for naturalistic explanation must be less demanding than full reductionism but more demanding than the "naïve" naturalism of many anti-reductionists. I claim that evolutionary continuity provides the proper naturalistic measure: a phenomenon must be shown to have arisen through a seamless course of evolutionarily explicable transitions. This yields a robust explanatory requirement, but one that is flexible enough to handle cases to which reductionism is ill-suited (like intentionality). Moreover, this brand of naturalism quite clearly calls for an interdisciplinary approach, as I believe naturalism should. In this case, that approach depends on continuing investigation into the evolution of cognition and the brain.
As for the second point, I think the tendency to give (2) a strongly metaphysical reading must be resisted. That is, intentional states should be seen not as fundamentally normative entities (as when beliefs are equated with states of commitment), but as causal entities that are essentially approximations to a dynamic normative ideal. The involvement of that ideal is enough to rule out straightforward reductions - the failure of dispositionalism has already been mentioned, and similar considerations can be used to support a broader anti-reductionist thesis - but the more modest naturalistic demand of evolutionary continuity can still be met. This requires an account of our capacity to construct and utilize normative models in our thinking. If it can be shown how we acquired the ability to run through various courses of reasoning and behavior off-line, and to have the results of that feed back and modify our cognitive dispositions, then we can make sense of our having intentional states, i.e. dynamic causal states that approximate to and are in some sense guided by a normative ideal. I cannot claim to tell the whole naturalistic story, but I hope to have made the project look tractable.