A reason-able ontology
Before one can begin any philosophical endeavour it is important to get clear what sort of explanation we are actually looking for. There has been a core of work done within the philosophy of reasons in recent years which does not seem to know where it is going, nor how we would recognise its destination when it arrived. One need not be scientistic to recognise that one very credible claimant to ontological priority is some form of subatomic particle or fundamental force. Naturally this does not necessarily mark the absolute foundation of the ontic and, if mistaken for anything other than an empirically testable explanation-model, scientism may inadvertently be smuggled in as well. To recognise the limits of scientific enquiry is not some sort of fatal blow to science struck by smug philosophers and theologians who can thereby sweep away the teachings of some of the only disciplines to have actually advanced in the last century; it is simply to posit that there may be some questions which are unanswerable and yet not nonsense.
This last point is important because unless our ontological commitments are clear and well established from the beginning, we can fall into an obscure philosophical pit and end up wandering dead-end tunnels of deranged speculation which have little or nothing to do with the sunlit, empirical world above. Such is much of the talk about reasons in contemporary philosophy. For long years the Humean model of motivation has been the consensus among philosophers of action. This consists in the straightforward and intuitive combination of a desire and a belief: I desire a glass of Malbec and I believe there is a bottle in the kitchen – thus I am motivated to go to the kitchen. This seems perfectly reasonable and accords well with psychology and neurology, albeit merely as a philosopher’s simplification and a rendering into plain English. Recently, however, there has been a general shift towards talk of reasons themselves having priority over the mental states of the Humean model. One argument for this is that, except in cases where we are mistaken, people do not tend to talk in terms of belief and desire, rather they cite facts about the world: “I am going to the kitchen because there is a bottle of Malbec there.” To say “because I believe there is a bottle of Malbec there” seems to imply a certain level of doubt, and “because I desire a glass of Malbec” seems to be implied already without needing to be said.
The usual objection to this account of motivation is that it seems to provide a fact about the world (and the location of bottles of Malbec) without any driving force behind it to actually lead to an action. It seems that without reference to a particular desire I could simply say, “There is a bottle of Malbec in the kitchen – and so what?” and remain seated rather than being motivated to do anything at all. Defenders of this model will cite the difference between my act of desiring and the object of my desire and use other examples to reinforce their intuition: I help a person who is drowning simply because they are in trouble, not because I believe they are in trouble and I desire to help them.
Many people do not share the intuitions of the reasons-centric views and are more inclined to accept the Humean model. On the other hand it is not initially obvious which deserves priority out of a Humean explanation and a reasons explanation, since they both seem to be completely interdependent or even interreducible: I believed there was a bottle of Malbec in the kitchen because there was a bottle of Malbec in the kitchen, but the fact that there was a bottle of Malbec in the kitchen could not motivate me unless I also believed this fact. We are left with a circle of explanations with no clear winner and so there seems to be a deadlock between the Humean and the reasons-theorist.
It is at this point that I reiterate my warnings about the relationship with science. The reasons-theorist will claim that their account is more parsimonious since it relies only on facts and states of affairs, rather than being forced to talk about states of affairs in combination with complex mental states. They have missed the point. A mental state belongs in the same ontological category as the states of affairs which lead to it. Facts are the strange things which seem to require us to inflate our ontology. What kinds of things are facts? At least we can have some sense of what something is to be a mental state and that there is a definite relationship between a mental state and a physical state of affairs. Allowing facts into our ontology leads to an explosion of entities since relational facts and facts about facts can go on indefinitely, and even then it is not clear what the relationship between a fact and a state of affairs is.
Of course, this is not to say that the Humean model is necessarily superior, only that it has the advantage of presupposing only a modest ontology – one which is informed by knowledge of natural science and has a distinct advantage for that reason. It may be that facts deserve their own ontological category, but this seems to be more a result of philosophers talking about talking, rather than talking about reality.