1. Reasons are causes.
2. Propositions are not causes.
3. (Therefore) Reasons are not propositions.
4. Reasons are either propositions or the subject’s mental states.
5. (Therefore) Reasons are the subject’s mental states.
Because people seem to think that Davidson's arguments from "Actions, Reasons, and Causes" support the causalist view (i.e., that rationalizing explanations are a species of causal explanation), people seem to think that the first premise in this argument must be true. Once that premise is in place, it is hard to see how one might reject the argument's conclusion. In Justification and the Truth-Connection, I defend the view that reasons aren't the subjects psychological states. I argue that reasons to believe, reasons to act, the reasons for which we believe/the reasons on the basis of which we believe, and the reasons for which we act/the reasons on the basis of which we act are facts. Specifically, they are the facts that agents have in mind when making up their minds about what to do or believe, not facts about their minds when they make up their minds about what to do or believe. I thought I'd write up a post here to try to defend that view. In so doing, I'm trying to show that there's little that supports the standard view in epistemology which says that our beliefs have to be based on our own psychological states. (In his contribution to the Routledge Companion to Epistemology, Neta observes that psychologism about the basis of belief seems to be the only game in town and he seems to credit this to Davidson's influence.)
Remember that there are three ways of reading (1):
1a. The reasons why the subject believes what she does are causes.
1b. The agent’s reasons for believing or what she does are causes.
1c. The reasons that bear on whether to believe what the agent believes are causes.
Typically, people think that it is possible to believe for good reasons. That is, they think that it is possible that the reason for which we believe are good reasons to believe. Thus, I shall assume that the reasons for which we believe belong to the same ontological category as normative reasons that bear on whether we should believe what we do. Since this is a debate about the ontology of normative reasons, the causal argument for psychologism has to establish (1c). If it does so, it does so is indirectly. First, the psychologists argue that explanatory or motivating reasons are causes. Second, they argue that (1c) follows from (1a) or (1b) because it is possible to act and believe for good reasons. Now, if we were feeling generous, we might grant (1a). Explanatory reasons or the reasons why someone acts need not be motivating reasons, the reasons in light of which they acted. Since (1a) does not entail (1b), we can accept (1a) and remain agnostic as to whether (1b) is true. And, if we can accept (1a) while denying that the reasons for which someone acted are psychological states, we can say turn the tables on the psychologists. Since it must be possible that the reasons we act for are good reasons, neither motivating nor normative reasons are psychological states. Thus, the psychologists have to show that (1b) is true. Typically, psychologists say that Davidson showed that motivating reasons are psychological states. In a later post, I shall explain why arguments for (1b) typically undermine the psychologist’s suggestion that (1b) and (1c) are both true. Here, I shall explain why Davidson’s arguments do not support (1b) and so cannot support (1c).
The argument that Davidson was supposed to provide for (1b) is found in “Actions, Reasons, and Causes”, which opens with these remarks: What is the relation between a reason and an action when the reason explains the action by giving the agent’s reason for doing what he did? We may call such explanations rationalizations, and say that the reason rationalizes the action. In this paper I want to defend the ancient – and commonsense – position that rationalization is a species of causal explanation. His aim was to show that the force of the ‘because’ that figures in a rationalization (e.g., “Audrey went outside because she believed Donna was waiting for her”) is the same as the force of the ‘because’ that figures in sentential causal explanations (e.g., “Coop went through the front door because he was pushed”). Davidson’s argument is contained in this passage:
Noting that non-teleological causal explanations do not display the element of justification provided by reasons, some philosophers have concluded that the concept of cause that applies elsewhere cannot apply to the relation between reasons and actions, and that the pattern of justification provides, in the case of reasons, the required explanation. But suppose we grant that reasons alone justify actions in the course of explaining them; it does not follow that the explanation is not also … causal … How about the other claim: that justifying is a kind of explaining, so that the ordinary notion of a cause need not be brought in? Here it is necessary to decide what is being included under justification. It could be taken to cover only … that the agent have certain beliefs and attitudes in the light of which the action is reasonable. But then something essential has certainly been left out, for a person can have a reason for an action, and perform the action, and yet this reason not be the reason why he did it. Central to the relation between a reason and an action it explains is the idea that the agent performed the action because he had the reason.His point was that if we want to understand the difference between (i) simply having reasons that could potentially justify an action but do not move you to act and (ii) acting for those reasons, we have to say that agents act because they have certain reasons. To say that she acted because she had these reasons is to say more than just that she simply had these reasons or had them in mind, for these reasons could be explanatorily idle (e.g., I might desire to amuse my roommate and annoy my neighbors and believe that tap dancing in my boots to Tupac would be a way of fulfilling both desires. If I start dancing, I might do so in order to amuse my roommate and not to annoy the neighbors or might do so in order to annoy my neighbors.). To distinguish cases where reasons are idle from cases in which the reasons are operative, we need to posit some causal difference between the agent’s desires and actions to decide which reasons are operative. Thus, we cannot understand how rationalizing explanations work unless the force of the ‘because’ in a rationalizing explanation is the same as in a causal explanation.
Suppose Davidson is right and rationalizations are causal explanations. What does this tell us about the relation between reasons and causes? Nothing. I realize that many people believe that it shows that reasons are causes, but this simply does not follow. Since it does not show that motivating reasons are the causes of the agent’s action or attitudes, it cannot support the crucial premise in the causal argument for psychologism. Remember that if the argument for psychologism has any hope of success, we have to assume that facts are not causes. If facts are not causes, then causes belong to a different ontological category than the explanantes that figure in rationalizing explanations. This is so even if rationalizing explanations are causal explanations because facts are explanantes and we have stipulated that facts are not causes. The Davidsonian thesis that rationalizing explanations are causal explanations is consistent with one of two views. The first identifies motivating reasons with the subject’s mental states and states that motivating reasons are causes rather than the explanantes of successful causal/rationalizing explanations. The second identifies motivating reasons with the explanantes of successful causal/rationalizing explanations and distinguishes them from the agent’s mental states/the causal antecedents of the agent’s actions. Both of these options are consistent with the conclusion of Davidson’s argument, but the second is incompatible with (1b) and incompatible with psychologism. Thus, even if Davidson’s arguments succeed, they do not support (1b) or psychologism.