Sunday, April 03, 2005
Tiffany's argument for Strong Internalism
Tiffany's argument for strong internalism rests on the following two claims:
- If strong internalism is not true, then whether or not some consideration is a reason will depend upon whether the consideration is compatible with externally warranted deliberative norms.
- Any theory which holds that a consideration's being a reason depends upon the consideration's compatibility with externally warranted deliberative norms will violate the central intuition of internalism.
The argument for 1 is straightforward. Weak Internalism articulates a necessary, but not sufficient condition for some consideration's being a reason for an agent. Whatever the fully specified sufficient condition might be, it will have to be such that it provides deliberative norms for moving from the agent's subjective motivational set to the conclusion that the consideration is (or is not) a reason. Since strong internalism requires that every condition relevant to a consideration's being a reason be contained in the agent's subjective motivational set, it requires a fortiori that the warrant of these deliberative norms be wholly internal. From this it follows that if strong internalism is false, then the warrant of these deliberative norms cannot be wholly internal.
On the basis of this argument I'm inclined to accept 1, though I will admit to some reservations deriving from uncertainty about what the notion of a deliberative norm comes to. For now, let me just note that Tiffany takes himself to be following the account given by David Sobel in "Subjective Accounts of Reasons for Action," Ethics 111 (2001): 461-92. I may return to this point after reading Sobel's paper.
In any case, the key premise of Tiffany's argument is given by 2. In order to evaluate this claim we need to know what Tiffany takes the central intuition of internalism to be, whether this really is the central intuition, and whether an appeal to external warrant violates the central intuition of internalism properly understood.
So, what is the central intuition of internalism thought to be? Explaining this will go easiest if we begin with an example, so consider how things are with an unrepentant sinner. A priest can say to the sinner that he has reason not to sin, but this appeal will get no hold on the sinner unless there is something within the sinner's interior psychological world that the priest can appeal to. So, for example, we might suppose that if the sinner believes (a) that there is an afterlife, (b) that sins will be punished in the afterlife, and (c) that being punished is bad, then the priest will be right that the sinner has a reason not to sin. If, however, the sinner rejects any of a, b, and c, then the priest's appeal will find no purchase and the sinner will not believe that he has a reason not to sin.
The underlying idea here is that we engage in talk about reasons as a way of explaining and influencing the actions of agents. This talk, though, will be empty unless the reasons we appeal to are such that they are actually capable of motivating the agent in question. Put another way, saying that some consideration is a reason for some agent when that consideration is incapable of motivating the agent is saying something false. Korsaard, in a passage I quoted in a recent entry, puts this point in a useful way:
If I judge that some action is right, it is implied that I have, and acknowledge, some motive or reason for performing that action. It is part of the sense of the judgment that a motive is present: if someone agrees that an action is right, but cannot see any motive or reason for doing it, we must suppose, according to these views, that she does not quite know what she means when she agrees that the action is right. On an externalist theory, by contrast, such a conjunction of moral comprehension and total unmotivatedness is perfectly possible: knowledge is one thing and motivation is another.
So far, then, there is the thought that being a reason is strongly tied to motivational efficacy. Tiffany, however, follows a number of theorists in thinking that there is something more to the central intuition of internalism. This something more comes out when we consider that not every consideration which is motivating for an agent is something which the agent would consider to be a reason. Some motivating considerations, the thought goes, are alien to the agent in important ways. In the following passage Tiffany articulates this intuition:
Not just any subjective attitude or desire can place a plausible constraint on reasons for action, for even the internalist would deny that all subjective mental states are reason-generating. Some are best considered as themselves alien and to be resisted, such as Watson's example of the mother who desires to drown her own child or the defeated squash player who desires to smash his racket into the face of his opponent. These desires are experienced as something more like impulses; they are felt as alien, as not coming from one's "true" self.
So, as Tiffany understands it, the central intuition of internalism has two parts. There is, first of all, the motivational component. This is supplemented by the requirement that the considerations in question be connected, in some sense, to deep facts about who the agent is. Put another way, this second requirement holds that in order for a consideration to count as a reason for the agent, it must be such that the agent takes it to be compatible with her most strongly held convictions about herself.
Clearly there is much more that needs to be said to make the second requirement clear and, indeed, one of the main tasks of Tiffany's paper is to clarify this intuition through the exploration of what he calls the alienation constraint. Enough, though, has been put on the table to get a useful understanding of Tiffany's argument for strong internalism.
Tiffany's worry is that any theory which licenses an appeal to externally warranted evaluative standards will necessarily impose constraints on the agent which alienate the agent from her own reasons. If this were so, if the agent had reasons which she was alienated from, then the second requirement would be vitiated.
I am inclined to reject this part of Tiffany's argument, but I'll save my objections for another post. For now, let me note several points that seem correct. First of all, it is right to say that when we are engaging in reason talk our point is to arrive at a particular sort of explanation of the behavior of agents. I also agree that this talk would be empty if it did not connect to the actual motivations of real agents. And, lastly, I am in agreement that considerations which are wholly alienated from an agent's most strongly held convictions cannot ground reasons for him. The upshot of all this is that I agree with Tiffany that the intuitions he appeals to ought to be incorporated in our theory of reasons.