Monday 8 March 2010

The M-Rationalist Objection

In my previous post I argued that one may instantiate a transitive inference even if one lacks the concept of a transitive inference. In this respect, I hold that transitive inferences are not like Davidson's conception of belief; for according to Davidson one can only have a belief if one also has the concept of a belief. However, as Ben points out in his comments, the claim that an agent can instantiate a transitive inference without the concept of a transitive inference falls short of the claim I wish to defend—namely, that an agent may instantiate a rational transition even though that agent lacks the concept of a rational transition. One may resist this further step by denying that the transitive inference performed by Hurley’s monkey is a rational transition. This strategy may seem attractive if one buys into the conception of a rational transition suggested by the following passage, due to R. Jay Wallace:
To fix terms, let us say that a process of thought is an instance of reasoning or ratiocination just in case it is governed by the principles or norms of rationality. To say that a principle or norm governs a process of thought is in turn to make an explanatory claim: it is to say, not just that the process of thought is in accordance with the rational principle or norm, but that the process of thought occurs because the person believes it to be in accordance with the principle or norm. [Italics his]
By Wallace’s lights, a psychological transition can only be said to be governed by the principle of transitivity if the agent engages in that transition because she believes it accords with that principle. But an agent can believe that a psychological process accords with the principle of transitivity only if that agent possesses the relevant conceptual resources. Consequently, if an agent lacks the conceptual resources to believe that a particular psychological transition accords with the principle of transitivity, then that agent could not perform an inference that is governed by that principle. If we assume that a psychological transition is rational only if it is governed by a rational principle, then Wallace’s account of what it means for a thought process to be governed by a rational principle entails that a transition is rational only if the agent performing the transition believes it to be such, and is motivated to engage in the transition by that belief. I will refer to his view as Motivational Rationalism (henceforth, M-rationalism). Roughly, M-rationalism amounts to the following claim:
M-rationalism
A psychological transition is rational only if the agent is motivated to complete the transition by her belief that it accords with a rational principle.
Significantly, I have unpacked Wallace’s use of the word ‘because’ in motivational terms. My motivation for doing so has to do with the overall aim of Wallace’s paper; namely, to impugn the Humean thesis that reasons cannot motivate. This aim suggests that Wallace is primarily concerned with motivation and that the ‘because’ he has in mind is motivational. This conclusion is further corroborated by the paragraph immediately following the passage quoted above, in which Wallace writes:
In these terms, the dispute between the Humean and the rationalist is a dispute about the capacity of rational principles or norms to contribute to explanations of motivation. The rationalist holds that such rational principles have a primary role to play in the explanation of motivation, that psychological processes which originally give rise to motivation can be processes which are governed—in the sense I have specified—by the principles or norms of reason.
There are at least two ways of interpreting the necessary condition implicated in the above definition of M-rationalism. The most natural reading is to see M-rationalism as requiring that the agent be motivated to complete some transition, T, by her belief that T accords with a rational principle. On this reading, T itself must feature as the object of the motivating belief. We may refer to this view as Token M-rationalism:
Token M-rationalsim
A psychological transition, T, is rational only if the agent is motivated to complete T by the belief that T accords with a rational principle.
However, we may see M-rationalism as imposing the weaker requirement that the agent have the belief that transitions of a certain type accord with a rational principle, while leaving open whether or not the agent has a belief about the particular transition in question. We may refer to this second weaker reading as Type M-rationalism:
Type M-rationalsim
A psychological transition, T, is rational only if the agent is motivated to complete T by the belief that transitions of type Y accord with a rational principle and T is a transition of type Y.
The distinction between Token and Type M-rationalism is not one that Wallace draws. But textual evidence suggests that he is committed to something along the lines of Token M-rationalism.

According to M-rationalism (in both its Token and Type variations), the transitive inference performed by Hurley’s monkey fails to constitute a rational transition. In future posts I will highlight some of the difficulties with M-rationalism.

No comments: