Smith, Michael - The Explanatory Role Of Being Rational

10/31/2008

This paper starts with a discussion of the classical Humean description of action, as being a belief and a desire put together in a particular causal way to produce an action. Author argues that there is actually another piece to this story, that the agent so doing is instrumentally rational. So, put this way we have:

a desire for an end + a means-ends belief + instrumental rationality = an action

Roughly, this corresponds to Hempel's deductivist claim about what schema action needs to conform to: (pg90)
1- Actor was in situation of type C
2- A was a rational agent
3- In a type C situation, any rational agent will do X
Conclusion: A does X
Author quotes Hempel, who defends this schema. (pg91-2) The instrumental rationality isn't just having the capacity, it's exercising it. (pg94-5)

This approach was criticized by Davidson, who claimed that instrumental rationality is the background assumption of action, not a peculiar constitutive element. (pg96-8) Author responds to this point by discussing a problem that Davidson himself had: that of 'wayward internal causation' (pg100-1). This is where the appropriate belief and desire are formed but something else ends up causing the intended action. The consensus on a fix of this is to use modal propositions about how the agent would have behaved under slight variations in the situation. Author argues this already imports elements of very local instrumental rationality. (pg101-2)

Author wants to expand this local rationality to mean a larger kind of rationality altogether (pg102). For this author uses a thought experiment of a John, who has a desire to be healthy and believes that flexing his bicep will do so, but so will flexing his triceps.(pg102-8) Good instrumental rationality should assign equal preference for either action if both are equally salubrious, lesser if one is lesser, etc.

The second part of the paper considers how rationality might play a non-constitutive part in action, namely playing a part in forming beliefs and forming desires. (pg110-120) Here there can be rational ways to form means-ends beliefs, and less rational ways. But what of desires? Surely a Humean will deny rational ways to form desires. Author here tries to convince us that we can appeal to reason about what we should desire in a non-instrumental way (pg116-121), perhaps showing that some situations are 'reason-giving'. The schema would go as follows:

1- A was in situation of type E
2- A was a rational subject
3- In a situation of type E and rational agent will desire the end that Q
Conclusion: A rationally desired the end that Q