Kennett, Jeanette & Smith, Michael - Frog and Toad lose control ... Mele, Alfred - Underestimating self-control: Kennett and Smith on Frog and Toad

03/27/2009

Frog & Toad: Analysis Vol 56 No 2 April 1996
Underestimating self-control: Analysis Vol 57 No 2 April 1997

Kennett & Smith:
This paper is written about the philosophical problem of self-control, specifically synchronic resistance, which is considered here by the authors as 'not to do what we most want to do' at the same time as we want to do it. This is separate from diachronic self-control, which is taking action now to resist acting on what you anticipate will be your strongest desire in the future (pg67-8). The premise of the problem is that action is produced from intrinsic desires of a certain strength mixing with beliefs, being transmitted rationally over 'the means-ends relation' into extrinsic desires of the same strength, then the strongest extrinsic desire becoming an action. The question is how you could avoid doing your strongest intrinsic desire; how self-control is possible synchronically.

The first explanation (1.) from Kennett & Smith is that there can be a failure in transmitting intrinsic desires into extrinsic ones: thus a stronger intrinsic desire for good health doesn't become a stronger extrinsic desire to avoid eating cookies (compared to a weaker intrinsic desire for sweets becoming the strongest extrinsic desire). This they consider to be a failure in instrumental rationality.

The second explanation (2.) for self-control is to be 'orthonomous', which is a kind of action that results when one is fully rational, not just instrumentally. To be 'fully rational' one needs to have 'knowledge of all the relevant facts' (pg66). Being orthonomous (having orthonomy) means acting according to what your desires would be if you were fully rational. So since it is fully rational to intrinsically desire good health over sweets, even if one had no extrinsic desires for good health, one could override extrinsic desires for sweets. Authors tell a tale for how this might be psychologically possible (pg66).

The next part of the paper (3.) tries to avoid a contradiction of the following form: I want most to eat sweets and want even more (more than most?) to prevent myself from eating treats. The solution is to treat the exercise of self-control as not a contradictory action (or a contradictory trying) but instead as the occurrence of thoughts and dispositions that allow the transmission of the relevant intrinsic desires for health across the means-ends relation (e.g. having thoughts that picture sweets as disgusting things). This 'doing' is not an action because, authors claim, it doesn't satisfy the standard picture that an action is a caused by a desire to x and then beliefs about how to get x. (pg69) Instead, having such thoughts enable orthonomy or a fix to one's instrumental rationality. (pg69-70)

Mele:
This response takes the premise of the Kennett & Smith paper to task as being faulty. The problem is that it's quite difficult to formulate how exactly the connection between intrinsic desires and intentional action is made. Author argues that Kennett & Smith's claim that 'whenever we do something [and whenever we try do to something] we want to do that thing more than we want to do anything else we can do' (K&S pg63) is just false, and gives a counterexample of sipping tea while reading article x. Certainly he wanted to read article x more than article y, but he didn't want to sip tea more than read article y. Yet he sipped tea nonetheless. The key here is that it is tricky to specify when and how desires will conflict in action and when they won't, and the possibility of instrumental irrationality makes any sort of mix possible. (pg120-121) (Here it seems author argues that because instrumental rationality is possible, so is it possible to desire to eat sweets and also desire to reduce that desire.) So it is certainly possible to have an intrinsic desire to reduce the power of the extrinsic desire of eating sweets (this would count as a second-order desire).

Mele also argues that, given instrumental irrationality, one can have a desire to engage in a 'picturing technique' that then enables synchronic self-control. But this desire would have been intentionally actional, a 'trying', and therefore allows that synchronic self-control can also be actional. (pg122)