Thu Mar 6, 2014
Howison Library, 4–6 PM
Sergio Tenenbaum (University of Toronto)
When Bratman puts forward his theory of intentions and plans in Intentions, Plans, and Practical Reasoning, he takes himself to be departing from a very entrenched orthodoxy. The orthodoxy is presented in Bratman’s book as the “belief-desire psychology”, a view according to which all reasons and rational requirements stem from our desires. Bratman argues that such views are incomplete; they do not take account of the importance that plans, future-directed intentions, policies, etc. have in the lives of agents with limited resources. More generally, this orthodoxy classifies as instrumentally rational any agent who, in light of her beliefs, chooses in each particular occasion the action that best satisfies her desires. Traditionally there were two very general ways to make this idea more precise. According to the first one, an agent is rational just in case that, given his beliefs, he chooses the best means to bring about what he desires; according to the second, an agent is rational just in case she maximizes expected utility.
Against the orthodoxy, Bratman argues that no theory of instrumental rationality is complete if we do not add to it rational principles governing the formation, maintenance, and execution of future-directed intentions, plans, etc. In fact, Bratman’s target extends much beyond belief-desire psychology, or any view that connects desires with reasons. Surely the more important claim is that any theory of instrumental rationality must provide pride of place to future directed intentions (FDIs).
It is a testimony to the force of Bratman’s arguments that, if anything, his view is now the new orthodoxy. I argue that the new orthodoxy is false and that future-directed intention might have no significant role to play in a theory of instrumental rationality. On the one hand, I agree that something was wrong with the old orthodoxy. However, the problem did not lie with disregarding the role of intentions in an account of practical rationality; rather, the old orthodoxy overlooked important features of the structure of action extended through time. I argue that a more appropriate conception of such actions will (i) show that the principle of instrumental reasoning has some surprising consequences when applied to extended action (ii) classify various things that have traditionally fallen under the separate rubrics of “projects”, “plans”, “general intentions”, etc., as ordinary instances of actions. These results allow me to defend what I call the “Policy as Action Model” (PAM). According to PAM, policies are best understood as ordinary actions, and the rational requirements that apply to such policies (and similar attitudes, such as plans, projects, etc.) are no different than the rational requirements that apply to ordinary, extended actions, such as the action of baking a cake. Finally, I argue that resulting theory of instrumental rationality in relation to policies (and similar attitudes) is quite different and a significant improvement over theories that rely on underivative norms and principles governing FDIs.