Davidson, Donald - Problems in the Explanation of Action

10/12/2007

Metaphysics and Morality; Essays in honour of JJC Smart, Petit, Sylvan & Norman eds. Blackwell, 1987

Author starts with discussing his views on intentional actions and actions explained by reference to intentions. Specifically, author investigates the question posed by Wittgenstein: 'What must be added to my arm going up to make it my raising my arm?' Author claims that there is nothing that must be added; first is a discussion that if a raising of an arm is an effect of a previous act of the agent, it seems that previous act must also have a previous act, ad infinitum. The case is simple in the case of the arm raising the arm that is raised. However, what if a rope is tied to a pulley that raises a paralyzed arm (I pull the rope with the other arm)? It seems here there is the pulling of the lever and the raising of the arm-- 2 events. Author rejects this (pg 37), saying the 'two events' are identical. In a sense, the rising of my arm is not part of my raising my arm.

The next difficulty has to do with possible objections raised to this 'identity' thesis that show a disjunct between cause and effect, or at least a period of time and/or space that separates the two, thereby plausibly questioning the identity of the two 'actions'. (e.g. my sending a thank you message and the recipient not getting it until later) (pg 38). Author replies that often causal verbs actually have two (or more) parts where x causes y, then y causes z. This can fix certain space/time problems.

Author moves to a concern for giving explanations for actions. Often the explanation takes the form of describing the intention for the consequence. Author takes back a previous claim he had made in "Actions, Reasons and Causes" where he claimed that 'there were no such states as intending, there were just intentional actions.' Author claims that 'it is not enough to ensure that an action was performed with a certain intention that it was caused by that intention' (pg 39) and gives examples of deviant causal chains, which leads author to conclude that 'concepts of event, cause and intention are inadequate to account for intentional action'.

Forming an intention requires a belief and a 'pro-attitude' or desire. Desires aren't the same as intentions, since, according to the author, a desire is a conditional, dispositional state that can be countermanded, while an intention is 'sandwiched between cause and effect' (pg 41). The complaint now is that it seems that only an explanation of which desires and beliefs formed the intention that is a reason for the action (plays the part of a cause). The problem here is that it seems that the explanation for the cause is dependent on how the events are described and can be considered fragile in this way compared to the hard sciences. Perhaps what is desirable are psycho-laws that, once you specify the belief&desire, lawfully cause the action. Since these haven't been found, author claims this is not a reason to say reasons don't cause actions, rather that Author attempts to fix this problem by saying that causal powers may be mentioned in different explanatory contexts, but in principle instantiate laws (pg 42, bottom). This has come under criticism.

Author considers various attempts to specify reason-explanation laws, that are mostly inadequate (pg 43-4). His conclusion is that 'laws relating the mental and the physical are not like the laws of physics, therefore are not reducible to them' (pg 45) The big complaint (that is relevant to our recent previous readings) is that events causes each other by virtue of the lawlike behavior of properties, but the only real properties are physical. Thus there can't be mental-physical causation. Author then says that there are all sorts of kinds of laws, and with that different kinds of explanatory schemes that will apply according to our interests. This will make different properties causally efficacious. In the physical universe only (free of our interests), all properties cause the effect. (pg 45-6)

Author finishes by saying that there is a further distinction between physics, the special sciences (which may theoretically be ultimately reducible to physics) and reason-explanation in the explanation of action. The distinction between any scientific explanation and reason-explanations, is the normative. Here author identifies the semantic content as vital to explanation, and semantic content is subject to our interests in consistency, correctness, etc. Author ends by arguing against a behavioral system of 'black box' psychology.