Wednesday, 14 May 2008

Davidson on Weakness of Will (Part 1)

According to Davidson, some agent a acts incontinently in performing some action x iff:
(D1) a does x intentionally;
(D2) a believes there is an alternative action y open to him; and
(D3) a judges that, all things considered, it would be better to do y than to do x.
For example, we can imagine an agent, let us call him Mr. Smith, who has the all-things-considered judgement that it would be better not to have a third glass of wine, and yet—freely, deliberately, and for a reason—proceeds to have a third glass of wine. Thus described, Mr. Smith would constitute an incontinent agent. Given the above definition of incontinence, the following three principles seem to be in conflict:
(P1) If some agent a wants to do x more that she wants to do y and she believes herself free to do either x or y, then she will intentionally do x if she does either x or y intentionally.

(P2) If a judges that it would be better to do x than to do y, then she wants to do x more than she wants to do y.

(P3) There are incontinent actions.
Davidson’s goal is to show that (P1)-(P3) do not contradict one another, and that we therefore do not need to give any of them up. He maintains that (P1)-(P3) only seem inconsistent because of a tacit assumption that moral principles take the form of universalised conditionals. The first step towards loosening the grip of this idea, according to Davidson, is to observe that judgements of the form ‘x is better than y’ that we make when deliberating about an action are always prima facie judgements. In the Mr. Smith example, the relevant prima facie judgements may be put as follows:
(a) Prima Facie, having a third glass of wine is better than not having a third glass of wine.

(b) Prima Facie, not having a third glass of wine is better than having a third glass of wine.
These judgements are prima facie in the sense that they fall short of the full-blown practical judgements that culminate in action. According to Davidson, prima facie judgements are always relative to some set of reasons in the same manner in which judgements of probability are always relative to a certain body of evidence. Weakness of will only arises in cases in which an agent has reasons for two mutually exclusive courses of action; what Davidson refers to as “cases of moral conflict”:
By a case of moral conflict I mean a case where there are good reasons both for performing an action and for performing one that rules it out (perhaps refraining from the action). There is conflict in this minimal sense whenever the agent is aware of considerations that, taken alone, would lead to mutually incompatible actions. . . Clearly enough, incontinence can exist only when there is conflict in this sense, for the incontinent man holds one course to be better (for a reason) and yet does something else (also for a reason). [Davidson [1969], p. 105, Italics mine]
Thus, by Davidson’s lights, Mr. Smith has reasons in support of both (a) and (b). For example, Mr. Smith’s reasons in support of (a) may include the fact that having a third glass of wine would be pleasant and it would ameliorate his depression. Thus, Mr. Smith thinks that in certain respects having a third glass of wine would be, prima facie, best. Let us call whatever reasons Mr. Smith may have in favour of (a) as r1. However, there are certain other respects in which Mr. Smith thinks that refraining from having a third glass of wine would, prima facie, be best. For example, we may suppose that Mr. Smith believes that having a third glass of wine would leave him feeling hung over and would jeopardise his job interview in the morning. Let us refer to the reasons that Mr. Smith has in favour of (b) as r2. Making good on the idea that both of Mr. Smith’s prima facie judgements are based on reasons we may restate (a) and (b) as follows:
(c) Prima facie, having a third glass of wine is better than not having a third glass of wine, given r1.

(d) Prima facie, not having a third glass of wine is better than having a third glass of wine, given r2.
To say that Mr. Smith has an all-things-considered judgement that it would be best not to have a third glass of wine means that when r1 and r2 are both considered, along with any other relevant factors, Mr. Smith arrives at the conclusion that not having a third glass of wine is best, given all the available reasons. We may express Mr. Smith’s all-things-considered judgement, where “R” represents the total set of reasons Mr. Smith possesses, as follows:
(e) Prima facie, not having a third glass of wine is better than having a third glass of wine, given R.
Since prima facie judgements are conditional, they are not of the right logical form to feature in a universal conditional. However, if we are to maintain that there are intentional actions (something Davidson takes as a truism) then there must also be unconditional judgements which culminate in actions. This leads to what may be described as a bipartite account of practical reasoning. According to Davidson, there are two categories of judgements that fall under the umbrella of practical reasoning: (1) conditional prima facie judgements and (2) unconditional sans phrase judgements. Only the latter implicates the wants and desires of an agent and thus, from a motivational point of view, only the latter can generate action. Since (P1) and (P2) both make reference to the wants of an agent, they must both implicate unconditional sans phrase judgements. The all-things-considered judgement of the incontinent agent (which figures in (P3)) are not connected to the agent’s wants or desires and is therefore an example of a conditional prima facie judgement. Such judgements do not culminate in action. Since the types of judgements that figure in (P1) and (P2) on the one hand, are different from the types of judgements that figure in (P3), on the other, there is no inconsistency in (P1)-(P3).


Roman Altshuler said...

Hi Avery, thanks for this. It's been a while since I've read the paper, and it's nice to have a compressed summary of what is--in Davidson's hands--a bit of a mess.

I am curious what you make of P2. It doesn't strike me as self-evident, the way Davidson claims; actually, it strikes me as self-evidently false. I judge that, all things considered, it would be better for me to work on my dissertation than look through blogs. But I want to look at blogs rather than work on my dissertation. In fact, judging something to be good while having no desire whatsoever to do it is probably far more common than the option given in P2. And since the difficulty with incontinence (at least in this essay) rests on P2, I wonder whether the difficulty isn't entirely artificial. (Incidentally: the little clause you take out of the Davidson quote seems crucial in this regard--it's the bit where Davidson claims that feelings of strife are unnecessary to the conflict.)


I think you’re right when you point out that common sense favours a distinction between judging that a course of action is best and wanting to pursue that course of action; between evaluation and motivation. It seems to me that most of the support for (P2) is theoretical and is due to considerations going at least as far back as Socrates in the Protagoras. The basic idea (in the Protagoras) is that to desire something is to (in some sense) make the evaluative judgement that it is good. Thus put, the Socratic picture does not seem all that implausible to me, even if it may go against our pre-theoretical or unreflective assessment of the matter.

As for Davidson, I believe he sees (P2) as important if we are to regard the actions of the akrates as intentional. On this view, to say that an agent does x intentionally means that her desire to do x implicates an evaluative judgement in favour of x. For example, consider your textbook non-intentional action: I reflexively pull my hand away from a hot stove. Such an action does not involve any judgement; I move my hand without seeing any value in the action. Moreover, it would be misleading to say that I moved my hand because I desired not to get burned, or that I was motivated to move my hand by my belief that the stove was hot. Talk of judging, desire, and motivation seem out of place in the reflex case. What distinguishes the reflex case from genuine intentional action, then, is that my desire for the end implicates an evaluative judgement in favour of that end. If moving my hand were intentional, then it would mean that I attached some positive value towards moving my hand. That just is to say that in some sense I judged moving my hand to be good. If this is right (and Davidson seems to think it is), then we need something like (P2) if we are to hold that the akrates acts intentionally and not from a compulsion (or what Aristotle calls, “impetuousity”).

Roman Altshuler said...

Davidson wrote the following in "Actions, Reasons, and Causes": "pro attitudes must not be taken for convictions, however temporary, that every action of a certain kind ought to be performed, is worth performing, or is, all things considered, desirable. On the contrary, a man may all his life have a yen, say, to drink a can of paint, without ever, even at the moment he yields, believing it would be worth doing." This seems quite right to me. Of course this claim is perfectly compatible with everything in the akrasia essay, since Davidson's point is that desiring X involves an evaluation that X is prima facie desirable, but not necessarily that X is desirable all things considered.

But here's the problem: The thought "I desire X, but X is not desirable" is perfectly coherent; I don't think there is even a tinge of paradox around it. The supposed paradox derives precisely from the thesis that desiring involves a judgment of desirability. The problem, I think, is that that thesis is introduced in order to figure out how to integrate desires into rationality; it is not, in other words, a thesis about desires, but rather a thesis about what desires must be like if we are to see the agents who act on them as rational, and rational in a strong sense. As a result, the thesis strikes me as fairly artificial.