Knowledge of People

I

Philosophers such as Elizabeth Anscombe and Donald Davidson have explained that we cannot derive predictions from judgments such as "he boasted from vanity." Such judgments are also the source of countless painful mistakes. However, are they necessarily unreliable? Often enough, even if only gradually and partially, we get people right. Assuming that we do is already assuming that there must be a connection, if not causal then at least casual, between what a person is, what she does, and how she acts; and that we may describe it correctly. If no ironclad generalizations can be made that connect the way a person is, her reasons, and her actions, how might we describe correctness in judgments such as this?

Anscombe believed there is a distinction between describing an action and interpreting it. For instance, "he boasted" would be a description of the action; "from vanity" an interpretation of that action. Like Anscombe, philosophers interested in the question of what (an) action is want to retain this distinction when treating action examples (e.g., omelets, houses, promises, switch flippings, goings-from-here-to-Kathmandu, and so on). They think as if from a spectatorial perspective, resembling Jimmy Stewart's and Grace Kelly's stances in Rear Window. Imperfective correctness seems more than enough to grasp the intentional form. In other words, if we can grasp that a given X is "a"-ing in order to "b" (for instance, breaking eggs in order to make an omelet), we may say we have grasped the action, and, arguably, what an action is. Any further considerations are deemed (as Lisa, Kelly's character in the Hitchcock movie, memorably puts it) "rear-window ethics."

Action theory is an episodic, impersonal genre. What philosophers, in a sense, see occurring ("what happens,"¹ to use Anscombe's formulation), comes up in examples given in action theory as a token of some action type, whose intelligibility is irrespective of considerations concerning the person that performs an action. (Say "John," or "X," is cooking an omelet, building a house, going from here to Kathmandu, and so on.) These action descriptions we are given—think of them as episodes bereft of narrative—give out an action's crude etiological configuration, detached from considerations concerning the protagonist. In the picture of action this manner of depiction suggests the person who does the action has no part in its intelligibility. Think, however, of how unlikely this sounds to nonphilosophers. No person is a given X. No one seriously thinks any given action is something just anyone would, or could, do. Omelets included. On the other hand, "something you might be doing out of generosity and goodness," says Wittgenstein, "is the same as you may be doing out of cowardice or indifference."² Might we say, then, that we have grasped the action (and what an action is) merely by identifying such-and-such movements from T to T₁ as a token of some action type?

There must be a way to describe what it is like to be right about the difference between an act of cowardice and an act of indifference, even if they look alike. As Wittgenstein seems to suggest, judgments like "he boasted from vanity" are possibly true as often as they are false. Truthful or not, they describe the protagonist as a certain kind of person: one whose actions (or many of whose actions) can be accounted relative to more or less fixed, second-natural dispositions: "P did/is doing Q so and so; P is so and so." So I want to call them second-natural judgments. Consider, for instance: "X signing up with a lesser club that pays more is a sign of greed," so "X is greedy"; "X signing up with a lesser club was a promise he made to his father," so "X keeps his promises" and "is a good son." In certain cases, needless to say, more than one reason can explain the action, which can be, all at once, an act of greed, a promise kept, and a display of filial affection. In countless other cases, though, one...