An Asymmetric Approach to Moral Responsibility — Epoche C2
The Control Principle, and the symmetry it looks like it carries Thomas Nagel's formulation of the Control Principle is that people cannot be morally assessed for what is not their fault, or for what is due to factors beyond their control, and his purpose in stating it was not to endorse it but to show that anyone who takes it seriously loses almost everything they wanted it for. In 'Moral Luck', collected in Mortal Questions (1979), he sets out four ways in which morally significant features of a case escape the agent's control. Constitutive luck governs the temperament and capacities one has; circumstantial luck governs which situations one is presented with; resultant luck governs how one's actions turn out; and causal luck governs the antecedent conditions from which one's choices issue. Strip away everything touched by these and the residue on which assessment could rest is vanishingly small — perhaps nothing at all. Nagel's own conclusion is worth stating precisely, because it is often mislaid: he offers no solution. He holds that the problem is a genuine paradox generated by the collision between an internal view of agency, on which we act, and an external view, on which we are among the things that happen, and that it has no resolution of the kind philosophers usually want. What generates the paradox is not the principle alone but an assumption tacitly bundled with it, and that assumption is this essay's subject. The Control Principle is normally read as symmetric: the degree and kind of control required to attribute responsibility is taken to be the same degree and kind whose absence removes it. Read this way the principle's exculpatory use is simply the contrapositive of its attributive use, and nothing else needs saying. The friction shows in cases of resultant luck. Take the essay's example, which is a good one and worth keeping: two generals, alike in planning, intention and execution, one of whom wins decisively while the other is destroyed by weather, by reinforcements nobody could have foreseen, by the failure of a unit to hold. The Sunzi , conventionally dated to the fifth century BCE, is unusually clear about the boundary here. Its chapter on dispositions holds that the skilled commanders of antiquity first made themselves unbeatable and then waited for the enemy to become beatable; invincibility lies in oneself, vulnerability lies in the enemy; and therefore one may know how to win and still be unable to bring it about. That is the Control Principle's own distinction, drawn nearly two and a half millennia before Nagel drew it, and it puts the two generals in exactly the same position with respect to everything either of them controlled. Yet the intuitive verdicts about them diverge, and the symmetric reading has no room for the divergence. Bernard Williams's contribution to the same 1976 exchange, reprinted in Moral Luck (1981), makes the divergence harder to dismiss as a mere error of the spectator. His target is the conception of morality on which moral worth is immune to luck, and his cases are chosen so that the agent's own view is what refuses the immunity. Gauguin abandons his family for painting; whether the decision was justified turns on whether he becomes the painter he might be, which is not his to settle in advance, so that any justification available to him is retrospective. The lorry driver who, through no fault whatever, runs over a child feels something Williams calls agent-regret — a response indexed to having been the one who did it, distinct in kind from the regret of a witness, and not obviously irrational. What matters here is the direction of the evidence: it is the unlucky agent, not the onlooker, who declines to treat "there was nothing you could have done" as closing the matter. Any account that flattens the two generals into equivalence has to explain that away, and it is not obvious that it should be explained away. A correction, and where the asymmetry thesis really comes from A correction must come before the argument, because it changes what the argument can lean on. In its earlier form this essay credited the asymmetry thesis to a 2019 article by S. Matthew Liao entitled 'The Asymmetry of Cognitive Control', said to have appeared in Mind & Language . I have not been able to verify that such an article exists, and nothing below depends on it. The bibliographical point is the smaller one. The larger is that the argument attributed to it would not license the conclusion drawn even if the article were exactly as described. That argument ran: experiments show that our cognitive systems attribute agency readily when actions succeed and scrutinise control rigorously before excusing when they fail; therefore our normative theory of responsibility should be asymmetric in the same way. Findings about how attributions are in fact made are evidence about attributors. A discovered bias in how readily we blame is at least as good a candidate for correction as for codification, and a theory that reads its norms off the dispositions it was built to discipline has given up the office. The thesis itself, however, does not need that support. It is older and better defended. Susan Wolf's 'Asymmetrical Freedom' (1980) argues that the conditions on praiseworthiness and blameworthiness are not mirror images. To be blameworthy, on her account, an agent must have been able to do otherwise — specifically, able to have acted well. To be praiseworthy no such requirement holds: an agent so constituted that they could not have brought themselves to do the vile thing is not thereby disqualified from praise, and we do not think worse of a person for finding cruelty unthinkable. Wolf's diagnosis is that the freedom that matters is the ability to act in accordance with reason and value, so that an agent determined by the good is not unfree in any way that bears on responsibility, whereas an agent determined by anything else is. Here the argument turns, and the turn is the