Affective Dispositions and the Roots of Responsibility — Epoche C2
Two candidates for what an attribution of responsibility tracks When we hold someone responsible for what they did, we may be tracking either of two quite different things: a capacity — the agent's general ability to recognise reasons and to govern conduct by them, a standing competence they would possess whatever they happened to care about — or a commitment , the particular set of concerns and attachments the agent actually has and could not easily be argued out of. The capacity is impersonal and portable; the commitment is not. Which of the two our practices of praise, blame, resentment and gratitude are actually responsive to is the question this essay pursues, and the answer proposed is that the appraisal is of the commitment while the capacity functions only as a condition of eligibility for appraisal at all. A point of scholarly hygiene before the argument. This line of criticism is sometimes framed as a response to a specific "attractor-based" theory of responsibility attribution. There is no canonical statement of such a theory under that name, and the essay should not be built on one. What there is, and what the criticism must engage if it is to bite, is the reasons-responsiveness family of accounts, of which John Martin Fischer and Mark Ravizza's is the most fully worked out. The dynamical vocabulary is worth keeping, because it makes the capacity view's commitments unusually visible, but it should be treated as an idiom rather than as somebody's doctrine. The capacity account, stated so that it can be attacked Fischer and Ravizza's account of moral responsibility rests on what they call guidance control, and its shape is dictated by a prior commitment: they accept Harry Frankfurt's 1969 conclusion that an agent can be responsible for an act they could not have avoided. Responsibility therefore cannot require regulative control — genuine access to alternatives — and must instead be a property of the actual causal sequence. Their two conditions are these. The mechanism issuing the action must be the agent's own , which they analyse historically as a matter of the agent having taken responsibility: coming to see oneself as an agent whose choices have effects in the world, accepting that one is a fair target of others' reactive attitudes on that basis, and holding both beliefs on the strength of the evidence rather than by indoctrination or manipulation. The mechanism must be moderately reasons-responsive : regularly receptive to reasons, meaning that across a suitable range of counterfactual scenarios it exhibits an intelligible pattern of recognising reasons, including at least one moral reason; and at least weakly reactive, meaning that in some scenario where it recognises a sufficient reason to do otherwise, it does otherwise. The asymmetry between "regular" receptivity and "weak" reactivity is deliberate and is the theory's most ingenious feature: recognising reasons must be systematic, or the agent is not tracking a normative landscape at all, whereas acting on them need only be possible once, or every agent who reliably succumbs to temptation would be excused by the reliability of their succumbing. R. Jay Wallace arrives at a structurally similar destination from an entirely different starting point, and it is worth flagging that he does so, because he is sometimes enlisted for the opposite conclusion. Wallace begins from the moral sentiments, treats holding someone responsible as holding them to an expectation, and then asks when doing so is fair . His answer is that it is fair when the agent possesses the general powers of reflective self-control: the power to grasp and apply moral reasons, and the power to regulate behaviour by that grasp. That is a capacity condition. Starting from the emotions does not, by itself, deliver an emotion-centred theory of responsibility. What the dynamical idiom adds, and where it fails An attractor, in the technical sense, is a compact set invariant under a system's dynamics towards which trajectories starting anywhere in some surrounding basin converge. Applied to an agent, the picture is that a person's stable dispositions are attractors of their behavioural dynamics, and that we hold them responsible for outcomes that are attractors of the system they are. The picture has an obvious appeal: it gives the capacity view something to point at that is observable and, in principle, measurable. It fails in three separable ways, and the failures are instructive because they are also failures of the capacity view in its plainer forms. Stability is not sufficient. Frankfurt's unwilling addict is the standing counter-example. He has a first-order desire for the drug, a first-order desire to abstain, and a second-order volition that the desire to abstain be the one that moves him; the addiction defeats the volition, and he takes the drug. His behaviour is maximally stable and maximally predictable — no trajectory converges more reliably. Yet Frankfurt's entire point is that the will that moves him is not his own: he is a bystander to his own action, and the regularity of the outcome does nothing to alter that. Frankfurt's contrast case, the willing addict, has an identical behavioural attractor and an entirely different standing, because he endorses the desire that moves him. If two agents with the same attractor differ in responsibility, the attractor is not what responsibility consists in. Stability is not necessary. A person who has never in their life told a serious lie tells one, once, with full knowledge of what it will cost the person lied to. Nobody treats the singularity of the act as an excuse; if anything the departure from character sharpens the blame, because it cannot be attributed to habit. A single act can be as fully imputable as a pattern. The notion is descriptive and the question is normative. That a trajectory converges is a fact about a system. That it was fair to demand a different trajectory is not a fact of that kind, and