Autonomy in Self-Formation and Moral Responsibility — Epoche C1
Introduction: re-evaluating control and responsibility The question of moral responsibility, and in particular its compatibility with determinism, has long captivated philosophers. Determinism is the thesis that every event, including every human choice, is fixed by prior causes together with the laws of nature: given the past and the laws, only one future is possible. If that is true, it seems no one could ever have acted otherwise than they in fact did — and a venerable principle connects that ability to blame. The Principle of Alternate Possibilities (PAP) states that a person is morally responsible for what she has done only if she could have done otherwise. Incompatibilists accept PAP and conclude that determinism would abolish responsibility; compatibilists deny that responsibility requires the sort of freedom determinism rules out. A prominent compatibilist line of attack, articulated by Harry Frankfurt in a celebrated 1969 paper, strikes at PAP itself: Frankfurt argues that an agent can be morally responsible for an action even when she could not have done otherwise, so that responsibility hinges on the agent's actual reasons for acting rather than on counterfactual alternatives — on what did move her, not on what else might have. The Frankfurt case, in detail Frankfurt's argument turns on a thought experiment whose structure deserves to be laid out fully, since everything depends on it. Suppose Black, a manipulator of great skill — Frankfurt lets him work by threat, hypnosis, drugs, or direct intervention in the brain — wants Jones to perform a certain action, say to cast a particular vote. Black prefers not to show his hand, so he merely monitors Jones. If Jones shows some reliable early sign of being about to decide otherwise, Black will intervene and force the decision he wants. But in the event Jones deliberates and decides, entirely for his own reasons, to do precisely what Black wanted. Black never lifts a finger; his presence makes no difference to what actually happens. Now ask two questions. Could Jones have done otherwise? No — Black's standing disposition guaranteed the outcome either way. Is Jones morally responsible for his vote? Intuitively, yes: he decided for his own reasons, exactly as he would have done had Black never existed. If both answers are right, PAP is false: the ability to do otherwise is not a necessary condition of responsibility. What did the work in our judgement of Jones was not the absence of alternatives — there were none — but the fact that the actual sequence of deliberation and decision was his. Frankfurt's conclusion is that responsibility is a matter of the actual causal history of the action, not of unrealised possibilities. (It should be noted that the argument has been contested: critics such as those discussed at length by Fischer and Ravizza point out that if Jones's world is indeterministic, the 'early sign' cannot guarantee his decision, so a residual alternative — a 'flicker of freedom' — survives; and if his world is deterministic, assuming he is responsible begs the question against the incompatibilist. The present essay grants Frankfurt his conclusion and asks what follows.) Consider the Korean proverb, 'Even a worm will turn' (지렁이도 밟으면 꿈틀한다), which speaks to an inherent will to resist. In a Frankfurtian scenario this 'turning' is rendered moot: the outcome is fixed regardless of the worm's struggle. Yet Frankfurt's point is that the struggle — or its absence — still matters morally. What matters is whether the action flowed from the agent's own will. Ownership of the will: the hierarchical model That phrase, 'the agent's own will', carries the whole weight of the position, and Frankfurt spelt it out in a second paper, 'Freedom of the Will and the Concept of a Person' (1971). There he distinguishes orders of desire. A first-order desire is a desire to do or have something: to smoke, to study, to vote for a candidate. A second-order desire is a desire about one's own desires: the unwilling addict has a first-order desire to take the drug and a second-order desire not to have that desire. When an agent wants a particular first-order desire to be the one that actually moves her to act, Frankfurt calls this a second-order volition , and it is here that he locates personhood and ownership: an agent identifies with a desire when she reflectively endorses it at the higher level. A creature with first-order desires but no reflective stance on them — Frankfurt calls it a 'wanton' — acts, but its actions are not expressions of a self. On this picture, Jones is responsible because the desire that moved him was one he endorsed; the unwilling addict who is moved by a craving she repudiates acts against her own will even though the movement is hers. Responsibility requires acting from a will one has made one's own — coercion and compulsion defeat responsibility not by removing alternatives but by severing the action from what the agent identifies with. The dialectical challenge: where identification runs out While Frankfurt's account effectively challenges the necessity of PAP, it arguably overlooks a deeper dimension of control: autonomy in the formation of one's motivational states. Two classic pressures expose the gap. The first is a regress, pressed by Gary Watson in 'Free Agency' (1975). If first-order desires need the endorsement of second-order volitions to be authoritative, what confers authority on the second-order volitions themselves? They are just more desires. Either they need endorsement from a third level — and so on without end — or the hierarchy stops at states whose credentials are simply assumed. Identification, as a purely structural fact about how an agent's mental states mesh at a moment, cannot by itself certify that the whole structure is the agent's own. The second pressure comes from manipulation cases. Imagine an agent whose entire hierarchy — first-order desires, second-order volitions, the lot — was implanted by c