free will problem the problem of the nature of free agency and its relation to the origins and conditions of responsible behavior. For those who contrast ‘free’ with ‘determined’, a central question is whether humans are free in what they do or determined by external events beyond their control. A related concern is whether an agent’s responsibility for an action requires that the agent, the act, or the relevant decision be free. This, in turn, directs attention to action, motivation, deliberation, choice, and intention, and to the exact sense, if any, in which our actions are under our control. Use of ‘free will’ is a matter of traditional nomenclature; it is debated whether freedom is properly ascribed to the will or the agent, or to actions, choices, deliberations, etc.
Controversy over conditions of responsible behavior forms the predominant historical and conceptual background of the free will problem. Most who ascribe moral responsibility acknowledge some sense in which agents must be free in acting as they do; we are not responsible for what we were forced to do or were unable to avoid no matter how hard we tried. But there are differing accounts of moral responsibility and disagreements about the nature and extent of such practical freedom (a notion also important in Kant). Accordingly, the free will problem centers on these questions: Does moral responsibility require any sort of practical freedom? If so, what sort? Are people practically free? Is practical freedom consistent with the antecedent determination of actions, thoughts, and character? There is vivid debate about this last question. Consider a woman deliberating about whom to vote for. From her first-person perspective, she feels free to vote for any candidate and is convinced that the selection is up to her regardless of prior influences. But viewing her eventual behavior as a segment of larger natural and historical processes, many would argue that there are underlying causes determining her choice. With this contrast of intuitions, any attempt to decide whether the voter is free depends on the precise meanings associated with terms like ‘free’, ‘determine’, and ‘up to her’. One thing (event, situation) determines another if the latter is a consequence of it, or necessitated by it, e.g., the voter’s hand movements by her intention. As usually understood, determinism holds that whatever happens is determined by antecedent conditions, where determination is standardly conceived as causation by antecedent events and circumstances. So construed, determinism implies that at any time the future is already fixed and unique, with no possibility of alternative development. Logical versions of determinism declare each future event to be determined by what is already true, specifically, by the truth that it will occur then. Typical theological variants accept the predestination of all circumstances and events inasmuch as a divine being knows in advance (or even from eternity) that they will obtain. Two elements are common to most interpretations of ‘free’. First, freedom requires an absence of determination or certain sorts of determination, and second, one acts and chooses freely only if these endeavors are, properly speaking, one’s own. From here, accounts diverge. Some take freedom (liberty) of indifference or the contingency of alternative courses of action to be critical. Thus, for the woman deliberating about which candidate to select, each choice is an open alternative inasmuch as it is possible but not yet necessitated. Indifference is also construed as motivational equilibrium, a condition some find essential to the idea that a free choice must be rational. Others focus on freedom (liberty) of spontaneity, where the voter is free if she votes as she chooses or desires, a reading that reflects the popular equation of freedom with ‘doing what you want.’ Associated with both analyses is a third by which the woman acts freely if she exercises her control, implying responsiveness to intent as well as both abilities to perform an act and to refrain. A fourth view identifies freedom with autonomy, the voter being autonomous to the extent that her selection is self-determined, e.g., by her character, deeper self, higher values, or informed reason. Though distinct, these conceptions are not incompatible, and many accounts of practical freedom include elements of each.
Determinism poses problems if practical freedom requires contingency (alternate possibilities of action). Incompatibilism maintains that determinism precludes freedom, though incompatibilists differ whether everything is determined. Those who accept determinism thereby endorse hard determinism (associated with eighteenthcentury thinkers like d’Holbach and, recently, certain behaviorists), according to which freedom is an illusion since behavior is brought about by environmental and genetic factors. Some hard determinists also deny the existence of moral responsibility. At the opposite extreme, metaphysical libertarianism asserts that people are free and responsible and, a fortiori, that the past does not determine a unique future – a position some find enhanced by developments in quantum physics. Among adherents of this sort of incompatibilism are those who advocate a freedom of indifference by describing responsible choices as those that are undetermined by antecedent circumstances (Epicureans). To rebut the charge that choices, so construed, are random and not really one’s ‘own,’ it has been suggested that several elements, including an agent’s reasons, delimit the range of possibilities and influence choices without necessitating them (a view held by Leibniz and, recently, by Robert Kane). Libertarians who espouse agency causation, on the other hand, blend contingency with autonomy in characterizing a free choice as one that is determined by the agent who, in turn, is not caused to make it (a view found in Carneades and Reid).
Unwilling to abandon practical freedom yet unable to understand how a lack of determination could be either necessary or desirable for responsibility, many philosophers take practical freedom and responsibility to be consistent with determinism, thereby endorsing compatibilism. Those who also accept determinism advocate what James called soft determinism. Its supporters include some who identify freedom with autonomy (the Stoics, Spinoza) and others who champion freedom of spontaneity (Hobbes, Locke, Hume). The latter speak of liberty as the power of doing or refraining from an action according to what one wills, so that by choosing otherwise one would have done otherwise. An agent fails to have liberty when constrained, that is, when either prevented from acting as one chooses or compelled to act in a manner contrary to what one wills. Extending this model, liberty is also diminished when one is caused to act in a way one would not otherwise prefer, either to avoid a greater danger (coercion) or because there is deliberate interference with the envisioning of alternatives (manipulation). Compatibilists have shown considerable ingenuity in responding to criticisms that they have ignored freedom of choice or the need for open alternatives. Some apply the spontaneity, control, or autonomy models to decisions, so that the voter chooses freely if her decision accords with her desires, is under her control, or conforms to her higher values, deeper character, or informed reason. Others challenge the idea that responsibility requires alternative possibilities of action. The so-called Frankfurt-style cases (developed by Harry G. Frankfurt) are situations where an agent acts in accord with his desires and choices, but because of the presence of a counterfactual intervener – a mechanism that would have prevented the agent from doing any alternative action had he shown signs of acting differently – the agent could not have done otherwise. Frankfurt’s intuition is that the agent is as responsible as he would have been if there were no intervener, and thus that responsible action does not require alternative possibilities. Critics have challenged the details of the Frankfurt-style cases in attempting to undermine the appeal of the intuition. A different compatibilist tactic recognizes the need for open alternatives and employs versions of the indifference model in describing practical freedom. Choices are free if they are contingent relative to certain subsets of circumstances, e.g. those the agent is or claims to be cognizant of, with the openness of alternatives grounded in what one can choose ‘for all one knows.’ Opponents of compatibilism charge that since these refinements leave agents subject to external determination, even by hidden controllers, compatibilism continues to face an insurmountable challenge. Their objections are sometimes summarized by the consequence argument (so called by Peter van Inwagen, who has prominently defended it): if everything were determined by factors beyond one’s control, then one’s acts, choices, and character would also be beyond one’s control, and consequently, agents would never be free and there would be nothing for which they are responsible. Such reasoning usually employs principles asserting the closure of the practical modalities (ability, control, avoidability, inevitability, etc.) under consequence relations. However, there is a reason to suppose that the sort of ability and control required by responsibility involve the agent’s sense of what can be accomplished. Since cognitive states are typically not closed under consequence, the closure principles underlying the consequence argument are disputable. See also ACTION THEORY, CLOSURE , DETER- MINISM , DIMINISHED CAPACITY, MIDDLE KNOWLEDGE , RESPONSIBILITY. T.K.