Epicurus on Freedom (review)

Epicurus is usually credited with being the first to recognize, and disavow, determinism as a threat to freedom of the will (see Pamela Huby, “The First Discovery of the Freewill Problem,” Philosophy 42 [1967]: 353–62). This common assumption has recently come under attack by Susanne Bobzien (“Did Epicurus Discover the Free Will Problem?”, Oxford Studies in Ancient Philosophy 19 [2000], 287–337; see also David Furley, “Aristotle and Epicurus on Voluntary Action,” Two Studies in Greek Atomism [Princeton: Princeton University Press, 1967]), and now also by Tim O’Keefe, who, in this rigorously argued but eminently readable book, examines the extant evidence for Epicurus’ views, and concludes that Epicurus was not concerned with the “modern” problem of free will at all.

On the traditional picture, Epicurus recognized that freedom of the will is incompatible with determinism: if the agent’s will is free, then her decisions and choices cannot be fully determined by preceding causes—it must be genuinely open to her to do otherwise than she does. But determinism tells us that every decision and choice is fully determined by preceding causes, which form an uninterrupted causal chain with other factors in the environment and past history of the agent. Hence, determinism and freedom of the will are incompatible. O’Keefe agrees with Bobzien’s contention that although Epicurus rejected the thesis of determinism, it was not because he thought it presented a challenge to freedom of the will—that is, the freedom to decide whether to do or not to do some action, which Bobzien calls “two-sided freedom of the will.” O’Keefe largely agrees with Bobzien’s argument that Epicurus, as well as Aristotle and his other contemporaries, held a conception of agency and moral responsibility which does not require two-sided freedom of the will. Epicurus did not espouse what Bobzien calls the “independent-decision-faculty model of agency,” which sharply distinguishes the deciding self from all factors, both internal and external, which could potentially determine one’s action. On this model, the “free” agent is completely undetermined by all internal and external causal factors. By contrast, Epicurus, like Aristotle, held the “whole-person model of agency,” according to which the agent is identified with the whole person, including her beliefs, memories, character dispositions, desires, and emotions. On this model, it is a necessary condition of my volitions being mine that they be fully caused by me as an agent and determined by the disposition of my mind. We aspire not to be free of any causally determining factors, but rather to act in such a way that our actions are fully determined by the right internal factors, i.e., by reason (note that there is nothing incompatible with determinism in the story so far). As O’Keefe puts it, “it is both much more plausible and more charitable to ascribe to Epicurus a concern to defend rational agency rather than libertarian free will” (8).

Epicurus does go on to reject determinism because of the threat it poses to rational agency; he was worried, according to O’Keefe, about the fatalism which he thought results from logical determinism, in the form of the Principle of Bivalence, according to which every proposition, even propositions about the future, is either true or false. Thus, O’Keefe stresses the parallels between Aristotle’s discussion of bivalence in De Interpretatione, 9 and Cicero’s discussion of Epicurus in his De fato. Epicurus thought that the Principle of Bivalence implies, given an Interentailment Thesis—namely, a bridge thesis between logical and causal determinism—that future propositions are true, and true in virtue of some facts about the present that are sufficient conditions for that future proposition’s being true. Hence, there are facts now that are sufficient to make that future proposition true. But this implies fatalism: our decisions will not make any difference to what happens, we will not be genuine causes of what happens, and our reasoning will not be causally efficacious (see chapter 3 for discussion...