Alexander’s Anti-Determinism

Free will was not invented all at once, but according to Olivier Boulnois, the most decisive step was taken by the later Greek philosopher Alexander of Aphrodisias (late 2nd to early 3rd century CE). Alexander is the first to articulate the metaphysical presuppositions that will be used by medieval and modern defenders of a radically free will, even though there is no phrase corresponding to “free will” in Alexander’s Greek. This line of thought emerges out of a polemic against the Stoic doctrine of an all-inclusive Fate. Here is a commented translation of the first half of Boulnois’s chapter on Alexander in Généalogie de la liberté.

Although “free will” is the standard English translation for the Latin phrase liberum arbitrium, in Latin it means literally “free arbitration”. In the title of Augustine’s famous early treatise De libero arbitrio voluntatis (“On Free Arbitration by the Will”), it is voluntatis that corresponds to will. The Latin arbitrium is an act of arbitration, not an actor. Arbitrio is a grammatical variant that does not change that meaning. The French arbitre on the other hand is a grammatical variant that does change the meaning. It refers to an agent rather than an act. The substitution of “will” in the English takes the further step of localizing and centralizing the agent’s agency into a distinct faculty or power.

My own perspective is that humans are free in several ways, one of which is that we are generally able to act based on an evaluation of alternatives. The evaluation of what is appropriate may be nearly immediate and only implicit, or it may involve an extensive process that develops unevenly, and remains open-ended. At least some of the time, we are also able to step back and expressly consider what criteria are most applicable to whatever we are dealing with. But these are kinds of practical freedom that are neither a metaphysical given nor a universal accomplished fact. Practical freedom is “acquired”, unevenly and provisionally, in a process that is never quite done. It involves feeling as well as thought. But what comes to be called free will is a far stronger claim than this; I think it is too strong.

Boulnois contests Nietzsche’s claim that free will is an invention of Christian theologians. That dubious honor goes to a philosopher, Alexander of Aphrodisias. The first Christian thinker to give a systematic role to free arbitration, the early Greek church father Origen of Alexandria, explicitly references Alexander’s discussion.

Yet while Alexander and the Greek church fathers already radicalize human autexousion (self-authority or self-power), they still treat it as a rational arbitration, grounded in deliberation. (Tertullian, the first Latin Christian theologian, will separately introduce voluntas or will into the Latin theological tradition, as a reified, independent faculty of arbitration. Using terminology from Roman law, Tertullian treats the voluntas as a dominus — absolute master over its own acts — wielding the legal authority (imperium) of a Roman emperor over the soul’s functions. But he still expects the will to ground its decisions in rational deliberation. Later writers explicitly elevate the posited will’s choice above reason, and remove its dependence on deliberation. This makes will into a power capable of choosing arbitrarily.)

According to Boulnois, Alexander’s metaphysicalization of human autexousion and his transformation of prohairesis from the resolution of a deliberation into a punctual “choice” already do the main work of radicalizing human freedom.

“Although Aristotle and the Stoics, each in a very different way, were able to provide a basis for responsibility without appealing to free will, the latter has become the key to human action. Where? How? Why?… For this we must look to Alexander of Aphrodisias” (p. 223, my translation throughout).

Though the Aristotelian commentary tradition did not begin with him, Alexander is the oldest commentator from whom a substantial body of work survived in later times. Relatively speaking, he is also the most purely Aristotelian. Alexander is the Greek commentator most highly esteemed by al-Farabi and the subsequent Arabic tradition. Aquinas defends Alexander’s account of soul and intellect against the criticism of Averroes — somewhat surprisingly, since the Latin tradition generally follows Averroes in rejecting Alexander’s “materialism”.

In comparison to Aristotle and Plato, the Stoics drastically narrow the notion of cause. They count only “productive” causes, and further claim that these can only be physical bodies. The Stoic logos and pneuma that give the world rationality and definite form are understood in terms of a subtle fiery “body” that permeates all things. Meanwhile, effects are said to be predicates that belong to a broader class of incorporeal “sayables” (lekta) that do not have independent existence. From this point of view, existence is only properly said of bodies. The Stoics also follow the Megarians in denying any independent reality to potentiality, a stance that had been sharply criticized by Aristotle.

Living in an era that was philosophically dominated by the Stoic system, Alexander sharply polemicizes against the Stoics. Nonetheless, he is deeply affected by the Stoic notion of causality. He adopts the exact Stoic phrase poietikon aition (making or doing or productive cause) as a replacement for Aristotle’s subtler and non-univocal notion of a “source or cause of being moved, or of being at rest”, which most centrally refers to things like the art of building. It is the Stoic and Alexandrist poietikon aition that was translated to medieval Latin as causa efficiens or efficient cause. While formally defending Aristotle’s view that the first cause acts only as a final cause or good to which things are attracted, Alexander undermines it from within. He basically accepts the Stoic view that to be a cause is to be productive, and in general prioritizes the new “efficient” cause over the final and formal causes that Aristotle regards as more primary. This will carry great weight in neoplatonism, Arabic philosophy, Latin scholasticism, and Western modernity.

Alexander’s readings of Aristotelian texts are also marked by an implicit acceptance of the Stoic idea that philosophy should take the form of a rigorous system. Thus he aims to show that Aristotle too is the author of a system, but a better one. A related claim of Alexander’s — that will be influentially taken up by al-Farabi — is that Aristotle’s Metaphysics should be read as presenting a demonstrative science of substance that yields certainty in knowledge, rather than, say, consisting in a dialectical or hermeneutic inquiry that remains in principle open-ended.

Whatever conclusions we may reach on these matters, Alexander forcefully rejects the central Stoic doctrine of an all-encompassing Fate. Even though Alexander’s Greek has no equivalent for the term “free will”, contemporary specialist scholarship has treated Alexander as the first ethical “libertarian”, or strong advocate of the concept we know as free will. (There are better, more Aristotelian grounds for rejecting Stoic determinism as a theory of causality, but these have been historically much less influential.)

At the same time, nature for Alexander has the same strict determinism that it does for the Stoics — only he exempts human deciders. This creates a sharp split between the human and nature that neither Aristotle nor the Stoics would have endorsed.

“Alexander expressly recognizes the existence of a causal determinism, but he limits it to nature. Like the Stoics, he admits that the same cause in the same circumstances always produces the same effect. In nature, each being necessarily produces its effect in conformity with its nature: ‘fate and nature are identical’. Thus animals move themselves naturally, according to their impulse, and at the same time under the action of fate: they necessarily effectuate the movements that proceed from their proper impulse. — But what is proper to the human, acting according to reason, opens for us another order. For all that falls under art (tekhne), including human action, is not subject to fate. The fundamental opposition is between nature and art, that is to say between necessity and human resolution: natures are in themselves the principle of their movement and of their becoming, while ‘the things produced by art and by resolution (proairesis) have the principle of their movement and their efficient cause outside of themselves, and not in themselves; and in their surging, in one who produces them, a discourse intervenes’ ” (pp. 224-225, emphasis in original throughout).

The strong claim that the same cause in the same circumstances always produces the same effect contradicts Aristotle’s literal text. Aristotle carefully says that what we regard as the “same” cause or reason why usually has the same consequences. This is not just a superficial “statistical” weakening of an otherwise univocal account of causes. Aristotle’s own view of causes as “reasons why” is not only pluralist but inherently non-univocal. The Stoics on the other hand insist that causality must be understood in a univocal way. Alexander is deeply swayed by this.

Alexander has been called an essentialist, in a somewhat derogatory way. Elsewhere this may have to do with an excessive emphasis on substance, which would seem to cut off the most interesting aspects of the Metaphysics. Here it shows up in the way he treats nature, as governed by necessity. To straightforwardly identify nature with necessity understood in a deterministic way is to lose all the “liveliness” that nature and natures have in Aristotle.

“There are two ways of acting: according to nature and according to art, that is to say according to reason (logos). In contrast to the products of nature, the works of art result from a choice operated by their productive cause, the human. These presuppose a power accompanied by logos, that is capable of contraries: to work, we must deliberate. Thus while nature operates necessarily, art comes from choice and from speech. The lyre is not tuned by fate; it could be untuned, or tuned a different way. For Alexander, art is an anti-fate; and it includes human action” (p. 225).

This dichotomy of nature and art in Alexander does not do justice to either nature or art in Aristotle. Nature is not rigidly deterministic, and at the same time there can be no art in Aristotle’s sense without determination. Determination in Aristotle is not deterministic. For Aristotle, to meaningfully and seriously say things and make assertions is not to claim that we have said the last word. On the flip side, not only does art in Aristotle’s sense of techne or technique not imply a deep indeterminism; art or craft does imply a basis in understanding that can be articulated in a definite (but not rigid) way, which is to say a kind of determination.

In Alexander we see, side-by-side as it were, a recognizably Aristotelian register or registers of speech-logos-reason-deliberation on the one hand, and what I would call a Stoic emphasis on production and “productive cause” on the other. This kind of unresolved or only partially resolved duality between textually Aristotelian elements and other non-Aristotelian elements or meanings is a common pattern in scholastic philosophy as well. Recognizably Aristotelian words are used, but given new senses, whose connection with what I would like to call their Aristotelian meanings is tenuous or nonexistent. (I don’t really believe that the real Aristotelian meanings are just there for inspection, waiting to be grabbed, but I do believe it is possible to make increasingly good judgments about them.)

“This analysis has an Aristotelian allure, but in identifying action [praxis] with art, and art with thought, it changes the meaning. For in Aristotle, we cannot operate this identification: certainly ‘what is produced in view of something’ falls under ‘all that can be produced by thought, and all that happens comes from nature’; but nature according to Aristotle is not deterministic. It includes spontaneity (automaton) and chance (tukhe). — Thus the principal effect of the distinction generalized by Alexander is to distinguish between nature and human activity; that will soon be to say, between necessity and choice” (pp. 225-226).

What Boulnois calls the Aristotelian allure of Alexander’s account has to do with its references to textually Aristotelian things like speech, logos, reason, and deliberation. I think it is fairly clear that none of these has anything to do with a radical indeterminism. Meanwhile, the emphasis on production and productive cause (despite its later pervasive adoption) implies a fundamentally non-Aristotelian (more Stoic) view of causes.

Something like this view of cause as essentially productive comes to be treated as canonical in Avicenna, Aquinas, and the scholastic mainstream. Now we have a much earlier example in Alexander. And given the large presence of Alexander in the background to Plotinus’s writing, it seems appropriate to also consider Alexandrist roots of the neoplatonic mainstream’s view of cause as productive. (In the case of neoplatonism, there are also middle Platonic sources independent of Alexander.)

“The question of decision (or of resolution) is henceforth thought starting from an opposition between animality and rationality. It will depend on a cut between animal impulse, considered as deterministic; and human choice, free because it is rational. From then on, the Aristotelian rootedness of action in animal movement is lost. The question is detached from its original grounding” (p. 226).

Choice that is considered free because it is rational is a good thing. It is not at all the inherent disaster that freedom as a metaphysical property of a metaphysical thing called will is. But as Boulnois points out, Alexander combines this with a radical division between human choice and animality. Aristotle says that a man and an ox are animals in the same sense. Our qualification as rational or talking animals in no way conflicts with our status as rational or talking animals. This makes it crystal clear that for Aristotle, animality cannot be equated with irrationality or unfreedom.

“Alexander agrees with the Stoics on the way in which animals move themselves. They receive an impulsion from desire. They orient their movements according to their ‘assent’. The conjunction of impulse and assent constitutes ‘what depends on them’. But their spontaneity is turned toward a single thing, the movement that is proposed to them by their representations: their capacity to act is monovalent. The Stoics draw a compatibilist conclusion: ‘they profess that what happens with the animals in being submitted to fate depends on the animals, but equally that it happens in a necessary way’. For the animals, the Stoics admit at once a power of acting by themselves, and a necessary determination of the impulses: the actions produced by the animals are no less necessary than other events; but they happen through them” (ibid).

(I actually doubt the claim that action in animals is strictly monovalent. The Stoic Chrysippus reportedly offered anecdotal evidence that a dog can be seen doing very simple reasoning about alternatives. Following a scent, the dog in Chrysippus’s anecdote comes to a fork in the road, sniffs one side and finds nothing, then immediately takes the other branch without sniffing it too. My own dog loves to play fetch, but also likes to jump up on the bed and play with the object. But he knows that if he does so, I may end the game. So he grabs the object, looks up toward the bed as if considering jumping up there, then turns back and brings me the object. He may not have concepts like a human does, but this behavior seems to show a practical grasp of elementary negation. He also often heeds a simple “No!”.)

“Does the specificity of the human radically alter the problem? Not for the Stoics, who see in the mind a unity that includes character, beliefs, and desires. Once the resulting impulse is triggered, it is impossible for the agent to act otherwise: there is no ‘alternative possibility’; indeed nothing that could be called a ‘free arbitration’ [libre arbitre]. As a result, there is no essential difference between the animal and the human” (pp. 226-227).

The Stoic concept of irresistible impulse as a kind of triggering cause seems very doubtful to me. (Even my dog visibly resists his impulse to jump up on the bed.) On the other hand, that we are governed by our character, desires, and beliefs seems like a tautology. We attribute character, desires, and beliefs to others based on what they say and do. These are sustained or repeated patterns of appearance.

“For Alexander, on the contrary, it is necessary to highlight this specific difficulty: like other living things, the human acts willingly, but it is proper to her to accomplish the actions that ‘depend on us’. An action depends on us if we are ‘its principle and the efficient cause’. It is necessary to clearly distinguish two concepts that, in Aristotle, are very close: [quote from Alexander:] ‘Willingness [hekousion] is not identical to “what depends on us”. For that action is willing that results from an unconstrained assent, and depending on us is the act that accompanies assent conforming to a reason and a judgment [krisis]. This is why if something depends on us, it is also willing, but not all that is willing depends on us’ ” (p. 227).

That they are not identical is true. But it is an unfortunate accident of history that our English terms “willing” and “voluntary” appear to presuppose a separate faculty of will. The root hekon in Homer applies to an action undertaken gladly or whole-heartedly. Before it became a philosophical term, it was used in Greek law as a basis for culpability. Someone who kills someone hekon is culpable for murder. Even the Latin voluntas originally referred to a specific inclination, intention, or wish, rather than a separate faculty of will.

”The Stoics confuse ‘what depends on us’ with willingness. For animals have voluntary movement, but they do not have what depends on us, the humans. It is proper to the rational adult to give her assent according to reason, by a deliberation that concludes in a judgment. The interiorization already at work in the Stoics changes the sense of the return to Aristotle: what depends on us is no longer a category of actions, but a power in us, reason, which allows us to judge before assenting. ‘Willingness’ is opposed to exterior constraint; ‘what depends on us’ is characterized by rationality, and this is why it is opposed to necessity” (ibid).

Ironically, on this point about what is up to us, Alexander is closer to the Stoics he criticizes than to Aristotle. Alexander accepts the Stoic view that only interior acts of reasoning, judgment, assent, and choice are up to us.

Aristotle is quite different. After a long enumeration of things it does not make sense to deliberate about, he concludes: “none of these things could happen through us. We deliberate about things that are up to us and are matters of action, and these are the ones that are left…. And among human beings, each sort deliberates about the things to be done by its own acts…. So it seems, as was said, that a human being is the source of actions and that deliberation is about the things to be done by oneself” (Nicomachean Ethics book III ch. 3, Sachs tr., pp. 42-43; see Willingness, Deliberation, Choice).

Boulois continues, “Alexander nonetheless admits an influence of nature in the human, under the form of character: the ‘character (ethos) of humans is their daimon, according to Heraclitus, that is to say their nature‘. Each receives from nature a temperament that is proper to her, and this explains why, in the same circumstances, different individuals make different decisions. Nonetheless this intervenes as a disposition, but not as the determining cause” (ibid).

“Alexander of Aphrodisias thus proposes a sort of wager on the autonomy of our resolution (which from now on well deserves the name of free arbitration [libre arbitre]). If we reduce human deliberation to nothing, there is no more merit or possibility of exhortation, and it is no longer necessary to make the least effort. Human life thus ceases to be livable. As a consequence, even if we don’t know that there is the ‘what depends on us’ (houtos eph’ hemin), any more than the contrary, it is better to hold, even if it is an error, that we are masters of our actions, than to admit that we are determined by fate. This point Nietzsche saw well: it is necessary to believe in free will [libre arbitre] because it is necessary for life. This belief is a wager; we admit it as a necessary error” (p. 228).

Alexander associates human resolution or “choice” with a kind of radical contingency. Aristotle on the contrary emphasizes the way in which proairesis intelligibly follows from the results of deliberation.

“Before Alexander of Aphrodisias, middle Platonism had already developed a bivalent concept of ‘what depends on us’. This brought with it a connection between the doctrine of action and the indetermination of the contingent. Alcinous declares in effect: ‘the possible […] is indeterminate’, and according to whether the ‘what depends on us’ inclines toward one side or the other, it becomes true or false. He applies the indetermination of the future contingent (Aristotle, On Interpretation, 9) to the concept of ‘what depends on us’, which thus becomes a faculty of decision. A part of the future being indeterminate, if an action depends on us, it is our resolution that ensures that it occurs, if we choose it, and that it does not occur if we do not choose it. But this implies that our choice is not predetermined. — The identification of ‘what depends on us’ with the Platonic sovereignty of the soul allows an escape from the determinism of desire evoked by Aristotle. This formula is finally summarized in the adage: ‘The soul is without master, and it depends on it to act or not to act,’ which establishes by collage an equivalence between Plato and Aristotle, between the Republic and the Eudemian Ethics, between the sovereignty of the soul and the ambivalence of action. — In the same way, pseudo-Plutarch identifies, within the contingent, that which can just as well occur as not occur, with an action that ‘depends on us and our resolution’: for example, walking” (pp. 228-229, ellipses in original throughout).

“Alexander founds this ambivalence on a doctrine of assent that integrates the Aristotelian analysis of action: the human is distinguished from other living things in that she possesses speech (logos). The latter is capable of contraries, and allows her to judge which things deserve to be chosen:” (p. 229).

[quote from Alexander:] “If, when submitted to an examination, the objects of representation are those that were shown from the beginning, she gives her assent to the representation, and in this way follows her objects, while if these appear different, or if something preferable appears, she chooses it, abandoning what appeared from the beginning to deserve to be chosen (haireton)” (ibid).

Here we see also that Alexander adopts the Stoic idea of what we call representation.

“It is reason (logos) that judges and that entails choice. It surpasses representations and gives proof of their truth. It allows us to prefer what really is better” (p. 230).

“Our mastery [sic] not only concerns our actions, but again our ability to decide: ‘we are at the same time masters of deliberating and of choosing what follows from the deliberation’. What is the sense of this deliberation? ‘The fact that because we have the power of choosing (exousian tes haireseos) actions, […] what appears according to reason as most deserving to be chosen, we choose and accomplish in place of (pro) the other’. Here, prohairesis (resolution, decision), becomes a power of preferential choice: the taking of one thing more than another. Deliberation, because it arbitrates between what we would have done in acceding to the first sensible representation, and what reason presents to us as better, allows us to choose between two eventualities. It opens two possible worlds. It is not a matter of demanding whether the agent could have done otherwise, in general — on the condition that her mental states and the circumstances were different — but: all things being otherwise equal, is there “the possibility of acting otherwise”? For the Stoics do not admit such an alternative possibility: for them, ‘it is impossible, given these circumstances, that we do not act’. Nonetheless it is thus that deliberation founds our capacity to choose between contraries. It gives us a ‘power of choice’. Alexander understands the principle of mastery as applying not to actions taken in a specific sense, but to each action seen in its singularity” (ibid).

This is Alexander’s new idea of proaireseis as a punctual act of choice. Alexander makes the unprecedented claim that in a human, there is an inner power of choice that does not depend on anything.

Not to put too much weight on a word, but I have a sort of Hegelian allergy to claims about mastery and sovereignty. These are too strong. What we have are abilities and capabilities. Mastery is not something we should seek in ethics. Unlimited mastery is what a tyrant wants. It is a defective concept of agency, which wants to insist in an unqualified way on the “independence” of the agent. This is an extreme example of what Hegel calls one-sidedness. As Brandom says, mastery is an evolutionary dead end. But Alexander is asserting here that we — individually, here and now — have just this sort of unlimited mastery over a strictly inner domain in which choices occur.

The ethical notion of agency — that what we do makes a difference in the way that things turn out — needs to be separated from all these supercharged, absolutist notions. This is precisely why the current investigation is so important.

“It is necessary to distinguish two senses of this principle: (1) The principle of mastery [sic] in the generic sense: in general, we have the power to lie or not to lie; when we lie, this depends on us, and when we do not lie, this also depends on us. (And this is the Aristotelian sense.)” (ibid).

This merely generic form of “mastery” that Boulnois attributes to Aristotle does not have the total and unconditional character that Hegel objects to. It is just an ability. Thus for the sake of clarity, I prefer to call it something else.

“(2) The principle of mastery, in the singular sense: here and now, I have the power to lie or not to lie. (And this is the sense retained by Alexander” (pp. 230-231).

This does have a total and unconditional character. It is far more than just an ability.

“The generic principle of mastery [sic] is perfectly compatible with the principle of determination by causes (action is determined by my beliefs and my desires). According to this principle, my action is determined, here and now, by my desires and my beliefs, even if here and now I cannot avoid lying. There is no contradiction between the principle (1): in general I have the power to lie and not to lie, and the principle of determinism, (3): every action is necessarily determined by its causes. Thus the Aristotelian theory of action is compatible with determinism. But there is a contradiction between the singularized principle of mastery (2): here and now, I have the power to lie or not to lie, and the fact (4): here and now (taking into account my beliefs and my desires), I am determined to lie. This is to say that in taking the principle of mastery in the singular sense (2), Alexander rejects determinism” (p. 231).

Here Boulnois speaks rather loosely of “determinism” in relation to Aristotle. But based on other passages, what he really means to attribute to Aristotle is that a) a radical indeterminism of choice is ruled out, and b) there is a relative “determination” by causes or reasons that gives definite form to things, without completely locking them down.

“This singular power gives a new significance to prohairesis: we choose one thing more than another, because reason presents to us not just one action represented as desirable, but a judgment that allows us to see another as preferable. The possibility of choice between actions pertains to our deliberation, itself understood as reason, able to see contraries. This is why our faculty of choosing is ambivalent” (ibid).

Deliberation is a weighing of alternatives. This is something we can do, without any super powers. But in his eagerness to combat Stoicism, Alexander gives deliberation an absolutist foundation it does not need. On the other hand, though, there is nothing obviously wrong with what he says about reason. And the argument does have the “Aristotelian allure” that Boulnois mentioned earlier.

“The root of punishability (indeed of responsibility) resides in the fact that an act depends on us, because we have chosen it: [quote from Alexander:] ‘To which acts do chastisements reasonably apply? To no others than those that are accomplished according to a bad choice (hairesin) that has been made. In consideration of this, in effect those who have the power of choice (exousia tes haireseos), who omit to pose as the aim of their proper actions the good and the law, but who, circumventing this and acting in view of profit or pleasure, do evil’ ” (ibid).

Aristotle is concerned with whether actions are good or bad. Alexander interiorizes this to decisions or choices.

“It is a bad use of the faculty of choice that makes action bad. Compared to Aristotle, Alexander pulls the principle of responsibility back a bit. This interiorization intervenes against the Stoics and those who follow them, because what is deserving of praise or blame is not the action, but the choice to do it. Alexander pulls back again in relation to Aristotle. Like him, he distinguishes willing actions (shared with the animals) from actions taken by choice, for which we have the power to do the contrary (and which are contingent). But from here on, it is always to this second level that praise and blame apply. The agent must respond, less for her action, (because it is hers: Aristotle) than for her choice (because she chose it in contingency: Alexander). Responsibility not only relates the agent to her action, but first of all relates the agent to her decision. Henceforth, responsibility and causality are dissociated. They are no longer two descriptions of the same phenomenon, like two sides of a coin. Quite to the contrary, starting from Alexander, responsibility (having become free will [libre arbitre]), diverges from causal dependency, and even comes to oppose it” (p. 232).

For Alexander, we are responsible for something not because we are its cause, but rather precisely because our choice is exempt from causality. Boulnois rightly identifies this with the later notion of free will, even though Alexander does not speak of will.

“According to Alexander, even one who professes determinism, when she deliberates, cannot avoid asking herself: ‘must I do this, or not do it?’ We have at once the ability to act or not to act (contradiction), and that of doing a thing or its contrary (contrariety). But this double ambivalence of action is rooted in that of choice: we have first of all the power to choose or not choose to act, and to choose an action or its contrary. Faced with the rigid determinism of nature, Alexander opens the way, precisely in the sphere of interiority, to the capacity of choice as a bivalent faculty” (pp. 232-233).

“In Aristotle, action accomplished following a prohairesis is called willing if its principle is interior to the agent. But there are other willing actions (those of children and animals). In Alexander, prohairesis, which has become the faculty of choosing between a thing and its contrary, defines ‘what depends on us’; it is rational and distinct from (animal) willingness” (p. 233).

“The contingency of choice is rooted in an even more fundamental dimension; the logos is capable of contraries: [quote from Alexander:] ‘As for the things that are produced according to the logos, they seem to be produced by the logos insofar as one who does them has also the power (exousia) not to do them (kai tou me poiein). For it seems that what is produced by the artisans, according to their art, is not produced by them according to necessity; assuredly, they do each of these things insofar as they also have the equal power not to do them‘ ” (ibid).

Not only reason, but speech, language, and discourse are enlisted in service of radical contingency. These things do essentially involve contingency — that there be some things that are not set in stone. The fallacy here is to claim that denial of one absolute leaves no alternative but to affirm the opposite absolute. That is just not the case. Everything real is somewhere in between these extremes.

“Choice and art produce action and the work, while they can just as well not produce them. Thanks to discourse, I can say a thing and its negation; indeed I can choose a thing and its negation. Thus speech opens art to contingency, to the equal possibility of contradictories: to act and not to act, to produce and not to produce” (ibid).

That discourse and speech open up a space of contingency is a very valuable insight. It is the presumption that this contingency must be absolute that is wrong.

“Human actions depend on the mastery [sic] that the human has over her movements. The key to this mastery is our capacity to decide according to speech:” (ibid).

Again, let’s replace mastery with ability. Our ability to decide is real, but not absolute. It is an ability to be swayed by reasons, which is the opposite of an arbitrary imposition.

[quote from Alexander:] “These acts depend on us, while it seems that we are masters (kurioi) of accomplishing them and not accomplishing them. It is no longer possible to say that fate is their cause, nor that there are certain principles and certain exterior causes that are predetermining in themselves, in an absolute way, that such a thing is produced or not produced” (ibid).

True, fate is not the cause of our actions. But this doesn’t make us masters. The causes of our actions are reasons and beliefs and feelings, and our weighing of them. A human is a composite whole, not a pure choice operating in a vaccum.

“When an action depends on us, that is to say that we are the masters of accomplishing it or not. The key to the resistance to fate is the bivalence of the disposition of self: what it depends on us to choose, it also depends on us not to choose. It is not sufficient to explain action by impulse and assent, as the Stoics do. This explanation remains monovalent. But it is necessary to root action in reason, as the power (exousia) of contraries. We can do a thing and its contrary, because we can think a thing and its contrary, and choose between them” (p. 234).

Again, the argument is mostly valid. But the power of contraries is not a complete exemption from determination.

“Alexander characterizes ‘what depends on us’ as a positive power to choose contraries. It is a power we have mastery of by proairesis, which rests on reason, capable of contraries, and thus becomes a capacity to choose between them. It ties together two very different theses of Aristotle. The first comes from Metaphysics Theta: ‘All the powers associated with the logos are powers of contraries’. The second is formulated in ethics: among our actions, ‘those […] on which humans can deliberate, are those that depend on us to do or not to do’. The conjunction of two theses (metaphysical and ethical) allows Alexander to affirm that human action depends on our rational power; it is, as a consequence, radically contingent. In making the ambivalence of choice the result of a rational deliberation, Alexander avoids the objection that the doctrine of Epicurus incurred: he does not resort to an indeterminate and irrational variation to explain choice (in an even more obscure way). It is indeed the agent who is the origin of contingency, and precisely insofar as she is rational” (ibid).

This part of Metaphysics Theta is exactly the part that will later be used by Duns Scotus. I’ll do a separate post on it some time. I think this part of Aristotle’s text overemphasizes the importance of binary contraries. There are other texts like this in Aristotle that do not seem to reflect the devastating critique of Platonic division (or classification by dichotomy) that is developed in Book I of Parts of Animals. In this case I am inclined to think that the critique may have been developed late in Aristotle’s life.

Epicurus sought to explain human freedom by positing a random, uncaused “swerve” of the atoms. This collapses too many levels. The higher achievements of humanity are not mere random variations.

“Nonetheless, the problematic field has changed since Aristotle. In spite of his rejection of determinism, Alexander refers to the Stoic problematic, and retains the importance of interiority. For him, ambivalence leads back to the capacity to choose, in each one of us. What depends on us is to act according to our reason and our judgment: ‘what depends on us’ is henceforth identified with proairesis, understood as the interior power to choose or decide following a deliberation” (pp. 234-235).

“But in insisting on this faculty proper to the human, distinct from that of the animals, Alexander makes the boulesis (a wish) into a will. First of all, it is this faculty that centralizes desires and gives them its assent. What is more, it is exercised following a judgment (krisis) and a rational deliberation; it is indeed a ‘rational desire’ — the classic definition of will, from Saint Thomas to Leibniz. And above all, it is not reduced to the activities of the rational soul, because it is posterior to them, and it is the principle of the impulse to act; coming after the assent given to representations, it triggers action; and the body, in its turn, through the mediation of the tendons, obeys it. These are exactly the metaphysical properties of a will” (p. 235).

What Boulnois attributes to Aquinas and Leibniz is the “intellectualist” view of will, which only becomes relevant when the existence of a separate faculty of will is assumed.

“What is more, in contrast to desire according to Aristotle, the impulsive soul is itself unmoved. The psychology of Alexander is thus articulated with his non-determinist doctrine of action: whatever our representations, it is not them that move us; it is necessary that an impulse be joined with them: the ‘rational actions’ (new expression) differ from the movements of animals in that they are triggered by an impulse resulting from a rational judgment — and only those that are rational merit the name of action in the proper sense” (ibid).

Building on Alexander, Plotinus will further generalize the claim that the soul is unmoved. That representations by themselves do not move us seems reasonable. But thoughts, beliefs, and feelings do move us. (This implies that they are more than just representations, which I think is true.)

“Alexander pursues the movement of singularization and interiorization of ethics that was begun by the Stoics. But he makes it a non-determinist interiority, concentrated in the logos, capable of proposing to choice a thing or its contrary” (ibid).

“In this, Alexander of Aphrodisias goes against the grain of Aristotle. Certainly, Aristotle’s Metaphysics Theta starts from the idea that a being capable of speech is capable of contraries. But he concludes that even if there is a modal symmetry between contraries (the one and the other are possible), there is no real symmetry between them: the one is better, and the other is attained by privation; medicine aims at health, but it can also fail. Between possible contraries, it is necessarily our desire that dominates, and that determines one more than the other. In stopping at the premises of reasoning, Alexander profoundly modifies the sense: he leaves aside the necessary determination by desire; indifference between contraries has become the principle of the indetermination of choice. All medieval and modern speculation about the freedom of indifference is no more than a variation on this interpretation of Metaphysics Theta by Alexander” (pp. 235-236).

Free Will in the Cambridge Platonists

Schneewind dedicates the final section of his chapter on the Cambridge Platonists to their views on the vexed topic of free will. There are quite a few interesting nuances here. But it is important to first recall a few generalities, beyond what Schneewind mentions. The following thumbnail sketch is far from complete, but will serve for this discussion.

What is called free will is in general a subject of immense linguistic confusion. Each time the term is used, we need to pause and consider in which of several highly divergent ways it is being said. First of all, there is no such term in ancient Greek. Second, not only have there been a great many highly developed positions on questions related to this, but different authors use key terms like “free” and “will” in quite different ways. We have to be careful when translators use such modern terms to translate ancient authors.

In the Nicomachean Ethics, Aristotle instead uses a number of more specific terms, in ways close to their meanings in ordinary speech. These include “deliberation” (bouleusis), “choice” (prohairesis), and what is “up to us” (eph hemin). He classifies human actions as voluntary (hekousios), involuntary (akousios), or mixed (miktos). Actions may also be considered abstractly (aplos), or in the context of an occasion (kata ton chronon).

What is most relevant here is that for Aristotle, an action is called “voluntary” when it is not forced and is not attributable to ignorance. We are unequivocally morally responsible for voluntary actions; responsible in a weaker sense for mixed actions; and not responsible for involuntary actions. It is my contention that insofar as they are justified, claims about the necessity of something called free will for ethics refer to this relatively common-sensical Aristotelian distinction.

In a much more specialized and systematic way, the late Stoic Epictetus (d. 135 CE) redefines prohairesis (choice) as the one and only thing that is eph hemin (up to us), and therefore makes us who we are. In Stoicism generally, all things are said to be governed by fate, understood in terms of cause and effect. Its main early systematizer, Chrysippus (late 3rd century BCE), is usually understood as advocating a “compatibilist” theory of human freedom. Choice in Stoicism is said to involve an inner hegemonikon or ruling principle in the human that is able to exercise or withhold assent (prokatathesis) to appearances and desires. It does not automatically have power over them, but can in principle discipline itself so as to achieve an inner freedom of choice. Only in the fully realized Stoic sage does it have full control.

Judaism, Christianity, and Islam all broadly emphasize that we are morally responsible for our actions, although their main concern in this seems to be not the dignity of the human, but rather to explain how God can be omnipotent, and yet have no responsibility for evil. Jewish, Eastern Christian, and Islamic traditions all seem to pretty consistently draw the conclusion that humans have what is translated as free will, even though God has foreknowledge of the outcome. But perhaps because it is less controversial in these contexts, the exact meaning of “free” and “will” is left relatively open.

Matters are far more complicated in Western Christian traditions, which came to be dominated by highly nuanced attempts to mediate between conflicting doctrinal concerns. Augustine (354-430) in his early work seems to be a strong advocate of free will, but in his later polemics against Pelagianism, strong doctrines of original sin and the insufficiency of human virtue come to overshadow this. It has been argued that Anselm of Canterbury (1033–1109) was the first defender a fully “libertarian” account of free will in the human. The first universally recognized advocate of such a position is Duns Scotus (d. 1308). Many other writers such as Aquinas make important uses of early Augustine’s term “free will” or liberum arbitrium, while giving it an “intellectualist” slant and taking great care to avoid imputations of Pelagianism.

The Cambridge Platonists sought to heal sectarian divisions in Christianity by promoting a new kind of Christian Platonism. Schneewind previously mentioned that they drew inspiration from the Greek church fathers, as well as Plato and Plotinus. Cudworth wrote a whole book on free will from this point of view.

“Cudworth’s objections to Smith’s consequentialist view of punishment occurs as a passing comment in his discussion of free will. Because our relation to God is such a central issue for the Cambridge thinkers, the complexities of the free-will problem could not be avoided. How can we explain our status as genuine agents, neither mere mechanical transmitters of divine activity nor servile subjects made to comply with commands we would otherwise resist, while acknowledging God’s supremacy and our total dependence on him? Whichcote did not produce any extended philosophical treatment of free will. Smith’s remarks on the issue are more interesting for the attitude they reveal than for the philosophical light they shed on it. More and Cudworth alone grappled with its more difficult aspects. Although Cudworth’s work was not published until the nineteenth century, we can use it to discover the difficulties seen by the most astute philosophical thinker of the Cambridge group as arising from an effort to reconcile the deification of the human, on which they all insisted, with that proper obedience to God, which none of them wished to deny” (Schneewind, The Invention of Autonomy, pp. 210-211).

We saw last time that Cudworth wants to defend the traditional view of reward and punishment in the afterlife, even though he is honest enough to admit that eternal punishment cannot in itself be a good of any kind. He agrees that other actions all aim at some good. But instead of rejecting eternal punishment on that Platonic ground, he wants or feels doctrinally compelled to uphold a traditional view of justice as retribution, alongside his broadly Platonic view of the good.

“As always in this period, the question of human free will leads unavoidably to the question of what divine free will is, or perhaps is not. Does our liberty make us similar to God, or different from him? He sees God as free not because God acts from ‘an absolute will’, doing as he pleases in an arbitrary fashion, but because God always wills what is best, in accordance with his own nature as shown him by his own understanding. So too in us. Liberty is reason, the ‘liberal election of, and complacency in, that which our understandings propound to us as most expedient’. To be free is to act voluntarily. It is doing what we see to be best, even in the particular situation and not only in general. When we do not see clearly what is best, we fluctuate in a kind of indifference and are in suspense about what to do. God is never in this unhappy state. Our indifference is therefore an imperfection. The perfection of freedom is to be moved by the universal good seen clearly in the particular case. It is to be moved wholly by the divine order of goodness in the universe” (p. 211, internal citations omitted throughout).

The idea that God wills what is best and that we should too sounds promising as an alternative to command and obedience, but the reference to expediency is a bit disappointing. This aspect sounds more like another anticipation of utilitarianism than an inspiration from Plato or any other Greek philosopher. Of course, there are broader and narrower conceptions of utility and expediency, and it is the narrower ones that are most troublesome.

Cudworth rejects theological voluntarism, but he wants to defend a version of human free will that is stronger than what he himself thinks is needed for ethics. Here we must pay close attention to exactly what notion of free will is at issue.

That liberty is reason and that God does not act from an absolute will but wills what is best, is good Platonism. That for a human to be free is to act voluntarily rather than involuntarily is good Aristotelianism, even though Cudworth generally wants to avoid Aristotle. That perfect freedom is to be moved by the order of goodness in the universe steers the Christian concern with freedom in a broadly Platonic direction that is a big improvement over the voluntarist emphasis on sheer will. Cudworth’s anti-voluntarist idea that freedom of indifference is an imperfection is original and interesting.

“Why should we ever have thought otherwise? Smith blames it on the Jews. Their notion of legal righteousness, now replaced by the inner righteousness of the Gospels, forced them to believe in a free will so absolute that it could do or refrain from any action of its own power. The law, for them, was merely the ‘object’ on which this power was to be exercised. Compliance earned absolute merit, and nothing more was needed from God than the law’s indication of his will” (ibid).

That in contrast to Christian emphasis on love, Judaism and Islam recognize only Law, is a cliché common among Christians. It is of a piece with other clichés about the moral inferiority of paganism, which are refuted by a better acquaintance with Greek philosophy. Apparently too, it was very common for 17th-century English Christians to characterize theological ideas they did not like as Jewish. All four of the Cambridge Platonists Schneewind discusses reportedly do this at times. In today’s polarized world, one need not be a defender of Zionist divine right to recognize that this is wrong. But apart from the red herring of blaming the evils of voluntarism on the Jews, these Christian Platonists make valid and telling criticisms of theological voluntarism. Schneewind does us a great service by bringing these angles to our attention.

“Smith calls Maimonides as witness to the Jewish belief that man’s ability to remain totally indifferent to good or evil, and to choose either, showed itself only after Adam’s sin. Thus Adam’s Fall occasioned ‘the rise of that Giant-like free will whereby [humans] were enabled to bear up themselves against heaven itself’ and live without any need for grace” (ibid).

Smith’s reference to Maimonides sounds distorted at best, since Maimonides was known as a strong “intellectualist” and not as a voluntarist. The connection between free will and sin has its classic source in Augustine. The earliest appearance of theological voluntarism is in Philo of Alexandria who of course was Jewish, but apparently Philo had essentially no influence on historical Judaism until he was rediscovered in the 19th century. On the other hand, contemporary scholarship finds a very significant influence of Philo on the Alexandrian church fathers like Clement (150-220 CE) and Origen (185-254 CE).

“Once we see that God is not to be served for wages, but out of love, we will see why we need not and should not claim to possess a free will of this kind” (ibid).

God is not to be served for wages, which is to say that for these Christian Platonists, promises of reward in the afterlife offer only motivation of a very inferior kind, compared to the Platonic and more generally Greek-philosophical motivation of seeking the good for its own sake. Cudworth will nonetheless also defend the traditional view of reward and punishment as socially needed, in order to keep the unphilosophical masses in line.

As we saw above, More’s initial explanation of free will uses what we can recognize as the Aristotelian distinction of voluntary from involuntary actions. To this he adds the Stoic notion of assent. Schneewind points out how far this is from late Augustine and Luther.

“More does not find things so simple. If there are any people who by nature always act for the best, they are indeed blessed; but they are in a small minority. More is concerned with the rest of us, who have to struggle to be good. Is there free will, entitling us to merit if we choose rightly? More’s first answer is that action from free will is simply one kind of spontaneous or voluntary action. We act voluntarily when we do what we ourselves see to be best. By contrast, we act from free will when we could, even seeing what is best, either act or refrain from acting. Only some external force or our ignorance can make action involuntary. But it is less clear what, on More’s view, might deprive us of free will. His problem arises because he holds that a truly honest man really cannot choose to do something base and vicious. It would seem that in forbearing, the honest man acts voluntarily but not freely. And although More speaks of ‘this power of not acting, when it regards things that are base’ as a perfection, he finally defines free will as a ‘power of abstaining from ill’. St. Augustine, with Luther following him, had said that since the Fall we are free only to choose what is sinful. More instead cheerfully says we are free only to resist evil — if our character is poor enough for evildoing to be an option for us” (pp. 211-212).

This distinction between acting voluntarily and acting from free will again effectively recovers Aristotle’s ethical criterion of voluntary versus involuntary actions, while appropriately putting aside voluntarist claims about a freedom of indifference. No extravagant metaphysical claims are needed or even relevant for the kind of freedom or “voluntariness” that is relevant to ethical judgments about responsibility for actions. But these Platonists seem to doubt their own claim that ethical goodness is within the reach of all normal adult humans.

“Even this asymmetrical freedom does not leave More happy. One of the objections to allowing it arises from the theory that the will necessarily follows the greatest perceived good. This of course is his own basic view; and he sees that it entails that sin arises from ignorance. That makes sin, by his own account, involuntary. But is it? Are we not all able to know the good? Here More suddenly develops doubts. ‘The bulk of mankind’, he says, ‘see little of themselves [and] can never discover what is the ultimate good’. If this is their own fault, then they are culpable, as having freely willed not to develop their potential insight into the good. But the ability to learn what is good is itself a gift, for which one can claim no merit. Not having it is also not a moral fault, even though its absence makes it impossible for one to be virtuous. ‘But whether any are so utterly deprived of this natural aptitude’, Smith confesses in despair, ‘is to me so hard and perplexing a question that I had rather wholly decline it, than involve myself with such mysteries of providence” (p. 212).

“More’s concern arises out of Smith’s position. Those who are so thoroughly moved by love that they need no law are also those who cannot bring themselves to act basely. For them the kind of free will claimed by the Jews [sic] is not needed, or rather, as More puts it, we should not say of them that they have free will. But More is less willing than Smith to take these blessed few to be meritorious. Are they not so fully tied to the divine order of goodness that they have no real agency of their own? Is it only the imperfect who need freedom in order to be blamed, and perhaps to be controlled by laws and sanctions? More sees the issue but not a solution” (pp. 213-213).

One who is thoroughly moved by love and reason and cannot bring herself to act basely needs nothing else. The reference to “merit” applies to views of Christian salvation that base it on something other than our moral goodness. Schneewind finds something similar in Cudworth, presented with a little more sophistication. Cudworth cannot quite free himself from a retributive concept of justice that runs deep in the Christian tradition, alongside celebration of the new dispensation that is supposed to be based on love. He nonetheless makes the point that contrary to what has been claimed for it, a liberty of indifference has no moral value.

“Cudworth thinks he needs a strong form of free will in order that ‘divine justice retributive, dispensing rewards and punishments’, may have a justifiable sphere. One kind of freedom poses no problem. We can choose between things that do not differ in goodness or badness at all, as when we pick one coin rather than another when someone offers us money. God also possesses this power. Though he always acts for the best, much about the world is in itself indifferent — for example, whether the number of stars is odd or even, or the exact date of the last judgment. But liberty of indifference of this kind makes for neither praise nor blame. Only where we choose what we clearly see to be the worse alternative can we be blamed. And this is where the problem lies” (p. 213).

Perhaps the problem here has to do with the intrusion of questions about efficient causality, conceived as something over and above good intentions and good consequences. On such a theory, we would not deserve credit for the good intentions and good consequences of our actions, unless it could be shown that we were also the efficient cause of those actions. But if a putative showing of this sort comes back to a kind of metaphysical claim that if applicable at all would always be applicable, we would not have added anything to our account of what makes this or that particular action meritorious.

“The ‘common psychology’ is at fault. Either it makes the will always follow the understanding’s judgement of good and ill, in which case the will is necessitated, not free. Or it allows the will to set the understanding to work on specific objects. But then the will must act blindly, and liberty amounts only to ‘mere irrationality and madness itself acting … all human actions’. A blind will independent of knowledge would make virtue and vice as impossible as praise and blame. What psychology must we call upon to allow freedom, and to avoid imputing all moral evildoing to God as the sole agent?” (ibid, ellipses in original).

More precisely, it is a certain received notion of hypostasized will as superior to reason that is the problem. A will that is not guided by understanding can only be blind. A will that is guided by understanding need not be “free” in what I would call the spurious sense of anthropological voluntarism. Cudworth at least begins to find a better model of freedom in the Stoic concept of self-governance, and even anglicizes the Stoic notion of the hegemonikon as superior to the voluntarist concept of will: “the ruling principle is none of these”.

“Cudworth’s answer is suggestive if not wholly clear. The division of the mind into faculties of will and understanding is a mistake: it is the individual as a whole who knows and chooses. The soul has many powers, or levels of activity. Its plastic nature, the source of its basic life functioning, is not within its control at all; desires are not directly under our control; conscience exerts itself whether we will it to or not, and joins the will sometimes in controlling desire. The ruling principle is none of these, nor is it the understanding alone. He uses the Stoic term ‘hegemonicon’ for the governing principle in us, or our self-power. It is, he says, ‘the soul as comprehending itself, all its concerns and interests, its abilities and capacities, and holding itself, as it were, in its own hand … redoubled upon itself more or less, in consideration and deliberation’. It can be self-impairing as well as self-improving, but either way it is that through which we make ourselves what we are” (ibid, ellipses in original).

The reification of a separate faculty of “will” as an internal efficient cause of our choosing as we do does indeed seem to be part of the problem here. On some accounts at least, the Stoic hegemonikon is not a separate faculty acting as a kind of interior cause, but an emergent property of self-governance in the whole human being. It makes good sense that self-governance in a human would be a kind of holistic property, rather than the effect of a discrete cause. This would also be a remote ancestor of something like the emergent unity of apperception in Kant.

The source of the power of assent or non-assent in Stoicism is left relatively open. In Kant, the unity of a unity of apperception transcends the order of factuality. Not only is it not attributable to a separate faculty, it does not name an existing state of affairs either. It is rather a kind of goal toward which we can be seen to tend — in effect, an Aristotelian telos or final cause.

“How does this reflexive hegemonicon operate? It does in the distinctively human world what plastic natures do in the purely material world. It serves as a source of order, under God but acting independently of God. It does not make us indifferent to apparent good and ill. It enables us, however, to consider carefully before we act. Haste is thus the source of blameworthiness. We might always have suspended choice and thought again. Grant that we have this power over ourselves and you grant that we are not always determined by ‘antecedent necessary causes’. Thus in the war between conscience and the passions, the understanding does not inevitably determine the agent one way or the other: ‘the matter wholly depends upon the soul’s hegemonic or power over itself, its exerting itself with more or less force and vigor in resisting the lower affections … this is not a single battle … but commonly a long lasting or continued war’. God praises or blames us as the battle turns out” (pp. 213-214).

In my lexicon at least, reflexivity or self-consciousness in us humans is also not a simple global property that we could be said to simply have or not have in a binary way. It is a matter of nuance and detail. It grows weaker and stronger at different times.

“Cudworth argues that the possession of a ruling power of this kind is a necessary attribute for beings as complex as we are. Without it, our various aspects would not function together to make meaningful action possible. In making us at all, God had to make us free; and he therefore had to make us capable of erring and sinning. It is thus no derogation from God’s power that he created self-acting beings other than himself, nor even that he created them with a freedom — that of choosing a known lesser good — that he himself does not possess. God’s fecundity is such that he makes all the possible kinds of being, even those that are self-acting; and out of them all he creates a harmonious whole, without constantly interfering in the world he has made” (p. 214).

Likewise, the freedom that matters for ethics is not something that we are simply given, or have or do not have. It is always a matter of degree, and it is again a matter of nuance and detail.

“What leads us, then, to use or not use our power over ourselves? How does the hegemonicon reflexively decide whether we shall follow our conscience or passion? On what principle does it accept or reject the promptings to action that come before it? Free choice is not the same, Cudworth insists, as determination by the good. Neither is it the same as chance determination, or pure contingency. We do not make ourselves damnable by ‘the cast of a die’. But on the details of the alternative to these rejected views, showing how self-acting agency is to be understood, Cudworth is silent” (ibid).

If my analysis is at all correct, these are badly framed questions. There is no single efficient cause that gives us, or could give us, freedom or self-consciousness. Our self-governance inheres, in varying degrees, in many different details of how we lead our lives.

Voluntarism

I think most people understand believing in free will simply as believing that humans are able to make genuine choices. This is entirely unobjectionable.

It is a very different matter to claim that the will is innately unconditionally free. That is what I call voluntarism, and “free will in the strong sense”.

This is different from the way the term “voluntarism” is used in the secondary literature on medieval philosophy. There it means the view that there is a separate faculty for choosing that is called will, and that this will is prior to intellect, in the sense of being able to overrule what our intellect tells us is true or right. In this view, will governs intellect rather than being governed by it. It is in this sense, for example, that Aquinas is called an “intellectualist” and not a voluntarist, because he holds that the will is governed by intellect, even though he also defends liberum arbitrium, or free will in the strong sense. (There is a separate question about how to understand complex views like this, which include contrasting elements. My simple answer is that we should give them the benefit of the doubt.)

Leaving aside the expanded concept of responsibility that Brandom recommends in A Spirit of Trust, I hold that moral responsibility is well explained by Aristotle. Aristotle says we are responsible for all “willing” actions, but not for “unwilling” ones. It really is no harder than that.

In ordinary life, without any philosophy, we already distinguish willing actions from unwilling or unintended ones that happen as a result of coercion or ignorance. These distinctions of willingness from unwillingness apply at the level of particular actions. Some things we do willingly, others we do not.

This distinction between willing and unwilling could not even possibly be a matter of metaphysical or anthropological truth, because metaphysical and anthropological conclusions are both too general to support this kind of distinction. A metaphysical or anthropological account could only give us ways of generalizing about action or human action as such. But we do not have responsibility in particular cases because of some general metaphysical or anthropological truth. Rather, we have responsibility in particular cases because of what is true in those cases.

We should notice that Aristotle’s whole discussion of willing and unwilling actions — and of choice — not only does not assume, but does not even mention a separate faculty called “will”, unless this has been introduced by a translator. Moreover, meaningful discussion of any definite will or intent revolves around its particular content, not its source. Meanwhile, the source of choice is well described by Aristotle as “either intellect fused with desire, or desire fused with thinking, and such a source is a human being” (Nicomachean Ethics, Sachs tr., p.104; see also Free Will in Aristotle?)

People should be broadly forgiven for having been brought up with the pervasive inflated language about free will, even though a more precise analysis shows it is not really needed to explain or justify the claim we want to defend: that humans are able to make genuine choices.

This is true especially because there is a whole family of old (largely 19th-century) clichés about free will and determinism that tend to force discussion of these topics into very narrow channels. “Free will” and “determinism” are commonly taken to be two opposite extremes. Either everything is determined in advance, or nothing is. Then we are told that in order to avoid the evils of a total determinism, the only alternative is to believe that we are not determined at all. This is a false dichotomy.

Historically, it was Augustine who coined the original Latin term liberum arbitrium, and he is often credited with originating the notion of what I call “free will in the strong sense”. Boulnois mentions the commentator Alexander of Aphrodisias (2nd-3rd century CE) as an important precursor.

Theological voluntarism is closely bound up what I have called strong omnipotence, for which the earliest source I know is Philo of Alexandria (1st century BCE to 1st CE). Some form of omnipotence is at least verbally affirmed by nearly all later authorities in all three monotheistic traditions. It is even more pervasive than free will in the strong sense. But again, most of those same authorities make other statements which seem to indicate that they do not at all mean really to endorse the extreme consequences that follow from strong omnipotence, or from free will in the strong sense. We should apply a good measure of interpretive charity in these cases.