Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 20 October 2019

Freedom of the Will

Abstract and Keywords

Freedom was a central notion within medieval psychological and ethical theory, and mattered as a metaphysical power, as a right, and as a desirable goal or condition. This article argues that the fundamental notion was of freedom as a metaphysical power to determine how people acted, a power which was seen as primarily enjoyed and exercised within the rational or intellectual mind, on most views at the point of the will, the capacity for choice and decision. In modern philosophy, the metaphysics and ethics of freedom are understood very differently, and are linked to very different theories of the mind and of human action. There is skepticism regarding freedom as a metaphysical power. The discussion examines the will and human action, freedom as a metaphysical power, the ethical significance of freedom, the metaphysics of freedom, and the transformation of ethics.

Keywords: moral obligation, metaphysical power, free will, ethical theory

1. Introduction

Freedom was a central notion within medieval psychological and ethical theory. It mattered as a metaphysical power, as a right and as a desirable goal or condition. But the fundamental notion was of freedom as a metaphysical power to determine for ourselves how we acted, a power that was seen as primarily enjoyed and exercised within the rational or intellectual mind, on most views at the point of the will, our capacity for choice and decision. Thanks to the dominant model of human action in the Middle Ages, freedom as a metaphysical power was characteristically seen as a freedom specifically of the will.

The will was a faculty for choice or decision that came to us with the intellect, the faculty for knowledge and cognition, and it was at the point of the will that the human capacity for action and so for freedom was immediately exercised. Will and intellect differentiated us from the lower, non-rational animals and left it true that we were made in the image of God. Like the intellect, the will was a faculty that operated without any bodily organ and, with the intellect, could survive bodily death.

Because the will was metaphysically free, it could be governed by and answerable to a moral law and so bound by moral obligation. This meant that our exercise of the will was the basis of merit or desert of reward or punishment, and it determined our ultimate destiny after bodily death.

The will’s metaphysical freedom could be used to explain and provide a basis for freedom as a value within ethical theory. Freedom as a right could be understood as a normative recognition of the metaphysical power in the form of a right to its unimpeded exercise. And freedom as a desirable goal or condition of liberation could be understood as a state in which the metaphysical power was perfected.

(p. 570) In modern philosophy the metaphysics and ethics of freedom are understood very differently and are linked to very different theories of the mind and of human action. There is scepticism regarding freedom as a metaphysical power; or if the existence of such a power is admitted, it is often given a reductive treatment in terms of other kinds of power found in wider nature. The primary locus of the power of freedom, as of action, is no longer the will, but what Hobbes termed voluntary or willed action—something that medieval philosophers regarded as a merely secondary form of action. And ideas of freedom as a right and as a desirable state of liberation are generally explained other than by reference to freedom as a metaphysical power—usually by reference to a conception of reason as a locus of autonomy.

2. The Will and Human Action

Medieval theories of human decision making and action assimilated a rich body of theory from Antiquity. In particular, from the thirteenth century on these theories were shaped by reflection on the newly reappropriated work of Aristotle, albeit an Aristotle read through a Latin philosophical vocabulary already embodying conceptions from Augustine and the Stoics.

The will was a motivational or appetitive faculty or capacity distinguished by its relation to reason and the intellect. Whereas our passions were shared with the lower animals, the motivations of the will were peculiar to rational or intellectual beings such as ourselves and the angels. Motivations of the will were directly responsive to cognitions of the intellect involving general concepts and, in particular, a general concept of the good, whereas passions, by contrast, were non-intellectual motivations responsive only to the senses and the imagination and their presentations of particulars—presentations that involved no conceptual generality. As befitting its intellectual grasp of general concepts, the will’s operation was non-corporeal, whereas passions were corporeal states.

As a capacity for intellectual motivation, the will also constituted our capacity for action in fully intentional or deliberate form. Action primarily occurred in actions of the will—such as in acts of choice or electio. What general conception of action lay behind this location of action in the will?

Intentional action occurs as the deliberately purposive; it involves goal directedness—the deliberate use of means to attain ends. So a central feature of any theory of action will be its account of purposiveness. And it was through its understanding of purposiveness in fully intentional or deliberate form—purposiveness as it was found in adult human agency—that the scholastic tradition located action primarily in the will.

In the Middle Ages, there was a widespread consensus about how such purposiveness was to be understood—a consensus that went back to Augustine and the Stoics and that, though less clearly present in Aristotle himself, was naturally read into his texts by medieval philosophers. Purposiveness, and with it intentional action, was (p. 571) understood to consist in a distinctively practice- or agency-constitutive mode of exercising reason. In exercising reason we respond in various ways to an object of thought. In exercising the intellect, we respond to truth, that is, to objects of thought—things believed—as true. In exercising the will, we respond to objects of thought—things willed or favoured—as good. More specifically, in willing outcomes we respond to objects of thought as goals to be attained—not only as goods, but as goods to be attained through our willing of them. We have then a model of the purposiveness constitutive of action that is practical reason based—as consisting in a distinctively practice-constitutive exercise of our capacity for rationality, to direct ourselves towards objects of thought as goals.1 The primary locus of action so understood was in decisions or choices of the will, which were termed actus elicitielicited actions of the will itself.

In some cases the action of the agent ended with these—as when the immediate object of the will was not a further willed action of the agent, but an object beyond the agent.2 For the will was a locus of emotion in intellectual form, as in the case of the intellectual love of God and neighbour, where the will’s object was another person. And such emotion was itself seen as a form of free action. But in other cases, the object of the will was further willed action—the operation of other capacities outside the will. Thus, I might will or decide to give alms, alms giving involving the extending of the hand, and so the exercise beyond the will of a corporeal capacity for motion. These further willed actions were generally termed actus imperaticommanded or imperated actions, or external actions, by contrast to the internal or elicited actions of the will itself. Externality here meant externality to the will, but not necessarily externally observable behaviour. For external or imperated action might also involve something that, though external to the will itself, was still mental or ‘inner’, such as the willed formation of a belief within the intellect, as in the occurrence of a meritorious and saving act of faith. Together elicited and imperated actions made up the category of the voluntarium, elicited acts constituting direct cases of agency, imperated acts only indirect cases, through their relation to the elicited acts that motivated them.

Why were elicited actions the only direct case of agency? That was because it was in the will that the defining case of agency—intellectual direction towards an object of thought as a goal—was immediately to be found. In imperated action considered in itself, there was, in the case of belief, only intellectual direction towards an object of thought as true, or else there was only the non-intellectual operation of a merely corporeal capacity. Goal-directed action could still be found in these imperated cases, but only derivatively, insofar as the imperated action was willed and motivated by a prior elicited action, inheriting the goal of that elicited action. Thus, I might choose to cross the road as a means to reaching the market, and my resultant willed road crossing would be an imperated action aimed at the same object of reaching the market through crossing the road.

Fundamental to the explanation of action in terms of its goal or purpose was formal causation. For the purpose of an elicited action of the will was given by the object of thought that informed it as its goal—an object of thought that then derivatively informed the further imperated act motivated by that elicited act. Aquinas described (p. 572) the completed human action as a form–matter complex, with the imperated act of a capacity external to the will standing as matter to the form provided by the elicited action of the will itself.3

This theory of action is a theory of purposiveness—of the relation of an action to its goal or object. It is to be distinguished from a theory of freedom, which is a theory of the capacity or power of an agent to determine for himself how he acts. For that second kind of theory is concerned with the importantly different relation of an agent to his action. And the relation of an action to its object is not obviously the same as that of an agent to the action he determines. The object of an action is an object of thought that, like an object of belief or of desire, may or may not be actual, whereas the relation of a free agent to his action is a relation between the exerciser of a power and an outcome that the power determines, where both entities must be actual. And it is far from obvious that each of these two very different relations involving, besides the action, two very different relata, should be constituted in the same way or explained by one and the same theory. Nevertheless, a fundamental difference between medieval and modern conceptions of freedom lies in a change in the relation between the theory of purposiveness and the theory of freedom.

Medieval theories of freedom certainly built on and presupposed the dominant practical reason-based theory of action, but they did not generally try to explain freedom in terms of the account that had been given of purposiveness. Whereas we shall find, especially in British philosophy following Hobbes, a tendency to use a radically new account of purposiveness that appealed exclusively to efficient causation also to provide a new account of freedom as a metaphysical power. A new theory of the relation of an action to its object was used also to explain the relation of an agent to the action he determined.

In medieval theories of freedom, there was a wide consensus about the constitution of action. But as we are about to see, because the account given of freedom was not straightforwardly adapted from the account of purposiveness, there was much less consensus about the constitution of freedom. It is true that the competing medieval accounts of freedom often involved disagreement about psychological questions, such as about the precise relation of the will to the intellect. Was the operation of the will closely tied to the intellect, as intellectualists such as Aquinas supposed? Or could the will operate to varying degrees independently of the intellect as voluntarists such as Scotus and Ockham supposed? But these differences involved no disagreement about the fundamental model of agency, which remained in all cases practical reason based.

3. Freedom as a Metaphysical Power

Medieval conceptions of freedom combined two elements in tension. The first element was the idea of a multi-way power, that is, a power that by its nature left it up to us which actions we performed, one and the same power being exercisable to (p. 573) determine one action or another. Peter Lombard referred to free will [liberi arbitrii], which the philosophers have defined as the ‘free judgement of the will’ [liberum de voluntate iudicium], because the very power and ability of the will and reason, which we said above was free will, is free regarding whichever alternative it pleases because it can be moved freely to this or to that.4

This multi-way nature was often characterised as involving liberty of exercise—a freedom to do A or to refrain, and in many but not all cases, a further liberty of specification—a freedom to do A or to do some specific alternative B.

But this power was also conceived in aretaic terms as being diminished by the ethical corruption of the agent, and as enhanced by his ethical perfection. Freedom was enjoyed in its perfect or most complete form by God, the infinitely good being, union with whom, though only as a divine gift and only through the help of divine grace, was the ultimate goal of the ethical life and so, too, the ultimate goal of human action. So the theory of freedom was part of a general account of creation, fall and redemption—a process of corruption and recovery both ethical and metaphysical in which freedom took correspondingly reduced and recovered forms.

Adam’s fall was a fall from an original state of created innocence into a state of ethical degradation—a state that was described as one of servitude to sin. And Christ’s redemption was described as a release from the same servitude, bringing ultimate ethical perfection in heaven, a state that was described as one of perfect liberty or freedom, a supernatural condition transcending our original created condition and approaching, as far as the retention of our created human nature could permit, the condition of God. These descriptions of various kinds of ethical condition obviously used a vocabulary of freedom and unfreedom. But more than that, the conditions were characterised by specific reference to variations in freedom considered as a metaphysical power.5 As we were originally created, the metaphysical power of freedom was a libertas minor—a power both to do good and also to do bad. Whereas the servitude of fallen humanity to sin involved a reduction in the power to do good—a weakened power to do good that became dependent on divine assistance through grace—the final liberation of heaven would involve our enjoyment of the libertas maior enjoyed eternally by God, then perfection of our power of freedom through the complete removal of any power to do bad or to sin.

Central to this moralisation of the power of freedom was the relation of freedom to reason. With some exceptions, most theorists held that any enjoyment of the power of freedom presupposed the possession of intellect or reason. On most views, human freedom was not to any substantial degree also shared by the lower animals. And in its perfect form, the exercise of freedom would always and infallibly follow the direction of reason. Indeed, it seemed that the libertas maior of heaven might, in effect, coincide with a perfected human capacity for reason.

Hence, in a writer such as Anselm we see a tendency to characterise freedom proper as really a capacity to do right. Thus, Anselm defined freedom of choice or will as ‘the power to preserve rectitude of will for its own sake’.6 The moralisation of freedom thus involved a certain tendency to rationalism about freedom—an assimilation of freedom to rationality.

(p. 574) But the assimilation could not be complete, as it was in considerable tension with the conception of freedom as a multi-way power. For the capacity for reason is not obviously by nature a multi-way power. In those cases where reason requires that we do one specific thing, it is not at all clear that the capacity for reason need involve any capacity to do otherwise at all, doing otherwise being something that reason actually excludes. Indeed, there was an obvious paradox arising from freedom’s being by nature both a multi-way power and a mode of excellence coincident, in perfected form, with reason. For the perfection of the power, as a libertas maior, appeared to involve the removal of one crucial kind of alternative—the removal of the power to sin. The perfection of freedom seemed then to involve, in one respect, its diminution as a power over alternatives. But this diminution was rarely admitted to be truly a diminution in freedom itself. What appeared to be a reduction in freedom was generally explained and celebrated as, correctly understood, its increase. As Lombard observed, relying on the authority of Augustine:

Indeed a choice [arbitrium] that is quite unable to sin will be the freer … after the confirmation of beatitude there is to be a free will in man by which he will not be able to sin; and this free will is now in the Angels and in the Saints, who are with the Lord; and certainly it is the more free, as it is the more immune from sin and the more prone to good. For one is the further from that servitude of sin, of which it is written: He who works sin is the slave of sin, as one’s judgment is freer in choosing the good.7

So freedom was a power that by its very nature made alternatives available. But it was also a power that was enhanced by ethical perfection, so that in its most complete form it had to exclude unethical and unreasonable alternatives, thus ruling out any possibility of its being exercised defectively.

This tension raised theological problems. It was difficult for freedom to remain a mode of excellence while still playing its normal role in ethical theory—a role that very much depended on its nature as a power to do more than one thing. Take, for example, the theory of the incarnation and redemption. Christ, as the perfect man, in his human life already fully enjoyed freedom in perfected form. But precisely because freedom in perfected form precluded any power to sin, in his case the power of freedom could no longer easily do the work its imperfect form could do for ordinary humans—which is serve as a basis of merit for doing right. Merit for doing right in this life depended on one having freedom in the form of power or control over whether or not one did right, and so merit for doing right depended on the power to do wrong. Now Christ’s human obedience to the command of the Father to sacrifice himself on the cross was redemptive only on the assumption that it was meritorious. But as a perfect human, Christ lacked, or so it seemed, the power to disobey that command.8

God not only served as the pattern of freedom, exemplifying freedom in perfect form. He was also its source in us as our creator. That could affect the theory of freedom as a metaphysical power in another way. For medieval philosophers had to devise a theory not only of how the power might be perfected and increased, but also of how far the power’s possession and exercise by the agent was consistent with the (p. 575) impact on the agent of power exercised from without, by other beings. Now the determination of an agent’s choice by some created object, such as an angel or a planet, were it possible, was widely viewed as something that would impose necessity and remove freedom. But since God was creator and the source of freedom, determination by God was not universally viewed as having the same implication. Even in this life, God’s determining, through grace, the agent to do good was held by Aquinas to be consistent with the agent’s retention of the power to do otherwise. On Aquinas’s view, such divine determination in the direction of the good by the very creator of our freedom preserved freedom and contingency, and it did not impose necessity.9

Not everyone accepted Aquinas’s doctrine. For example, in early-modern Jesuit writers such as Molina we find a clear sense that determination of the agent’s action by any external cause, even by God, is a threat to freedom—and so a kind of determination that any acceptable theory of the operation of divine grace must clearly reject.10 But this Molinist view was not yet uncontroversial. So the free will debate in its modern form did not yet clearly exist. Nowadays the debate between incompatibilists and compatibilists is assumed to be about the implications of causal determination from without just considered in itself, irrespective of the metaphysical and ethical dignity of the external cause. Thus, determination of the agent by causation from without is taken by modern incompatibilists to be an automatic threat to freedom in every case, whether the external cause be creator or created. But this assumption has only become general in modern times.

If freedom was based on reason, what explained its apparent limitation to the practical sphere? For reason extends to the theoretical, and indeed, it might appear that it is the cognitive or immediately theoretical faculty of the intellect that underpins our capacity for reason generally. So what explains why freedom is a freedom primarily of the will rather than of the intellect, if that, indeed, is how freedom is to be understood?

In Aquinas, freedom was explained in terms of an aspect of our theoretical grasp of the good. The good is the object of reason in its practical form. But the good is also the subject of cognitive or intellectual enquiry: the good falls within the truth at which the intellect is directed. And it was as such a subject that Aquinas used the good to explain the basis of our freedom. For the good was seen by him as attaching in this world to a variety of options—options that realised it in incomplete or finite form. So, in an enquiry about what it is best to do, that left the practical intellect free to judge any one of a variety of options as the best to pursue. And this freedom at the point of the intellect explained why the will was free to choose one option rather than another. And, indeed, the freedom of the will had to be related to a corresponding freedom of the intellect. For Aquinas was an intellectualist about the operation of the will. Electio at the point of the will was a function of the agent’s intellectual judgement about which option was best.

So while freedom was a characteristic of the will as its subject, Aquinas taught that its cause was to be found in reason, freedom of choice or liberum arbitrium being described by him as the free judgement of reason.11 And the will’s status as an intellectual or rational faculty was derivative, owing to its acts being informed by those of the intellect. The liberating state of beatitudo or happiness, which was the (p. 576) destiny of the saved in heaven, was primarily a state of the intellect as involved in the beatific vision or contemplation of God.

For Scotus, however, the will’s freedom depended on a duality in the good that was its object—as consisting both in the good that lay in the happiness of the agent and in justice, where the good was pursued for its own sake. But the nature of freedom was not exhausted by the nature of the will’s intellectually presented object. Freedom was understood by Scotus to consist in a power of the will to determine itself, and the presence of this power was only revealed to us in our experience of our own choices.12 Moreover, the will determined its actions as an efficient cause of them, judgements of the intellect serving only as secondary, partial or non-determining efficient causes. The will could operate to a degree independently of judgements made by the intellect. Indeed, appealing to the characterisation of rational power by Aristotle in Metaphysics Book IX as involving a potential to opposites, in Scotus’s view, thanks to the possibility of freedom, it was the will rather than the intellect that was the truly rational power. For Scotus, reason did, after all, involve by its very nature a capacity for alternatives, and it was not the intellect but the will itself that, thanks to its freedom, could constitute the primarily and properly rational faculty.13 Again, it was the will rather than the intellect that was the primary source of the liberating happiness of heaven, as the faculty by which we direct ourselves to God in love of him.

Besides freedom as a power over alternatives, there is also voluntariness in the sense in which the term would exclusively be used by Hobbes. This is the property of occurring as and through being willed. Voluntariness and freedom have often been identified, but they involve, on the face of it, quite distinct powers. Freedom is a multi-way power over alternatives, whereas voluntariness is not similarly multi-way. Certainly one may both be able to do A on the basis of willing to do A and able to refrain on the basis of willing to refrain. But these are distinct one-way powers, not a single two-way power. One power may be possessed without the other: someone may be able to stay in a room on the basis of willing so to do but, because unbeknown to them the room is locked, be unable to leave on the basis of willing so to do.14 And certainly only one of these distinct powers can ever be exercised at a given time.

And though an agent may have the capacity to act as he wills, he may lack freedom of will; he may not be able to determine for himself what he wills. In which case, if freedom of action depends on a freedom of will, there may for this reason be voluntariness without any accompanying freedom. On the other hand, there seems nothing in the idea of freedom that immediately implies voluntariness. For it may be up to me what I decide or will, so that I have freedom of will. But it may not be possible for me to take specific decisions on the basis of decisions so to decide; I may not be able to will as I will. Even if there is freedom of will, there may not be such a thing as a voluntariness specifically of the will.

It was widely held by many in the medieval tradition that voluntariness did not suffice for freedom. But was it necessary for freedom? This question about voluntariness arose from within the theory of action. Voluntariness was a feature of imperated actions, which by their very nature occurred on the basis of a prior willing of them. But was it in any form a feature of elicted acts of the will themselves, and if so, a (p. 577) feature of action and so of free action generally? One view that had supporters from Augustine and Anselm through Scotus and down to Suarez was that elicited actions of the will, and so also free actions of the will, must by their very nature be willed. But this characteristic of being willed did not involve any relation to a prior act of will distinct from and motivating of it—which would involve a vicious regress—but was a reflexive feature of the act of will itself and something internal to it: as Suarez put it, citing Anselm: ‘omnis volens ipse suum velle necessario vult’—anyone who wills necessarily wills his own willing.15

Other thinkers objected that the appeal to voluntariness did no work in the theory of how elicited acts constituted purposive doings—that work was done by the practical reason–based model—nor, given the differences mentioned above between freedom and voluntariness, in any account of their freedom. In any case, it was obvious that elicted acts of the will were not reflexively and internally self-willed. Acts of will no more willed themselves than acts of seeing saw themselves.16

A further question concerned the nature of freedom itself. Freedom seemed to involve a kind of power to determine. But what kind of power, and how was it related to other powers in nature? It was natural to think that its exercise involved efficient causation, certainly at the point of imperated actions, which were widely seen as efficiently caused by the elicted acts that motivated them. But it was also natural to see efficient causation as involved in the free performance of the elicited acts themselves. The free agent could be seen as an efficient cause of his own decisions and choices. The Scotist tradition put emphasis on this conception of freedom as an efficient causal power of agents, and it was under the influence of that tradition that Suarez gave this view its most sophisticated expression in the Metaphysical Disputations. Suarez distinguished between necessary and free or contingent causes. Necessary causes were non-rational substances, such as stones, which under any given condition could produce but one effect, as when a stone once thrown at a window with a given force would only break it. Free causes were rational substances, such as humans or angels, and these operated contingently, with efficient causal power taking multi-way form. Under a given set of conditions, a free cause was sufficient for a variety of possible effects, and which effect it produced was up to the cause. Our knowledge of our possession of this distinctive causal power was based on our experience of our own agency, which revealed its existence to us.17

4. The Ethical Significance of Freedom

In medieval ethical theory, freedom was a central ethical notion along with law, and the two phenomena were treated as existing in a complex harmony. By contrast the early-modern period saw the development of a very different kind of ethical theory. (p. 578) In attempting to rid ethical theory of any dependence on freedom as a metaphysical power, Hobbes was to argue that freedom and law are quite opposed notions, freedom properly understood requiring the absence of law and law serving simply to limit freedom.

In scholastic moral theory, law provided humans with direction towards the good that carried the force of obligation—a direction that freedom in the form that we enjoy it in this life, as a power to do evil as well as good, especially required. Moreover, insofar as law imposed obligations, it presupposed the very power of freedom that it directed. This presupposition operated through what was often seen as a constitutive connection between obligation and desert of punishment for its breach. For desert or meritum required the agent to have had control over whether or not he breached the obligation.

The same dependence of law and obligation on freedom also came from the relation of obligation to blame. Obligation was a standard to breach which was to be blameworthy for doing wrong. But blame was often held to be a criticism that presupposed a power of freedom on the agent’s part. Consider Aquinas’s characterisation of blame:

Hence a human action is worthy of praise or blame in so far as it is good or bad. For praise and blame is nothing other than for the goodness or badness of his action to be imputed to someone. Now an action is imputed to an agent when it is within his power, so that he has dominion [dominium] over the act. But this is the case with all actions involving the will: for it is through the will that man has dominion over his action … Hence it follows that good or bad in actions of the will alone justifies praise and blame; for in such actions badness, fault and blame come to one and the same.18

Blame was understood by Aquinas as a criticism in which the badness of an agent’s action was imputed to the agent, so that the agent counted as bad to have done what he did. And what warranted this imputation was the agent’s control or dominium over his action—his freedom to determine for himself which action he performed.

The primary form of law as it governed humans was the natural law. And natural law was reason itself in preceptive form—a form in which it did not merely recommend through consilia but imposed obligations through praecepta. In other words, the natural law was a directive form of reason—reason that with the binding force of obligation served to direct the proper use of freedom. Now any directive force of reason must directly address and apply to the will. For it is at the point of the will—the point at which we choose or decide to perform this action rather than that—that we immediately respond to directives of practical reason. So it was the will that the binding directives of reason in its obligatory or preceptive form had immediately to address. That meant that the natural law had immediately to bind the will. Obligations of the natural law were immediately obligations on the will. As Aquinas made clear, under natural law we were bound to will obligatory external actions, and bound not to will prohibited ones, so that the existence of an obligation to give alms implied a corresponding obligation to decide and intend to give alms.19 Behind this lay a conception of obligation, not simply as a kind of command—though obligations (p. 579) might be imposed through commands—but as a demanding mode of justificatory support or, to use a term from Suarez, a demandingly preceptive vis directiva or justificatory force.20

If natural law involved a directive force of practical reason that, like any such, had directly to address the will, and this particular justificatory force, as obligatory and preceptive, was by its nature freedom-directive, then that had an obvious implication. The will would have to be a locus of free action; freedom would have to be exercised at the point of the will. So the location of the metaphysical power of freedom in the will followed both from the dominant practical reason–based model of action, but it was also something required by the dominant model of natural law and obligation as involving a distinctively obligatory or preceptive justificatory force.

Not only did the general conception of natural law as a demanding justificatory force of reason presuppose a freedom and agency of the will; such a freedom of the will was also presupposed by the view taken of what the natural law specifically required. It was the content as well as the general nature of the natural law that dictated the location of freedom. For the natural law primarily required love of God and of neighbour as bearing the image of God. And love was, as we have seen, conceived as involving a free action of the will that had God or our neighbour as its object.

Law not only presupposed freedom as a metaphysical power. It also constituted freedom as a right and in its highest form, as the law of grace of the new covenant, directed us towards the perfect freedom of heaven—towards freedom as a state of liberation.

The term dominium could pick out both the metaphysical power of freedom—as we saw in Aquinas’s account of the conditions for fair blame—but was also used to refer to a right of control or possession, which involved a right to exercise the metaphysical power. A phrase used by an author such as Suarez was dominium libertatis, or dominion over one’s freedom. Nature not only gives us a power of freedom; through the law of nature it also gave us the right to exercise it:

If, however, we are speaking of the natural law of dominion, it is then true that liberty is a matter of natural law, in a positive, not merely a negative sense, since nature itself confers upon man the true dominion of his liberty … For liberty rather than slavery is a precept of the natural law, for this reason, namely, that nature has made men free in a positive sense (so to speak) with an intrinsic right to liberty, whereas it has not made them slaves in this positive sense, strictly speaking.21

5. Towards a New Metaphysics of Freedom

Modern theories of freedom are in various ways the legacy of Hobbes’s attack on the very reality of freedom as a metaphysical power of self-determination. This attack was deeply influential. If freedom remained as any kind of power thereafter, (p. 580) it often did so reduced to a complex case of some other power of a kind that Hobbes did admit into his metaphysics, such as voluntariness. Hobbes’s attack on freedom as a metaphysical power was part of a wider attack on the scholastic understanding of freedom’s ethical significance. And this had an equal effect on subsequent philosophy. It led ethical theory generally to begin to detach itself from commitment to freedom as a metaphysical power, and it also led, especially within the Anglo-Saxon tradition, to conceptions of that metaphysical power that were increasingly detached from ethical theory. In particular, there was an abandonment of the aretaic model that tied the perfection of the metaphysical power to ethical excellence.

Central to Hobbes’s scepticism is his reduction of action in general to what had previously been the special and secondary case of willed or imperated action—what Hobbes was now to term ‘voluntary action’. Hobbes completely abandoned any theory of purposiveness as consisting in a distinctively practical mode of exercising reason located in the will as a special intellectual or rational appetite. Indeed, he found the idea of intellectual or rational appetites distinct from the passions unintelligible:

For I do not fear it will be thought too hot for my fingers, to shew the vanity of words such as these, Intellectual appetite, conformity of the appetite to the object, rational will, elective power of the rational will; nor understand I how reason can be the root of true liberty, if the Bishop (as he saith in the beginning) had the liberty to write this discourse. I understand how objects, and the conveniences and inconveniences of them, may be represented to a man by the help of his senses; but how reason representeth anything to the will, I [do not] understand …22

Instead, Hobbes introduced a new model of purposiveness that could only apply to the case of what had previously been called imperated action. To act purposively was always to do something voluntarily, on the basis of a prior will to do it—a will that, in line with much medieval action theory, Hobbes conceived to be an efficient cause of the action it motivated. But now this motivating will was no longer itself a prior case of action, but passive. This was because, as Hobbes insisted, the will itself is not voluntary as are the actions willed that it motivates:

I acknowledge this liberty, that I can do if I will, but to say, I can will if I will, I take to be an absurd speech.23

Given Hobbes’s view that the will is not itself voluntary, his new model of action as nothing more than voluntariness just could not apply to willings or motivations.

Hobbes also denied the status of actions to willings because of his particular conception of voluntariness. For Hobbes voluntariness was never an internal or reflexive phenomenon. Rather, it always involved efficient causation of the action willed by a prior willing distinct from it. That meant that if actions were by nature voluntary, the willings of them could not be actions too, on pain of a vicious causal regress.

The object or content of the motivating will to act still provided the goal at which the action willed was directed. But since the motivating will was not itself a purposive action, but merely a passion or passive desire, purposiveness—being a (p. 581) response made to a goal in order to attain it—was restricted to the action willed. It followed that all explanation in terms of purposes now involved reference to a motivating efficient cause. So Hobbes could afford to claim that all explanation of action in terms of the purposes for which it was performed had to be in terms of efficient causes. In fact, Hobbes claimed to find the idea of motivation by causes other than efficient ones, such as by formal causes, completely unintelligible:

Moved not by an efficient, is non-sense.24

That meant that it was the very nature of action to involve the exercise of efficient causal power—a kind of power that, as we have seen, had already been appealed to in order to characterise the power of freedom. On Hobbes’s account of action, though, the immediate bearer of this power was no longer the agent, but rather a passive motion of the will within the agent. And that motivation as passive was not the product of any free causation on the agent’s part, but rather, the determined effect of a chain of necessary causes for which the agent had no responsibility.

Hobbes did not even begin to claim that the operation of this passive willing involved the agent’s determining for himself how he acted. Indeed, Hobbes regarded the very idea of self-determining freedom as unintelligible because it was inherently and viciously regressive:

And if a man determine himself, the question will still remain what determined him to determine himself in that manner.25

He also regarded as unintelligible the Suarezian attempt to model freedom as a special kind of contingent or multi-way efficient causal power. That was because any power that was genuinely sufficient to produce an effect must, in his view, when present actually produce it, so that a cause sufficient for a range of contrary effects—a cause involving indetermination, as Hobbes put it—must simultaneously produce them all, which was impossible:

But that the indetermination can make it happen or not happen is absurd; for indetermination maketh it equally to happen or not to happen; and therefore both; which is a contradiction. Therefore indetermination doth nothing, and whatsoever causes do, is necessary.26

So all efficient causation was in terms of necessary causes; every cause could produce but one effect, and there was no contingency about how it would operate.

Freedom was no longer a kind of contingent power since all power operated through necessity. That meant that far from being itself a kind of power, freedom could only amount to an absence of obstacles to other kinds of power:

Liberty is the absence of all impediments to action, that are not contained in the nature, and in the intrinsecal quality of the agent.27

As such, freedom was not restricted to rational creation, nor was its perfection tied to ethical excellence:

how reason representeth anything to the will, I understand no more than the Bishop understands there may be liberty in children, in beasts, and inanimate (p. 582) creatures. For he seemeth to wonder how children may be left at liberty; how beasts imprisoned may be set at liberty; and how a river may have a free course.28

Moreover, in involving obligations law was not harmonious with freedom, but its essential opposite. Law in imposing obligations imposed a kind of block to the power of appetites. And as with Hobbesian liberty generally, there was nothing more to liberty as a right than the absence of such a block—in this case through the absence of some restricting law. Just as there was no metaphysical power of freedom, there was no room for a normativity of law of the kind that natural law was understood to constitute by the scholastic tradition: there was no room for law as a special directive force of reason that gave normative recognition to the power of freedom both by imposing obligations to direct its exercise (lex or ius) and by affording rights to protect its exercise (ius):

For though they that speak of this subject, use to confound Ius, and Lex, Right and Law; yet they ought to be distinguished; because Right, consisteth in liberty to do, or to forbeare; Whereas Law determineth, and bindeth to one of them; so that Law, and Right, differ as much, as Obligation and Liberty; which in one and the same matter are inconsistent.29

But although Hobbes himself denied that freedom could be a metaphysical power, this denial was not maintained so obstinately by his successors. For the power that was used by Hobbes to characterise purposiveness—the efficient causal power of motivations—was used by Locke and many successors also to give a positive account of freedom as a power. The Hobbesian theory of an action’s relation to the object at which it was directed was turned into a theory of the agent’s relation to the action he determines. The agent was identified with the motivations of his will—even though these were now passive motivations that preceded his agency and were no longer his deliberate doing. The actions efficiently caused by those passive motivations were, nevertheless, to count as determined by the agent.

In effect, a reductive account was given of freedom as a complex, two- or multi-way case of voluntariness. The theory of freedom familiar from classical English-language compatibilism was born. The power of freedom to do A or not is now a complex, composed of the power to do A as an effect of wanting or willing to do A, and of the distinct power to refrain from doing A as an effect of wanting or willing to refrain.

Even those theorists of freedom who opposed this compatibilist theory, perhaps in the cause of some form of libertarianism, often did so to some degree or other within the metaphysical framework that Hobbes imposed.

Thus, conceptions of the metaphysical power of freedom were increasingly detached from any theory of the agent’s metaphysical and ethical perfection. Conceptions of freedom were increasingly neutral on what would constitute the ethical improvement of the agent. And the operation on the agent of the infinitely good creator whose image they bore no longer was treated as having distinctive implications for human freedom. Determination of one’s action by God was as much (or as little) a threat to freedom as its determination by any other created being.

(p. 583) 6. The Transformation of Ethics

Just as ethical theory ceased to shape and condition accounts of freedom as a metaphysical power, so at the same time that metaphysical power ceased to play a role in ethical theory. Either the ethical life ceased to be directed at the liberation of the agent, or the condition of liberation increasingly ceased to be understood by reference to freedom as a metaphysical power. And notions of responsibility, of obligation and of liberty as a right were similarly detached from the theory of freedom as a power.

We may see two tendencies within ethical theory following Hobbes. One is a naturalism very much in the spirit of Hobbes himself, and the other a rationalism in the spirit of Kant. Both traditions, naturalist and rationalist alike, eventually detached ethical theory from any commitment to freedom as a multi-way metaphysical power, and this is true even of the post-Kantian rationalist tradition despite Kant’s own commitment to the noumenal existence of freedom as a power. Unlike the naturalist tradition, though, the rationalist tradition continued to conceive of the ethical life as involving and aiming at a condition of liberation, this time involving not a perfection of metaphysical freedom, but rather, the perfection of reason in the form of autonomy.

The post-Hobbesian reduction of freedom to a complex case of voluntariness had one especially important consequence for ethical theory. Freedom as a multi-way power ceased to be a condition of moral responsibility. And that is because on the reductive account of freedom as a complex case of voluntariness, a power to act otherwise is no longer being exercised to determine action. Freedom is no longer a single power that if possessed at all must be a power to do more than one thing. It is now a combination of distinct powers each to do but one thing—a power to do A on the basis of a will or desire to do A and a distinct power or will to refrain from doing A on the basis of a will or desire to refrain—each power being exercised without the other, and each being capable of being possessed without the other. So now when an agent does A, no multi-way power of freedom—no power to act otherwise—is ever actually exercised by him. The only power to act being exercised is the one-way power to do A as an effect of willing to do A. The power to refrain is the quite distinct power to refrain as an effect of willing to refrain, and this power is obviously not being exercised at all. But how can the responsibility of the agent depend on a power to do otherwise that is never actually exercised? It is voluntariness—doing what one does because one wills or wants to do it—on which moral responsibility now comes to depend; the presence of a freedom to do otherwise becomes irrelevant.30

The conception of self-determination as involving multi-way freedom came under pressure within the medieval tradition primarily from rationalism about freedom linked to an aretaic conception of freedom as something perfected by growth in virtue. The perfection of freedom involved an assimilation of freedom to reason and so the exclusion of some power over alternatives. By contrast we can (p. 584) now see that the conception of self-determination as inherently multi-way was coming under pressure within post-Hobbesian philosophy from quite another direction. This was a divorce of self-determination from any aretaic conception of it and, instead, a naturalistic reduction of freedom to a complex case of voluntariness—an inherently one-way efficient causal power also to be found in non-rational nature and in particular in the actions of lower animals.

But it is also true that the determination of willed action by passive motivations is not a particularly plausible model for understanding self-determination. For why should we identify the agent with motivations within him that are not within his control and not of his own doing?

The long-term effect would be to detach the ethical significance of action conceived as voluntariness from any connection with genuine self-determination. A modern example of this is the Hartian project of understanding responsibility for the voluntary not in terms of any theory of desert based on self-determination, but rather, in terms of a fair choosing system. On this model, which Hart intended for legal responsibility within a legal system, responsibility for doing wrong is understood by reference to the liability it brings to pressure and sanction.31 But the fairness of a sanction is not explained in terms of its being the deserved response to the misexercise of some metaphysical power of self-determination. Instead, the fairness of the sanctions is understood both in terms of their general deterrent and reformative utility, but also in terms of the fairness of their application. And they will be fairly applied if those subject to them are given a fair chance of avoiding them. Hence, their application is restricted to the voluntary, since voluntary action—action subject to the will—is action that we can perform or refrain from at will, as a means to avoiding sanctions.

Obligation, the standard we are responsible for meeting, is no longer being conceived on this Hartian approach as a special force of reason directive of a metaphysical power of freedom, and it is instead understood simply as a sanction-backed standard on the voluntary. And with this reconstruction of the idea of obligation comes a thinning out of the very theory of reason or rationality. Within medieval ethics the directives of reason in practical form were tied to a rich metaphysics of the self, of which the power of freedom was a central feature. Practical reason could take the distinctive form of obligatory law precisely because it directed not just a bare capacity for reason, but a genuine capacity for self-determination. But once that capacity was lost from ethical theory, or disregarded by it, the theory of practical reason would have to be correspondingly simplified.

And so we arrive at modern theories of ethical rationality that see ethical standards as no more than general standards of reason, with the blame that asserts them as no more than a general form of rational criticism. On such views blame is no longer what Aquinas conceived it to be—a condemnatory criticism that presupposed a personal metaphysical dominium over one’s actions and that was specifically directed at free action. It becomes, instead, on some views a mere assertion of unreasonableness that can apply equally to voluntary actions and to passive desires and emotions.32

(p. 585) How then are rights to liberty and ideas of liberation to be understood, if not by appeal to some theory of freedom as a metaphysical power? A crucial concept here is that of autonomy. The idea of autonomy in modern philosophy derives from Kant’s conception of rational self-legislation, but in a form detached from his metaphysics of noumenal freedom—which was already a withdrawal from a serious theoretical metaphysics of freedom, the theoretically knowable phenomenal world having already been surrendered by Kant to universal causal determinism.

The post-Kantian rationalist tradition has much in common with the rationalising trend in medieval theories of freedom, insofar as freedom is assimilated to a certain excellence in one’s capacity for rationality. But the idea of a power to determine action for oneself has been abandoned. Instead of the idea of a power of self-determination that operates independently of other powers, we have the idea of a capacity for rationality that operates independently of external standards. This can take a number of forms. There is the Kantian idea of rational or ethical standards as the legislated or constructed product of human reason.33 Or there is the Millian idea of a capacity to live according to values of one’s own, developed apart from the standards of one’s culture.34

Autonomy so understood can involve a kind of liberation, insofar as one’s reason develops to a level that liberates one from dependence on standards or values external to one. And liberty as a right can be explained in terms of a right to freedom from kinds of pressure or coercion that would obstruct the attainment of autonomy.

What is unclear is how desirable the enjoyment of such a form of rationality might be. Certainly there is nothing in rationality that generally demands or requires the preservation of alternative options. Indeed, in theoretical reason, removal of all options but one—through evidence or proof—when available, is highly desirable. If within the practical sphere there is an intuitive right not to be coerced, that may turn out to be precisely because we have, in metaphysical freedom, a genuine multi-way capacity to determine for ourselves how we act—a capacity that would be obstructed by coercive pressure in one particular direction, no matter how favoured by reason that one direction might be.


Anselm. 1968. S. Anselmi Cantuariensis Archiepiscopi Opera Omnia, ed. F. S. Schmitt. Stuttgart-Bad Cannstatt, Germany: Friedrich Fromann Verlag.Find this resource:

Garrigou-Lagrange, Reginald. 1936. ‘Prémotion physique’. In Dictionnaire de Theologie Catholique, volume 13, 31–77. Paris: Letouzey et Ané.Find this resource:

(p. 587) Hart, Herbert. 1968. Punishment and Responsibility. Oxford: Clarendon Press.Find this resource:

Hobbes, Thomas. 1656. The Questions Concerning Liberty, Necessity and Chance, Clearly Stated between Dr Bramhall Bishop of Derry, and Thomas Hobbes of Malmesbury, London.Find this resource:

Hobbes, Thomas. 1996. Leviathan, ed. Richard Tuck, Cambridge: Cambridge University Press.Find this resource:

John Duns Scotus. 1997. Quaestiones super libros metaphysicorum Aristotelis libri VI–IX. New York: St Bonaventure University.Find this resource:

Kant, Immanuel. 1996. ‘Groundwork of the metaphysics of morals’. In Practical Philosophy, ed. Mary Gregor. Cambridge: Cambridge University Press.Find this resource:

Locke, John. 1973. An Essay Concerning Human Understanding, ed. P. H. Nidditch. Oxford: Clarendon Press.Find this resource:

Mill, John Stuart. 1989. On Liberty, ed. Stefan Collini. Cambridge: Cambridge University Press.Find this resource:

Molina, Luis de. 1953. Liberi arbitrii cum gratiae donis, divina praescientia, providentia, praedestinatione et reprobatione concordia, ed. J. Rabeneck. Madrid: Sapientia.Find this resource:

Peter the Lombard. 1981. Sententiae in IV Libris Distinctae. Grottaferrata, Italy: Collegii S. Bonaventurae ad Claras Aquas.Find this resource:

Pink, Thomas. 2004. ‘Suarez, Hobbes and the Scholastic Tradition in Action Theory’. In The Will and Human Action: From Antiquity to the Present Day, ed. Martin Stone and Thomas Pink, 127–53. London: Routledge.Find this resource:

Pink, Thomas. 2005. ‘Action, will and law in late scholasticism’ In Moral Philosophy on the Threshold of Modernity, ed. Jill Kraye and Risto Saarinen, 31–50. Dordrecht, the Netherlands: Springer (The New Synthese Historical Library 57).Find this resource:

Pink, Thomas. 2009a. ‘Natural law and the theory of obligation’. In Psychology and Philosophy, ed. Sara Heinämaa and Martina Reuter, 97–114. Dordrecht, the Netherlands: Springer (Studies in the History of Philosophy of Mind 8).Find this resource:

Pink, Thomas. 2009b. Power and Moral Responsibility. Philosophical Explorations 12:127–51.Find this resource:

Scanlon, Thomas. 1998. What We Owe to Each Other. Cambridge, MA: Harvard University Press.Find this resource:

Suarez, Francisco. 1856–70. Opera Omnia, ed. C. Berton. Paris: Vives.Find this resource:

Suarez, Francisco. 1994. On Efficient Causality: Metaphysical Disputations 17, 18 and 19. New Haven, CT: Yale University Press.Find this resource:

Thomas Aquinas. 1961. Summa contra gentiles. Turin: Marietti.Find this resource:

Thomas Aquinas. 1950. Summa theologiae. Turin: Marietti.Find this resource:

Vasquez, Gabriel. 1611. In primam secundae Sancti Thomae I. Ingolstadt, Germany.Find this resource:


(1) . For discussion of the practical reason-based model, see Pink (2004).

(2) . For discussion of this possibility, see Suarez (1856–70, volume 4, 308).

(3) . See Aquinas Summa Theologiae (1a2ae, q.17, a. 4).

(4) . Sententiae II, d. 25, cap. 1§2; Peter the Lombard (1981, 461).

(5) . For a classic discussion, see Peter the Lombard (Sententiae II, d. 25; 1981, 461–69).

(6) . Anselm (1968, volume 1, 212).

(7) . Peter the Lombard (Sententiae II, d. 25, cap. 4; 1981, 463).

(8) . For the difficulty, and a lucid and comprehensive discussion of attempted solutions within the scholastic tradition, see Suarez (1856–70, volume 18, 282–92).

(9) . See Aquinas (Summa contra Gentiles III, 94; Aquinas 1961, 138–41); Aquinas (Summa theologiae I, q.19 a. 8, and q. 22–23, especially q. 23 a. 6) and Thomas Aquinas (1950, 114–33). For a classic discussion from a neo-Thomist Dominican perspective, see Garrigou-Lagrange (1936).

(10) . See, for example, Molina (Liberi arbitrii … concordia, Part 7, q. 23, a. 4–5, disputatio 1, membrum 7; 1953, 501–10).

(11) . Summa Theologiae 1a2ae, q. 17, a. 1, ad 2; Aquinas (1950, 81).

(12) . Quaestiones super libros Metaphysicorum (q.15); John Duns Scotus (1997, 682–83).

(13) . Quaestiones super libros Metaphysicorum (q.15); John Duns Scotus (1997, 675–99).

(14) . See John Locke, An Essay Concerning Human Understanding (II, 21 §10), and Locke (1973, 238).

(15) . For defence of this view and discussion of prior support for it, see Suarez (1856–70, volume 4, 196).

(16) . ‘And in like manner a willing is produced from the will, but is not itself willed through its own production; just as a seeing … is not itself seen through its own production.’ Vasquez (In primam secundae Sancti Thomae, disputation 23, chapter 2; 1611, 165).

(17) . See metaphysical disputation 19 ‘On necessary and free causes’, section 2, §13, in Suarez (1856–70, volume 25, 697), and in Suarez (1994, 291).

(18) . Summa Theologiae (1a2ae, q. 21, a. 2); Thomas Aquinas (1950, 112).

(19) . Summa Theologiae (1a2ae, q.100, a.9); Aquinas (1950, 463).

(20) . For discussion of this Force model of moral obligation, its history and its implications, see Pink (2005, 2009a).

(21) . De legibus (book 2, chapter 14, §16); Suarez (1856–70, volume 5, 141).

(22) . Hobbes (1656, 35–36).

(23) . Hobbes (1656, 29).

(24) . Hobbes (1656, 59).

(25) . Hobbes (1656, 26).

(26) . Hobbes (1656, 184).

(27) . Hobbes (1656, 285).

(28) . Hobbes (1656, 35–36).

(29) . Thomas Hobbes (1996, chapter 14), ‘Of the first and second Naturall Lawes, and of Contracts’.

(30) . For a discussion of the effect on modern ethical theory of this transformation in the understanding of moral responsibility and its basis in self-determination, see Pink (2009b).

(31) . Hart (1968).

(32) . See Scanlon (1998, 22).

(33) . See Kant (1996).

(34) . See Mill (1989).