Because my overall aim in this book is to discuss a representative sample of Locke’s philosophical thought, concentrating on certain issues that I take to be of perennial interest and importance, there are good reasons to include some discussion of his philosophy of action. One is that Locke’s account of agency has provoked strong reactions, many of them highly critical. Another is that, despite such criticism, Locke is widely acknowledged as having laid the foundations for all subsequent philosophical debate concerning the nature of voluntary action and the so-called problem of ‘free will’. A third is that – again, despite such criticism – some modern philosophers of action, amongst whom I would include myself, believe that Locke’s account of action and the will is fundamentally correct. Finally, a fourth reason is that a discussion of Locke’s philosophy of action serves as a natural bridge between his metaphysics and epistemology, on the one hand, and his political philosophy on the other – which I shall discuss in the next chapter. For in political philosophy we are above all concerned with human beings in their capacity as agents. Moreover, for a political philosopher like Locke, who is especially concerned with the liberty of the citizen and government by consent, human freedom of action must obviously be something of paramount importance. For if, by our very nature as human beings, we lacked such freedom, there could be no sense in trying to base the legitimacy of civil government on the freely given consent of the governed, nor in making the protection of the liberty of citizens a distinguishing feature of just political authority.
Towards the end of Chapter XXI of Book II of the Essay Concerning Human Understanding – entitled ‘Of Power’ – Locke remarks, with all the appearance of sincerity and genuine modesty, that
Impartial deductions of reason in controverted points being so very rare, and exact ones in abstract notions not so very easy, especially if of any length … I should think my self not a little beholding to any one, who would upon these or any other grounds fairly clear this subject of Liberty from any difficulties that may yet remain.
[Essay, II, XXI, 72]
This chapter was substantially revised by Locke for the second edition of the Essay and it seems clear from the foregoing remark, as well as from other expressions of hesitancy in the chapter, that Locke was never entirely happy with his treatment of the topic which is the chapter’s dominant concern – free agency and the so-called problem of ‘free will’. Locke himself notoriously regarded the expression ‘freedom of the will’ as a linguistic monstrosity because, as he puts it:
the Will is nothing but one Power or Ability, and Freedom another Power or Ability: So that to ask, whether the Will has Freedom, is to ask, whether one Power has another Power, one Ability another Ability; a Question at first sight too grosly absurd to make a Dispute, or need an Answer.
[Essay, II, XXI, 16]
But, fortunately for philosophical posterity, Locke did not leave the matter at that, recognizing that ‘Philosophy … when it appears in publick, must have so much Complacency, as to be cloathed in the ordinary Fashion and Language of the Country, so far as it can consist with Truth and Perspicuity’ (Essay, II, XXI, 20). Certainly, Locke recognized that there is a genuine philosophical issue that has, however misleadingly, been called ‘the problem of free will’, and he does his best to try to identify and solve it – though with how much success it is hard to judge, not only on account of the complexity and difficulty of the problem itself, but also because obscurities and apparent tensions in Locke’s text have provoked a number of different interpretations of his views on the subject.
Many present-day commentators interpret Locke as being a so-called compatibilist in the matter of free will, although some of these acknowledge – wisely, I think – that the claim that Locke is a compatibilist is at least contestable (see, for example, Yaffe 2000, p. 142, n. 5). Here I take a ‘compatibilist’ to be someone who holds that freedom of the will is compatible with universal causal determinism – that is, who holds that even if the operations of the will are all causally necessitated by prior events, the products of those operations, in the form of our bodily and mental actions, may nonetheless be said to be ‘free’ in the fullest sense of the term and hence properly subject to moral appraisal as commendable or blameworthy. Even if such an interpretation can be defended, however, it seems clear that there is more to be said about what precisely it is, in Locke’s view, that confers upon human agency a liberty ‘worth the Name’ (Essay, II, XXI, 50). In short, we still need to be given an account of what Gideon Yaffe has called ‘the Elusive Something’ (Yaffe 2000, p. 19): the crucial factor that distinguishes, for Locke, human action that is fully morally accountable from behaviour that is not.
Now, for all his heavy sarcasm about the phrase ‘freedom of the will’, Locke is in fact clear enough about what should be meant by freedom of action. According to Locke, one is free to act in a certain way – for instance, to raise one’s arm – just in case both (1) if one were to will to raise one’s arm, one’s arm would rise as a consequence and (2) if one were to will to forbear to raise one’s arm, one’s arm would fail to rise as a consequence (see Lowe 1986, p. 154). As he himself puts it:
[S]o far as a Man has a power to think, or not to think; to move, or not to move, according to the preference or direction of his mind, so far is a Man Free. Where-ever any performance or forbearance are not equally in a Man’s power; where-ever doing or not doing, will not equally follow upon the preference of his mind directing it, there he is not Free … that Agent is under Necessity.
[Essay, II, XXI, 8]
To illustrate his conception of what it is for an agent to be acting under necessity, Locke offers the famous example of a man asleep in a locked room, who wakes to find a friend there with whom he is happy to stay and talk. This man, Locke says, stays in the room not freely but under necessity, because if he were to will to leave it he would not succeed in doing so, the door being locked. The example is supposed to support a further contention of Locke’s, namely, that ‘Voluntary … is not opposed to Necessary’ (Essay, II, XXI, 11), on the grounds that, while the man stays under necessity, his stay is nonetheless voluntary – though whether it really does support this contention is not in fact as clear as it might seem to be (see Lowe 1986 and Lowe 1995, pp. 128–32). (Some readers may be tempted to see in this contention of Locke’s – that an action may be voluntary even though it is done ‘under necessity’ – an implicit endorsement of compatibilism, but I think that that would be mistaken: for at this point Locke is not discussing freedom of will but only freedom of action, and compatibilism is a doctrine concerning freedom of will.)
So far, we have only been considering what it is, according to Locke, for an agent to be free to act in a certain way – for instance, to stay in a certain room or to leave it. However, as Locke himself acknowledges, after some preliminary beating about the bush, a question which it seems natural to raise at this point is ‘Whether a Man be at liberty to will which of the two he pleases, Motion or Rest’ (Essay, II, XXI, 25). It is natural to ask this precisely because willing, as Locke himself seems to conceive of it, is itself a species of action, albeit a mental action. For Locke, volition or willing is ‘a thought or preference of the mind ordering, or as it were commanding the doing or not doing such or such a particular action’ (Essay, II, XXI, 5). But Locke dismisses the question peremptorily as one that is manifestly absurd, on the grounds that
to ask, whether a Man can will, what he wills; or be pleased with what he is pleased with [is a] Question, which, I think, needs no answer: and they, who can make a Question of it, must suppose one Will to determine the Acts of another, and another to determinate that; and so on in infinitum.
[Essay, II, XXI, 25]
But this is surely just bluster on Locke’s part. If willing is a kind of action – a question to which I shall return later – it should certainly make sense to ask whether one can will what one wills. And it should be possible, in principle, to give a positive answer to this question without committing oneself automatically to an infinite regress of willings. What is more to the point, though – and perhaps this is really all that Locke means to convey – if ‘freedom of will’ were merely a matter of our being free, at least sometimes, to ‘will what we will’, then this would not be a kind of freedom that we don’t already possess in good measure in respect of other actions that we perform, such as raising our arms. Hence, freedom of will, so conceived, would not be what Yaffe calls ‘the Elusive Something’. If there is a deeper or more significant species of ‘freedom’ to which we can and should aspire, beyond our Lockean freedom of action, we must seek it elsewhere than in a freedom to ‘will what we will’.
Here it is worth considering Gideon Yaffe’s own answer to the question of what it is that Locke identifies as ‘the Elusive Something’, for it is both interesting and novel. The key to this question, Yaffe thinks, lies in Locke’s remarks concerning the kind of freedom possessed by God and other ‘superiour Beings above us, who enjoy perfect Happiness’ (Essay, II, XXI, 49). For, Locke says, if we reflect upon their condition,
we shall have reason to judge that they are more steadily determined in their choice of Good than we; and yet we have no reason to think they are less happy, or less free, than we are. And if it were fit for such poor finite Creatures as we are, to pronounce what infinite Wisdom and Goodness could do, I think, we might say, That God himself cannot choose what is not good; the Freedom of the Almighty hinders not his being determined by what is best.
[Essay, II, XXI, 49]
So for Locke, according to Yaffe, we are free in the deepest and most significant sense just to the extent that our choices are ‘determined by what is best’, that is, by the good. And ‘determined’ here means necessitated. Of course, the kind of necessitation in question cannot straightforwardly be causal necessitation, but it may still rest upon causal necessitation: for if our choices are causally determined by our desires and our desires are for the good, then there is a clear sense in which our choices – and so our consequent actions – are determined by the good. And Locke does seem to hold that our choices – our volitions or exercises of the will – are causally determined by our desires or ‘uneasinesses’, for in answer to the ‘Question, what is it determines the Will?’ (Essay, II, XXI, 29) he replies that the ‘true and proper Answer is … always some uneasiness’ (Essay, II, XXI, 29). As for the implied identity of uneasiness with desire, Locke himself remarks a little while later that ‘This Uneasiness we may call … Desire; which is an uneasiness of the Mind for want of some absent good’ (Essay, II, XXI, 31).
The Locke that emerges from this analysis is a sophisticated compatibilist, who holds that our actions are all causally necessitated and yet may be ‘free’, in a deeper sense than that which attaches to them merely in virtue of their being dependent upon our will or choice, in so far as they are determined by the good, which they may be precisely to the extent that the desires that causally determine our choices are desires for the good. If our choices are so caused, we are no less free than God and the angels, whose choices are also determined by the good and who are, as a consequence, in a constant state of perfect happiness. It would be madness to hanker after some sort of rationally undetermined ‘freedom’ instead:
Is it worth the Name of Freedom to be at liberty to play the Fool, and draw Shame and Misery upon a Man’s self? If to break loose from the conduct of Reason, and to want that restraint of Examination and Judgment, which keeps us from chusing or doing the worse, be Liberty, true Liberty, mad Men and Fools are the only Freemen.
[Essay, II, XXI, 49]
But how are we supposed to get ourselves into a condition in which our choices are determined by the good, given that our psychology is not divine or angelic in nature? This, I think, is where both Locke and Yaffe’s interpretation of Locke meet a certain difficulty. Locke makes at one point the following crucial remarks:
There being in us a great many uneasinesses always solliciting, and ready to determine the will, it is natural … that the greatest, and most pressing should determine the will to the next action; and so it does for the most part, but not always. For the mind having in most cases, as is evident in Experience, a power to suspend the execution and satisfaction of any of its desires, and so all, one after another, is at liberty to consider the objects of them; examine them on all sides, and weigh them with others … This seems to me the source of all liberty; in this seems to consist that, which is (as I think improperly) call’d Free will. For during this suspension of any desire, before the will be determined to action … we have opportunity to examine, view, and judge, of the good or evil of what we are going to do.
[Essay, II, XXI, 47]
The difficulty for Locke, now, is this. Clearly, ‘suspending’ the satisfaction of some present desire, whatever else it may be, must surely qualify as an action of some kind and, presumably, one that we can engage in voluntarily. But if it is, thus, an action which is determined by the will, the question arises as to what determines the will when, by an act of will, we ‘suspend’ the satisfaction of some present desire. Locke has just said that ‘it is natural … that the greatest, and most pressing [uneasiness or desire] should determine the will to the next action; and so it does for the most part, but not always’. So what is the case when, by an act of will, we ‘suspend’ the satisfaction of a present desire? A dilemma seems to loom for Locke. Either he must say that, in such a case, what happens is what happens ‘for the most part’, namely, that our most pressing desire determines our will: but this seems to imply the absurdity that our most pressing desire may determine the ‘suspension’ of its own satisfaction. Or else he must say that, in such a case, what happens is what happens ‘not always’, namely, that we exercise our power to ‘suspend’ the satisfaction of our most pressing desire. But then we seem to be setting off upon a regress, of precisely the sort that Locke wished to avoid, for now we have invoked our power of ‘suspension’ in order to explain how an exercise of that power is not determined, on a given occasion, by our most pressing desire. It seems to me, in fact, that Locke, in appealing to our alleged power of ‘suspension’, is ultimately falling back on a libertarian conception of ‘free will’ which is intuitively appealing but completely at odds with his official doctrine, at least as this is interpreted by Yaffe and other commentators who see Locke as straightforwardly being a compatibilist. (By a ‘libertarian’ conception of free will, I mean one that holds that genuine free will exists but is not compatible with universal causal determinism, because free choices and the actions that follow from them are not causally necessitated by prior events. It is worth mentioning, incidentally, that Locke’s discussion of the ‘suspension’ of desire was a later addition to Chapter XXI, a fact which may help to explain some of the tensions that commentators find in this part of the Essay.)
There is one feature of Locke’s position, rightly emphasized by Yaffe, which I have not so far mentioned. This is Locke’s contention that the mere recognition of some future happiness, whether in this life or the next, as being a good for oneself is not enough, by itself, to supplant the force of present pleasures or pains as determinants of one’s will. Deliberation, for Locke, is not a merely intellectual activity, but plays also the vital motivating role of making contemplated future pleasures or pains sufficiently vivid that they can come to weigh in the balance against present pleasures and pains and, at least in some cases, outweigh them and thereby determine one’s will in one’s own best interest:
And thus, by a due consideration and examining any good proposed, it is in our power, to raise our desires, in a due proportion to the value of that good, whereby in its turn, and place, it may come to work upon the will, and be pursued.
[Essay, II, XXI, 46]
However, interesting and important though this element in Locke’s position undoubtedly is, it does nothing to resolve the difficulty that he faces in explaining how it is that one can ‘suspend’ the satisfaction of present desires in order to enable deliberation (or ‘examination’) to have the effect on one’s will that he describes.
It is worth recalling that Locke’s discussion of human action occurs in a chapter entitled ‘Of Power’. Locke sees the notions of agency, causation and power as being intimately related to one another, as indeed they plausibly are. Causation is clearly involved in many, if not all, instances of human action, as well as in processes taking place in the realm of wholly inanimate things. However, when we ask what exactly causation is, we are apt to find ourselves perplexed. Present-day philosophers tend to assume that causation is, fundamentally, a relation between events – and some of our ways of talking about causation seem to support this view, as when we say, for example, that the explosion of a bomb caused the collapse of a bridge: for explosions and collapses are clearly species of events. But other ways of talking about causation suggest a rather different view, as when we say that the bomb caused the collapse of the bridge – for a bomb is an object, or individual ‘substance’, rather than an event. And it is to things like bombs that we attribute causal powers, such as the power to destroy things like bridges. According to this way of thinking about causation, it is, in the final analysis, substances that cause effects of various kinds, which they do by exercising or manifesting various of their causal powers. Moreover, if one adopts this view of the matter, it is natural to think of human agency along the same lines, conceiving human agents as being substance-causes – or, more specifically, as so-called ‘agent-causes’ – of various effects, brought about by the exercise of a distinctive kind of power.
Because present-day philosophers tend to think of causation exclusively as a relation between events, some of them are prone to assume that Locke himself is no advocate of substance-causation and so, a fortiori, no advocate of agent-causation – not wishing, perhaps, to foist upon Locke an approach to causation and agency which they deem to be philosophically untenable. This tendency seems to me at once somewhat anachronistic and philosophically less well motivated than its representatives appear to think. It is a tendency that is exhibited, for example, by Gideon Yaffe, who expressly represents Locke as being a thoroughgoing event-causalist. I suspect, however, that commentators like Yaffe, in declining to interpret Locke as a substance-causalist, have an inappropriate conception of substance-causation in mind. A substance-causalist holds that causation is primarily a relation between individual things or substances – the bearers of properties – and only in a derivative sense, at best, a relation between events. This is a natural position to hold for a philosopher who, like Locke, puts causal powers and liabilities at the heart of his doctrine of causation: for such powers and liabilities belong primarily to individual substances rather than to events. By exercising their causal powers and manifesting their causal liabilities, individual substances act on, and are acted upon by, other individual substances. Seen in this light, the canonical form of a singular causal statement should be taken to be something like this: ‘Substance S1, by Fing, caused substance S2 to G’ – for instance, ‘The sun, by radiating heat, caused the lump of wax to melt’, or ‘The bomb, by exploding, caused the bridge to collapse’. And, indeed, Locke himself speaks very much in these terms when he talks about voluntary human action, as when he defines what is meant by ‘will’, ‘volition’ and ‘voluntary’:
[W]e find in our selves a Power to begin or forbear, continue or end several actions of our minds, and motions of our Bodies, barely by a thought or preference of the mind ordering, or as it were commanding the doing or not doing such or such a particular action. This Power … we call the Will. The actual exercise of that power … we call Volition or Willing. The forbearance or performance of that action, consequent to such order or command of the mind, is called Voluntary. And whatsoever action is performed without such a thought of the mind, is called Involuntary.
[Essay, II, XXI, 5]
For here the implication is that in an episode of voluntary human action, the agent, by willing – that is, by exercising his or her will – causes, for example, motion or rest in some part of the agent’s body. I set aside, for the time being, the complication that, for Locke, human persons are not in the most fundamental sense substances, but will return to this matter later.
Now, it is true enough that Locke also uses the language of event-causation, even in the context of talking about voluntary human action. For instance, he remarks at one point, in Book IV, that
all our voluntary Motions … are produced in us only by the free Action or Thought of our own Minds … For example: My right Hand writes, whilst my left Hand is still: What causes rest in one, and motion in the other? Nothing but my Will, a Thought of my Mind: my Thought only changing, the right Hand rests, and the left Hand moves.
[Essay, IV, X, 19]
Here Locke is saying that a volition or act of will is a cause of a motion in one of my hands – and volitions and motions plausibly belong to the category of events. However, it is perfectly consistent for a substance-causalist to speak in event-causation terms, since the substance-causalist’s thesis is not – or should not be – that there is no such thing as event-causation, but only that the notion of substance-causation is prior to that of event-causation and that the latter notion is derivable from the former. For instance, it may be held that to say that one event, c, is a cause of another event, e, is just to say that there is some substance, S1, and some substance, S2, such that, for some manner of acting, Fing, and some other manner of acting, Ging, c consisted in S1’s Fing and e consisted in S2’s Ging, and S1, by Fing, caused S2 to G (see Lowe 2002, ch. 11). Thus, for example, to say that a certain volition caused a certain motion is, by this account, just to say that a certain agent, by willing, caused some body-part to move in a certain way. In other words, a substance-causalist may happily resort to event-causation talk simply as a kind of shorthand, without compromising his thesis that substances, rather than events, are the primary relata of causal relations.
However, some of the recent literature on ‘agent-causation’ clearly encourages the thought that any substance-causalist is committed to the improbable thesis that, in an episode of substance-causation, a substance causes an event to occur without itself doing anything whatever to cause that event. It seems to be this conception of substance-causation that Yaffe, for example, has in mind when he mentions, only to dismiss, a certain line of thought which, as he puts it,
pushes toward a substance-causal reading of Locke’s account of action: A motion is begun by a substance just in case the motion, or the modification of the substance by virtue of which the motion is caused, is caused by the substance itself and not by any modification of the substance.
[Yaffe 2000, p. 81]
However, Yaffe seems to be assuming here that we have only two mutually exclusive alternatives to choose from: (1) to follow the event-causalist and say that a modification of the substance – the substance’s being F, or its Fing – is directly or indirectly the cause of a certain motion, or else (2) to adopt what Yaffe apparently takes the substance-causalist’s position to be and say that ‘the substance itself’ and not any modification of it is directly or indirectly the cause of the motion. But, as I hope my earlier remarks make clear, this is a false antithesis, for a sensible substance-causalist can say that the way in which ‘the substance itself’ causes the motion is precisely by Fing, that is, by acting or becoming ‘modified’ in some specific way – and that for this very reason we can also say, albeit only in a derivative sense, that the modification is a ‘cause’ of the motion.
I have dwelt upon this issue at some length only because commentators like Yaffe seem so keen to interpret Locke in event-causation terms to the exclusion of any substance-causal reading of him, almost as if this were necessary in order to make Locke’s position at all acceptable to the present-day philosophical reader. Yaffe himself acknowledges, in fact, that ‘Locke did not distinguish explicitly between event causation and substance causation’ (Yaffe 2000, p. 154, n. 16), but then adds, rather revealingly, ‘I believe (somewhat controversially, I know) that an appeal to substance causation should be avoided for philosophical reasons, and hence the account … that I offer here invokes only event causation’ (Yaffe 2000, pp. 154–5, n. 16). I would only say that this illustrates the dangers of interpreting the great philosophers of the past in the light of present-day perceptions of what constitutes a philosophically tenable thesis on some subject. One such danger is that we do not learn as much as we might from the texts of such authors, whose perennial value consists not least in their ability to prompt us to challenge some of the most deep-seated philosophical prejudices of our own time.
An important issue that arises in attempting to understand Locke’s theory of voluntary action is the question of whether, for Locke, it is both necessary and sufficient for an action’s being voluntary that the event concerned should have been caused by a volition of the agent to perform such an action – or, to put the same point in substance-causation terms, that the agent should have caused the event concerned by willing to perform such an action. That Locke held this to be a necessary condition for voluntariness seems to be borne out by, for example, the passage from Book IV of the Essay quoted a little earlier. But an important consideration that might make one hesitate to say that Locke also thought this to be a sufficient condition for voluntariness is that this would apparently render Locke’s account vulnerable to counter-examples involving so-called ‘deviant causal chains’, of the kind supplied by Donald Davidson’s famous case of the nervous climber (see Davidson 1980). I shall describe this case in a moment, but first I want to declare an interest in the issue at stake here. In earlier work (Lowe 1986, p. 150), I maintained that the following schema captures the essence of Locke’s conception of what it is for an agent S to act in a manner A voluntarily:
S As voluntarily if and only if an A-result is caused by a volition of S’s to A.
Here, as I say, S is an agent and A is a manner of acting, such as, for example, raising one’s arm. I shall explain the notion of an ‘action-result’ more fully later, but in this instance it would be an event of one’s arm’s rising. So, in this particular instance, my contention was that, according to Locke, S raises his arm voluntarily if and only if an event of S’s arm’s rising is caused by a volition of S to raise his arm. In saying this, I committed myself to the view that, for Locke, causation by a volition of the agent is both necessary and sufficient for voluntary action – in short, that Locke espoused both of two claims, which we may call, respectively, the necessity claim and the sufficiency claim. In his recent book, Gideon Yaffe agrees with me about the necessity claim, but not about the sufficiency claim, remarking:
The account of voluntary action that Lowe attributes to Locke is not philosophically sound, because causation by volition simply isn’t sufficient for voluntariness. The account runs into difficulties, in particular, from cases involving deviant causal chains.
[Yaffe 2000, p. 104]
In my earlier work, I acknowledged the possibility that the account might suffer from this difficulty, but commented that ‘I propose to waive such issues precisely because Locke was not addressing them’ (Lowe 1986, p. 161, n. 3). That still seems to me to be a perfectly reasonable comment. However, since Yaffe himself has raised the matter, I shall now address it and argue that, in fact, it is not as clear as Yaffe supposes that ‘causation by volition simply isn’t sufficient for voluntariness’, when one takes every relevant consideration into account. But first let us return to Davidson.
In Davidson’s famous example, a climber is desperately holding on to a rope supporting his companion, strongly desiring to let go of it and believing that if he does so his companion will fall to his death – but this desire and belief so unnerve him that they actually cause him to let his companion fall. The problem is that, according to a causal theory of intentional action like Davidson’s, it would seem that the climber must be said to have let his companion fall intentionally – because his action of letting go was apparently caused by an appropriate combination of belief and desire – and yet our intuition is to say that this agent in fact acted unintentionally. Now, Davidson himself is no volitionist and his example is meant to illustrate something about the concept of intentional, rather than voluntary, action. But Yaffe adapts it to his own purpose thus:
If I have a volition to let go, and reflection on the fact that I had such a volition causes me to become so nervous that I let go, then my volition caused my letting go, despite the fact that I did not do so voluntarily.
[Yaffe 2000, p. 104]
As with regard to the distinction between event-causation and substance-causation, Yaffe acknowledges that nothing in Locke’s text suggests that he was alive to this issue and hence that the problem of deviant causal chains doesn’t really constitute evidence against the thesis that, for Locke, causation by volition is sufficient for voluntariness. Yaffe thinks, as we shall see, that there is, none the less, textual evidence of another sort against the thesis. But why even mention the problem of deviant causal chains, in that case? The thought seems to be that Locke’s supposed grounds for denying the sufficiency claim can, in fact, be brought to bear on the problem of deviant causal chains and help in its solution, thereby demonstrating the strength of his account, as interpreted by Yaffe. However, I am rather doubtful whether the problem of deviant causal chains really does beset the sufficiency claim in any case: Yaffe is too ready, on the basis of a very sketchily drawn example, to suppose that it does. A careful volitionist will take pains to specify very precisely the intentional content of the volition that is causally implicated in any instance of voluntary action. In the case of a climber who voluntarily lets go of the rope bearing his companion, what he wills to do is to cause his grip to loosen now, directly by this very act of will – so that the intentional content of the volition or act of will has a self-referential aspect (see Lowe 1996, pp. 149ff.). But such a description does not fit the case of the nervous climber, who may plausibly be conceived to possess a general desire to let go sometime soon and be so upset by the thought of having such a desire that he unintentionally lets go now, but who cannot coherently be conceived to will to let go now, directly by means of this very act of will, and yet let go in the way that he willed to only as an unintended consequence of reflecting on the fact that he had such a volition. For if such a volition does not immediately give rise to a letting go, but only indirectly via an episode of reflection and consequent state of anxiety, then it does not cause what its own intentional content requires it to cause in order for the ensuing action to qualify as voluntary.
But what more is required, according to Yaffe’s interpretation of Locke’s position, for an act to be voluntary than causation by volition? Yaffe begins by distinguishing between ‘a volition causing an action, and an action satisfying the content of a volition’ (Yaffe 2000, p. 108). Then he says that, for an action A to satisfy the content of an agent’s volition to A,
the agent’s A-ing must have come about as a result of the [agent’s] … volition. Further, the agent’s conception of her volition as the exertion of a power that is, usually, an active power to A must be confirmed by the occurrence of the action. The action’s occurrence will not confirm the agent’s conception of her volition unless the volition causes her action in a particular way, in the way that the agent expects her volition to cause her action.
[Yaffe 2000, pp. 108–9]
Finally, Yaffe proposes that, for Locke, an action A of an agent S is voluntary if and only if A ‘satisfies a volition to A on the part of S’ (Yaffe 2000, p. 109). This account of voluntary action, Yaffe maintains, is not vulnerable to counter-examples involving deviant causal chains. For instance, in the nervous climber case, ‘the letting go was not caused by the volition in the appropriate way’ (Yaffe 2000, p. 110), because ‘if my letting go is accounted for in part by appeal to the fact that I got nervous, then the aspect of myself that explains why I got nervous … is also involved in accounting for the occurrence of my letting go’ (Yaffe 2000, p. 110). However, it may by now seem that Yaffe’s proposed solution to the problem of deviant causal chains boils down, in effect, to something very much like my own earlier proposal for dealing with this supposed problem: in both cases, it may seem, the key lies in exploiting an essential feature of the intentional content of a volition, namely, that it is always a volition to cause an occurrence ‘in a particular way, in the way that the agent expects her volition to cause [the occurrence]’, to use Yaffe’s own phrase. This phrase might seem designed to capture once more the self-referential aspect of volitional content. But then my point would be that when such a constraint on volitional content is acknowledged, it may be argued, as I argued earlier, that the sufficiency claim is not, after all, vulnerable to counter-examples invoking deviant causal chains.
However, on closer inspection, it is clear that the two proposals are in fact significantly different. Yaffe’s proposal does not appeal to a constraint on what we might call first-order volitional content, but rather to a condition on the agent’s ‘conception’ of the causal role of his or her volition and thereby to a constraint on the content of a second-order mental state that is distinct from the volition itself. This connects with another peculiar feature of Yaffe’s understanding of Lockean volitions, namely, his contention that volitions, for Locke, are not intrinsically volitional in nature. As Yaffe puts it:
It is a necessary condition for a particular act of my mind to count as a volition that I conceive of it in a particular way: namely, as an exercise of my ability to control my body (or some other part of myself) … So, an act of the mind does not count as a volition simply by virtue of its intrinsic, nonrelational properties. That is, we could demote a mental act from its status as a volition … merely by changing the surrounding psychological facts.
[Yaffe 2000, pp. 89–90]
Yaffe calls this condition ‘the Conception Condition on volition’, and in support of Locke’s adherence to it he cites, for example, the following passage from the Essay:
Volition, ‘tis plain, is an Act of the Mind knowingly exerting that Dominion it takes it self to have over any part of the Man, by imploying it in, or witholding it from any particular Action.
[Essay, II, XXI, 15]
However, there is a difficulty for Yaffe’s view that seems quite serious. This is that Locke takes a volition to be a species of thought – ‘a thought or preference of the mind ordering, or as it were commanding the doing or not doing such or such a particular action’ (Essay, II, XXI, 5) – and he also notoriously holds that all thought is self-conscious in character:
[C]onsciousness … is inseparable from thinking, and as it seems to me essential to it: It being impossible for any one to perceive, without perceiving, that he does perceive.
[Essay, II, XXVII, 9]
How, then, could Locke contemplate the possibility of one’s engaging in an act of mind which falls short of counting as a volition only because one does not think of this act of mind as an exercising of one’s ability to control part of one’s body – that is, which in every other way but this qualifies as a volition? For it could only qualify as a volition at all if it were a thought ‘ordering, or as it were commanding the doing or not doing such or such a particular action’ – it would have to be a thought with that sort of intentional content – and yet, at the same time, one would have to ‘perceive’ it (that is, think of it) as being just such a thought, because, according to Locke, all thought is conscious in this sense.
Returning to my own earlier proposal for dealing with the problem of deviant causal chains, I readily concede that there is no clear evidence that Locke himself thought that volitional content is ‘self-referential’ in the way that I suggest, although there is nothing in his position inimical to this idea. If the problem of deviant causal chains had occurred to him, he could certainly have adopted this sort of solution and thereby have defended the sufficiency claim. As for the question of whether there is clear evidence that Locke espoused the sufficiency claim – which Yaffe, of course, denies – I shall discuss that in the next section. Certainly, I think that there is no clear evidence that he rejected it. What I do maintain, though, is that the package that Yaffe offers on Locke’s behalf – a rejection of the sufficiency claim together with an acceptance of the Conception Condition – is not one that sits at all comfortably with some central tenets of Locke’s philosophy of mind and action.
I now want to address the main objection that Yaffe raises against attributing the sufficiency claim to Locke. But I should advise readers that this section of the present chapter unavoidably involves some very detailed exegesis and argumentation and can safely be omitted by those who are concerned only to understand in broad terms the character of Locke’s theory of voluntary agency. I have included the section because I think that the issues at stake are important for the philosophy of action and I hope that this discussion of them will provide a useful example of the way in which close engagement with a famous historical text can help to illuminate and enrich present-day philosophical concerns.
Yaffe’s main objection arises out of Locke’s remark that ‘[W]hatsoever action is performed without [a volition], is called Involuntary’ (Essay, II, XXI, 5). Yaffe initially canvasses two possible interpretations of this remark, both of which he intends to reject. According to what he calls the ‘First Interpretation’, the remark may be interpreted as merely making the claim that ‘absence of an appropriate volition is sufficient for the involuntariness of an action’ (Yaffe 2000, pp. 104–5). Call this claim claim (1). The first thing I want to point out is that Yaffe is wrong to assert as he does that claim (1) ‘is the contrapositive of the claim that causation by [an appropriate] volition is necessary for voluntary action’ (Yaffe 2000, p. 105). Call the latter claim claim (2). Claim (2) is, of course, what I earlier called the necessity claim and it is a claim that both Yaffe and I want to attribute to Locke. The reason why claim (1) cannot be the ‘contrapositive’ of claim (2) is that the contrapositive of a claim is logically equivalent to it and yet claim (1), unlike claim (2), makes no mention of causation.
A few words of explanation are called for here. One normally speaks of the ‘contrapositive’ of a conditional statement. The contrapositive of a conditional statement of the form ‘If p, then q’ is another conditional statement, of the form ‘If not q, then not p’ – and two such statements are standardly taken to be logically equivalent. When such a pair of statements is true, we may say that p is sufficient for q and q is necessary for p, or, equivalently, that not q is sufficient for not p and not p is necessary for not q. By extension, then, one might say that the statement that p is sufficient for q is the ‘contrapositive’ of the statement that not q is sufficient for not p and hence also of the equivalent statement that not p is necessary for not q. Now, claim (1) states that absence of an appropriate volition is sufficient for the involuntariness of an action and it will be seen that this is the contrapositive of the statement that non-absence of an appropriate volition is necessary for the non-involuntariness of an action – or, in other words, that presence of an appropriate volition is necessary for the voluntariness of an action. Claim (2), however, states instead that causation by an appropriate volition is necessary for the voluntariness of an action. So Yaffe is mistaken in asserting that claim (1) is the contrapositive of claim (2).
But let us move on. Yaffe urges next that the First Interpretation of Locke’s remark is unsatisfactory, on what seem to me to be the untenable grounds that ‘it means that Locke never (as far as I know) explicitly offered necessity conditions for involuntary action (that is, sufficiency conditions for voluntary action) … [and so] means that Locke left an important lacuna in his philosophy of action’ (Yaffe 2000, p. 105). As I shall make clear shortly, Locke undoubtedly did explicitly offer sufficiency conditions for voluntary action, quite independently of the remark now under discussion – and, oddly enough, Yaffe himself later acknowledges as much. Be that as it may, Yaffe now considers a Second Interpretation of Locke’s remark, according to which it is to be taken as meaning that absence of an appropriate volition is not only sufficient but also necessary for the involuntariness of an action. This Second Interpretation, he maintains, conflicts with my view that the sufficiency claim, as I call it, is attributable to Locke:
If the Second Interpretation captures Locke’s meaning, then he cannot have consistently held the view that Lowe attributes to him. He cannot, that is, have held that causation by [an appropriate] volition is sufficient for voluntariness.
[Yaffe 2000, p. 105]
Yaffe’s point seems to be that if absence of an appropriate volition is necessary for the involuntariness of an action, then it follows that presence of an appropriate volition is sufficient for the non-involuntariness – in other words, for the voluntariness – of an action, even if that volition plays no causal role in the production of the action. As Yaffe puts it, ‘Under the Second Interpretation, actions that are preceded by volitions that are not causally responsible for them are … voluntary’ (Yaffe 2000, pp. 105–6). However, this does not in fact conflict with the sufficiency claim but only with what I have been calling the necessity claim, which Yaffe himself is happy to attribute to Locke. This is easier to see if we formulate these claims as conditional statements. In this form, the sufficiency claim states that if an action is preceded by a volition that is causally responsible for it, then it is voluntary. But this is perfectly consistent with its also being the case that if an action is preceded by a volition that is not causally responsible for it, then it is voluntary. What the latter is not consistent with is the necessity claim, which states that if an action is voluntary, then it is preceded by a volition that is causally responsible for it. For from ‘If an action is preceded by a volition that is not causally responsible for it, then it is voluntary’ and ‘If an action is voluntary, then it is preceded by a volition that is causally responsible for it’ we may infer the absurd conclusion ‘If an action is preceded by a volition that is not causally responsible for it, then it is preceded by a volition that is causally responsible for it’.
We see, thus, that the Second Interpretation only creates a difficulty for an aspect of my account of Locke’s position with which Yaffe himself is in full agreement. It creates no more difficulty for me than it does for Yaffe and he is obliged to reject it for exactly the same reason that I am. As it turns out, Yaffe in fact rejects both of the canvassed Interpretations of Locke’s remark and favours a third, based on the proposal that I examined in the previous section. He surmises, however, that ‘Lowe would probably want to avoid the inconsistency between the Second Interpretation and his claim that, for Locke, causation by [an appropriate] volition is sufficient for voluntariness by returning to the First Interpretation’ (Yaffe 2000, p. 107). As we have just seen, in reality there is no such inconsistency for me to have to avoid, only an inconsistency between the Second Interpretation and the claim that both Yaffe and I want to make that, for Locke, causation by an appropriate volition is necessary for voluntariness. But to avoid this inconsistency I do not, in point of fact, favour Yaffe’s First Interpretation. Rather, I prefer to see in Locke’s remark an implicit affirmation of the claim that absence of causation by an appropriate volition is sufficient for the involuntariness of an action. Call this claim claim (3). Claim (3), by the way, really is the ‘contrapositive’ of the claim that causation by an appropriate volition is necessary for the voluntariness of an action – claim (2) above, or the necessity claim, as I have been calling it – and so is merely equivalent to that.
To see why I favour this interpretation of Locke’s remark that ‘[W]hatsoever action is performed without [a volition], is called Involuntary’, note that this remark immediately follows Locke’s assertion that ‘The … performance of [any] Action, consequent to [a volition], is called Voluntary’ (Essay, II, XXI, 5). Now, as I argued in the paper that Yaffe is criticizing, ‘consequent to’ here cannot plausibly be interpreted in a purely non-causal sense, implying mere temporal succession (see Lowe 1986, p. 150). Indeed, I see in this assertion of Locke’s the clearest evidence we have that he subscribed to what I have been calling the sufficiency claim – the claim that causation by an appropriate volition is sufficient for the voluntariness of an action. But, by the same token, when Locke speaks, in the immediately following remark, about an action’s being performed ‘without’ a volition, this too is not plausibly to be read in a purely non-causal sense – a fact, incidentally, that Yaffe himself later acknowledges, if somewhat obliquely (see Yaffe 2000, p. 111). Plausibly, what Locke has in mind here is an action which is not brought about by means of the agent’s will. For just as performing an action ‘with’ a volition is naturally construed as performing it by exercising one’s will, so an action’s being performed ‘without’ a volition is naturally construed as its not being performed by exercising one’s will. At the very least, it should be acknowledged that this reading of Locke’s remark is quite as natural as one which interprets it as asserting merely that absence of an appropriate volition is sufficient for the involuntariness of an action – claim (1) above. It should be noted, though, that claim (1) is implied by the reading of Locke’s remark that I favour. On my reading of Locke’s remark, it makes claim (3), that absence of causation by an appropriate volition is sufficient for involuntariness. But absence of an appropriate volition is obviously sufficient for absence of causation by an appropriate volition and so, given that the latter is sufficient for involuntariness, so too is absence of an appropriate volition sufficient for involuntariness. For if p is sufficient for q and q is sufficient for r, then p is sufficient for r.
I now come to the crucial question of whether Yaffe’s denial that Locke espoused – or, at least, was committed to – the sufficiency claim can be rendered consistent with the text of the Essay. I believe that it cannot, with any plausibility. Yaffe seems, however, to be confused about one vitally important matter here. He too notes that Locke’s remark that ‘[W]hatsoever action is performed without [a volition], is called Involuntary’ is immediately preceded by Locke’s assertion that ‘The … performance of [any] Action, consequent to [a volition], is called Voluntary’. But, without actually quoting this preceding sentence at this point – although he does so earlier (see Yaffe 2000, p. 99) – Yaffe describes it here as the ‘claim made in the preceding sentence (namely, that causation by volition is necessary for voluntariness)’ (Yaffe 2000, p. 105, my emphasis). And, indeed, Yaffe offers as a further reason for rejecting the First Interpretation of Locke’s subsequent remark concerning involuntary action that, on this interpretation, that remark would be ‘just a reiteration of [the] claim made in the preceding sentence’ (Yaffe 2000, p. 105). But this cannot be correct. The preceding sentence, as we have just seen, is ‘The … performance of [any] Action, consequent to [a volition], is called Voluntary’ – and this cannot be construed merely as a necessity claim concerning voluntariness. For it clearly implies that if the performance of an action is consequent to a volition, then it is voluntary and thus that an action’s being performed consequent to a volition is sufficient for its being voluntary. This still leaves it open that it may be intended to imply necessity as well as sufficiency and also leaves open whether ‘consequent to’ should be construed in a causal or in a merely temporal sense – or perhaps in some third sense. But that it constitutes a sufficiency claim concerning voluntariness cannot be in doubt.
Now, if ‘consequent to’ should be construed in either a causal or a merely temporal sense, Yaffe clearly cannot maintain that Locke is not committed to what I have been calling ‘the sufficiency claim’ – that is, the claim that causation by an appropriate volition is sufficient for the voluntariness of an action. For we have just seen that Locke is undoubtedly committed to a sufficiency claim concerning voluntary action, namely, that being ‘consequent to’ an appropriate volition is sufficient for the voluntariness of an action. But precisely because the temporal sense of ‘consequent to’ is weaker than the causal sense, Locke is committed to what I have been calling ‘the sufficiency claim’ whichever sense is attributed to ‘consequent to’. The point is once again a simple one concerning the logic of sufficient conditions. Being caused by an appropriate volition is a sufficient condition for being preceded by an appropriate volition, because any cause necessarily precedes its effect. Hence, if Locke means that being preceded by an appropriate volition is sufficient for the voluntariness of an action – in accordance with the merely temporal interpretation of ‘consequent to’ – then he is committed also to holding that being caused by an appropriate volition is sufficient for the voluntariness of an action.
Clearly, then, Yaffe must deny that Locke understands ‘consequent to’ in either the causal or the merely temporal sense. And that is precisely what he does do, in the following passage:
[Locke] says that the voluntary actions are those that are performed ‘consequent to’ … a volition. This is a peculiar phrase and suggests that he wanted to avoid using the language of causation when offering his sufficiency condition for voluntary action. The remark is, of course, inconclusive, but I think that it suggests that things are more complicated than Lowe allows.
[Yaffe 2000, p. 109]
Notice that Yaffe here contradicts his earlier mistaken supposition that the sentence containing ‘consequent to’ merely states a necessary condition for voluntariness and at the same time undercuts the main reason he offered for rejecting the so-called ‘First Interpretation’ of Locke’s remark about involuntariness. I would only add that Yaffe’s attempt to construe Locke’s phrase ‘consequent to’ as having neither a causal nor a merely temporal sense seems to me speculative in the extreme. If Locke really did have a special meaning in mind, would he not have taken pains to spell it out carefully, given the evident importance of the issue at stake? Yaffe’s own construal of ‘consequent to’ is, of course, intimately bound up with his proposed new account of Locke’s conception of voluntary action, discussed in the previous section, and so stands or falls with that proposal. And I have already explained why I think that the proposal is untenable. That leaves us, I think, with no good reason to deny that Locke either espoused or was at the very least committed to the sufficiency claim.
I have considered it worthwhile to go into this matter in such detail for two reasons, in addition to those mentioned at the beginning of the section. The first is that it really is very important to get as clear as we can about how Locke understood the relationship between volition and action, not least because this has an intimate bearing on how we understand his views about freedom of action and freedom of will. The second is that, as will no doubt have become evident, it is very easy to get confused in discussions concerning necessary and sufficient conditions. Moreover, since Locke himself does not explicitly use the terminology of necessary and sufficient conditions, great care needs to be taken in attributing to him claims made in those terms. I hope that the discussion in this section has provided, at the very least, an object lesson in the perils attending philosophical debate about such conditions.
As we saw in Chapter 3, much is often made of Locke’s views concerning the relationship between the identity of persons and their capacity to remember episodes in their past lives. Rather less tends to be said, however, concerning links between his conception of personal identity and his views about free agency. But such links are certainly very well worth exploring. Locke, we should recall, denies that personal identity consists in the identity of substance of any kind, material or spiritual, making the notoriously obscure and controversial claim that it consists instead in ‘consciousness … as far as this consciousness can be extended backwards to any past Action or Thought’ (Essay, II, XXVII, 9). However, although Locke’s discussion of personal identity explicitly concentrates on ‘backward-looking’ consciousness, of the sort that characterizes memory, it is clear that ‘future-directed’ consciousness of the sort that is involved in action is implicitly quite as central to Locke’s overall conception of personhood and personal identity. Moreover, it is important that Locke’s denial that personal identity consists in the identity of substance of any kind should be understood in context. As we saw in Chapter 3, his point is that human persons are not persons in virtue of their purely material or biological characteristics, nor in virtue of any purely spiritual characteristics that they may have, but rather in virtue of their psychological characteristics. But this is not best taken to imply that, for Locke, human persons just are, or are ‘reducible to’, certain ‘bundles’ of psychological states, or temporally extended sequences of such ‘bundles’: they may still qualify as ‘substantial’ beings and indeed as physical beings, even though their persistence over time does not rest upon the persistence of any particular material or spiritual substance. Here we should beware, once more, of the danger of interpreting Locke anachronistically, this time in the light of conceptions of personhood and persistence that have really been advanced explicitly only in recent years. I do not, for example, think that it is very helpful, in trying to explain Locke’s conception of personal identity, to talk in terms of ‘consciousness between person-stages’, in the way that some modern commentators tend to do (see, for instance, Yaffe 2000, p. 120).
In one of the relatively few recent discussions of the connections between Locke’s conception of personal identity and his views about human agency, Gideon Yaffe interestingly raises what he calls the ‘Where’s the Agent Problem’ (Yaffe 2000, p. 121), which he thinks afflicts all causal theories of agency. The problem is supposed to be something like this. A causal theory of agency represents actions as being caused by antecedent mental episodes of a suitable kind – in the case of volitionist theories, volitions or willings. But these mental causes of actions, it is said, are not themselves actions and have in turn other non-actions as causes, in an unbroken chain of event-causation which stretches back indefinitely far in time. Hence, an agent’s actions appear to be nothing more than elements in the ongoing stream of event-causation and the ‘agent’ seems thereby to be denied any genuine role as the author of his or her actions. Yaffe himself seems to think that the Conception Condition which he attributes to Locke helps with this problem, because the fact that, according to this view, a volition must be the object of self-conscious awareness to count as a volition makes it somehow especially internal to the person whose volition it is and thereby provides a robust sense in which an action caused by such a volition has that person as its agent and author.
I think it is wrong to suppose that all causal theories of agency, or even all volitional theories, have the structure that Yaffe describes. First of all, many causal theorists would deny, in my view rightly, that bodily actions – such as the action of raising one’s arm – have mental episodes, such as volitions, as their causes. Such theorists would distinguish between the action of raising one’s arm – that is, causing one’s arm to rise – and the event of arm-rising that is an essential ingredient or part of that action. To distinguish the two, the arm-raising may be called the action proper and the arm-rising the ‘action-result’. Yaffe himself always talks in terms of actions and their causes, never in terms of action-results, because he thinks that Locke himself never used the distinction between actions and action-results (Yaffe 2000, p. 157, n. 24). However, although Locke does not explicitly make the distinction, it would be surprising if it were not implicitly present in his writing, since it is built into the very semantics of natural language – for instance, in terms of the distinction between transitive and intransitive senses of the verb ‘move’ (see Hornsby 1980, pp. 2–3). To moveT an object O is to cause O to moveI, where ‘moveT’ and ‘moveI’ are, respectively, the transitive and intransitive forms of the verb ‘move’. Recall again the passage from the Essay quoted earlier, where Locke says that
all our voluntary Motions … are produced in us only by the free Action or Thought of our own Minds … For example: My right Hand writes, whilst my left Hand is still: What causes rest in one, and motion in the other? Nothing but my Will, a Thought of my Mind.
[Essay, IV, X, 19]
Here Locke is saying that a movementI in my hand is caused by a volition of mine. It is true that he speaks of this motion or movementI in my hand as being ‘voluntary’, but that doesn’t necessarily imply that he takes my action in this case to be the movementI in my hand, as opposed to my movingT my hand by an act of will. In any case, whatever Locke’s own view may or may not be, I would urge that the proper view for a volitionist to take in this matter is that my action in this case is my movingT my hand by willing to moveT it in a certain way and that what is caused by my so willing is not my action but the action-result of my hand’s movingI in the appropriate way. The upshot is that, on this view of the matter, actions such as hand-movingsT are not, after all, mere elements in the ongoing stream of event-causation, because they are not events but causings of events, and do not have events – not even volitions – as their causes (see Lowe 2002, ch. 12).
To this it may be responded that still, on this view, action-results and the volitions that are their causes are mere elements in the ongoing stream of event-causation, so that the real difficulty has not been allayed. But here I would question another assumption that Yaffe makes on behalf of causal theorists of agency, namely, that the mental causes which they invoke in their theories are not in any special sense actions, but mere ‘happenings’: ‘causal theorists of agency build doings out of happenings and causal relations’ (Yaffe 2000, p. 121). Well, that may be true of some causal theorists, such as Donald Davidson, but it is certainly not true of all volitionists, many of whom certainly qualify as causal theorists. A volitionist may contend that although willing is not causing, in the way that arm-raising is causing one’s arm to rise, it is nonetheless a kind of action, not a mere happening or event. Carl Ginet is an example of such a volitionist, though one may not find altogether felicitous his description of volitions as having an ‘actish phenomenal quality’ (see Ginet 1990, pp. 11ff.). I would rather make the point myself by assimilating willing to trying or endeavouring (without wishing to identify the two notions exactly), for trying incontrovertibly falls within the sphere of genuine action (see Lowe 1996, pp. 157ff.). The upshot is that there is plenty of scope for a causal theorist of agency to reject the idea which Yaffe thinks gives rise to the ‘Where’s the Agent Problem’, namely, that all doings are built out of happenings and causal relations – especially if one takes into account the considerations raised earlier in favour of the primacy of substance-causation.
Despite all of the problems undoubtedly besetting Locke’s account of personhood and personal identity, some of which we examined in Chapter 3, my own opinion is that his views concerning human agency strengthen rather than weaken that account, and do so precisely because those views are most naturally construed as belonging to the substance-causation tradition. Locke’s conception of the person as agent is, to my mind, considerably more compelling than his conception of the person as a repository of memories, even though the latter is typically given much more prominence by present-day commentators on Locke’s work.
Locke is a volitionist: he holds that voluntary action consists in the causation of certain bodily or mental events by volitions or acts of will, that is, by exercises of a special mental power or capacity, the will. He is also, I believe, a substance-causalist, holding that it is substances, rather than events, that are in the most fundamental sense causes of anything. For Locke, then, voluntary human action is a special case of substance-causation. An agent is free to act in a certain way, according to Locke, just in case whether or not he acts in that way depends on whether or not he wills to act in that way. On the other hand, freedom of the will, he is inclined to say, is at best a misnomer: but he agrees that there must be more to human freedom in the fullest sense of the term than is simply involved in our being free to act in various ways. What this ‘more’ exactly is, according to Locke, is hard to pin down, although it is clear that he thinks that it has something to do with our ability, at least sometimes, to ‘suspend’ the satisfaction of our most pressing desires. Whether Locke is properly to be regarded as a compatibilist or, perhaps, as an implicit libertarian is also very hard to determine. What is very clear, however, is that Locke thought that our capacity for free agency is absolutely central to our status as persons, quite as much as our capacities for reason, reflection and memory. The chapter ‘Of Power’ in the Essay resists easy interpretation, because Locke himself revised it extensively and was never entirely satisfied with it. But it is also one of the most philosophically rewarding chapters of that work and repays constant re-reading.
Chappell, Vere 1994: ‘Locke on the Freedom of the Will’, in G. A. J. Rogers (ed.), Locke’s Philosophy: Content and Context (Oxford: Clarendon Press).
Clarke, Randolph 2003: Libertarian Accounts of Free Will (New York: Oxford University Press).
Magri, Tito 2000: ‘Locke, Suspension of Desire, and the Remote Good’, British Journal for the Philosophy of History 8, pp. 55–70, reprinted in Udo Thiel (ed.), Locke: Epistemology and Metaphysics (Aldershot: Dartmouth, 2002).