Suppose we accept the truth of hard determinism. Suppose, in other words, we come to believe that everything that happens is necessitated to happen, so that for every event that takes place there are conditions (such as antecedent physical causes plus the laws of nature, or our genetic make-up and environment, or the decrees and foreknowledge of God) whose joint occurrence is logically sufficient for the occurrence of that event. And suppose further that we come to believe that, since everything that happens is determined to happen, there is no free will, or at least no free will of the sort required for genuine responsibility, blameworthiness, or desert. What consequences (other than the repudiation of some form of moral responsibility and concomitant notions) would the adoption of such a view have for morality?
Many, of course, have thought that nothing less than moral nihilism would result: `If all is determined, everything is permitted'. Peter van Inwagen, for example, writes that, «If there is no free will, then morality as it is ordinarily conceived is an illusion».Foot note 1 Similarly, Ishtiyaque Haji has argued that in a deterministic world we would be bereft of any `moral anchors', that is to say, moral deontic normative statuses like those of being morally right, wrong, or obligatory.Foot note 2 There are, however, a growing number of dissenters who maintain that the consequences for morality, given hard determinism, are not as dire as is ordinarily feared. Indeed, some have gone on to argue that hard determinism offers us a `purer' or `higher' form of morality than would otherwise be available.Foot note 3
In looking for a way to settle this dispute, I will concentrate on four central moral notions: moral responsibility (and the allied notions of praiseworthiness and blameworthiness), moral obligation, moral rightness and wrongness, and moral goodness. The question, then, to be addressed is: Must the hard determinist adopt a `scorched earth' policy, setting all such moral categories to the flames, or can some or even all of them be retained or at least rehabilitated?
At the outset, however, I would like to counter a possible line of response to this question. One might contend that even if there is little or no evidence that we are free and morally responsible, the consequences of denying moral responsibility are so grave that we should continue living under the (perhaps false) impression that we were free in the sense required for moral responsibility. This line of thought goes as far back as Kant, and has been vigorously defended in more recent times by Saul Smilansky.Foot note 4 According to Smilansky, libertarian free will does not exist, but once we remove it from our picture of reality, our fundamental values, practices and attitudes -- such as the belief in our potential for blameworthiness according to what we do, our appreciation of achievement in ourselves and others, the sense of value and meaning in our lives, and so on -- would be placed under great risk, if not largely destroyed. Since the consequences for humanity of widespread belief that we lack libertarian free will would be utterly destructive, Smilansky exhorts us to maintain and even promote the illusion that we possess such free will. As he puts it, «In the context of free will,...we cannot live well with the full truth. So, we shall have to make do with illusion, to a significant extent.»Foot note 5
This is an interesting response, but I think it can be easily dispensed with. Firstly, I suspect that most people would find it psychologically difficult to engage in the required sort of pretense. Smilansky, to be fair, is not suggesting that the illusory belief in libertarian free will needs to be induced in people, for as he is aware, that belief is already part of our cognitive endowment. For Smilansky, however, what is crucial is that there is a `double illusion' in place: most people hold the false belief that we have responsibility-entailing free will, while being oblivious to the fact that their belief is false. There is therefore a widespread illusion about the presence of illusion.Foot note 6 The question, however, remains as to what people with a psychology like ours would likely do once they discovered that libertarian free will is merely an illusion. Smilansky's answer seems to be that the illusion of free will fulfills such important emotional needs in people that it is unlikely that many people would be able to accept the harsh reality with regard to free will.Foot note 7 But surely some people will hit upon the truth and will be brave enough to accept it -- Smilansky himself is a case in point. Would these brave souls be required to engage in some sort of self-deception in order to carry on with their daily lives? It is not clear, in other words, how someone in Smilansky's shoes, who is fully aware of the illusory character of free will, can (without discounting this awareness) continue thinking of himself and others as free and morally responsible agents.Foot note 8
Secondly, Smilansky's line of response presupposes that much of what is important in human life (e.g., morality, interpersonal relations) would be undermined if we came to believe that hard determinism is true. Smilansky, to be sure, has his reasons for thinking that an acceptance of determinism, including hard determinism, has these damaging effects. However, if hard determinism has no such implications, then the proposed pragmatic justification for belief in free will and moral responsibility would be undercut. Although not a criticism of Smilanksy, this does indicate that it is incumbent on us to determine what the moral consequences of hard determinism in fact are before even considering whether we should opt for a pragmatic defence of `illusion'.
Thirdly, when we believe that someone is blameworthy for committing a serious moral offense, we are normally angry with them, we hold them in contempt, and we endeavour to punish them by, for example, incarcerating them. Clearly, these actions (of expressing anger and contempt towards someone, of punishing them) tend to be very harmful to the offender. But it is arguably the case, though some wish to deny this, that it is permissible to intentionally inflict harm on someone when they do not deserve to be harmed (i.e., when they are not genuinely blameworthy for their behaviour) only if one has very good reason to believe that by inflicting such harm they will be procuring some great benefit for the sufferer.Foot note 9 However, in many cases of the relevant sort, we would lack any reason to think that our treatment of the offender is likely to benefit them -- indeed, in some cases (e.g., those involving the death penalty) we would have good reason to think that our treatment will bring more harm than good to them.Foot note 10
(An autobiographical note at this point might be appropriate before proceeding further. This investigation is intended as a prelude to a somewhat distant topic in the philosophy of religion, viz., the problem of evil. In particular, I want to see how this problem can be tackled by a theist who also happens to be a hard determinist. But since the problem of evil is at bottom a moral problem, it will be helpful to first ascertain what the moral world of the hard determinist -- and by extension the theistic hard determinist -- looks like.)
My procedure in this paper is to take the views of Derk Pereboom, one of the ablest contemporary defenders of hard determinism, as my point of departure.Foot note 11 Pereboom argues that some, but not all, of our central moral categories can be salvaged from the ruins of hard determinism. My view is that Pereboom does not go far enough. In particular, I will argue that all of our central moral notions, and not merely some of them, can be reconciled with the hard determinist outlook -- though such reconciliation, I argue, will be possible only after the moral notions in question have, at least to some degree, been reconceived. But first it will be important to get clear about some terminology.
I. Hard Determinism and Hard Incompatibilism
Following William JamesFoot note 12, determinists are usually grouped into two camps. There are soft determinists or compatibilists, who believe that determinism does not undermine any free will or responsibility worth having, and there are hard determinists, who take a harsher line: determinism is true and therefore free will (or at least the kind of free will required for moral responsibility) does not exist. More perspicuously, hard determinism -- at least as classically expressed by such thinkers as Spinoza and HolbachFoot note 13 -- is committed to the following three theses:
(1) Free will (in the strong sense required for moral responsibility and moral desert) is not compatible with determinism.
(2) There is no free will (in the above sense), because:
(3) All events are determined.
Contemporary hard determinists, however, are unlikely to accept the third of these theses, preferring instead to remain non-committal about the truth of universal determinism (particularly in the light of developments in twentieth-century physics). Derk Pereboom is a case in point. He describes himself as a `hard incompatibilist' rather than a `hard determinist', where hard incompatibilism involves a commitment to the following two theses:
(4) All of our actions and choices are either alien-deterministic events (i.e., events such that there are causal factors beyond our control by virtue of which they are causally determined), or truly random events (i.e., events not produced by anything at all), or partially random events (i.e., events such that factors beyond the agent's control contribute to their production but do not determine them, and there is nothing that supplements the contribution of these factors to produce the events).
(5) An action is free in the sense required for moral responsibility only if the decision to perform it is not an alien-deterministic event, nor a truly random event, nor a partially random event.
The idea behind (5), or what Pereboom calls `the Causal History Principle', is that an agent's responsibility for an action is explained not by the existence of alternative possibilities open to her, but rather by the action's having a certain causal history -- viz., one which allows the agent to be the source of her action in such a way that she has control over the production of her decision to perform that action.Foot note 14 But according to (4), all our actions and choices lack the requisite kind of causal history, as they are produced by (deterministic or random) processes over which we have no control. From (4) and (5) it clearly follows that we do not have free will of the sort required for moral responsibility.Foot note 15
By contrast, classical hard determinism holds that we lack free will because we live in a fully deterministic world. Pereboom, however, does not want to commit himself to universal determinism. He would rather leave the matter open, arguing instead that, since our actions and choices are events that lie on the continuum from alien-deterministic through partially random to truly random events, we have no freedom in the sense required for moral responsibility. But as Pereboom is careful to point out, his position is not to be confused with the `no-free-will-either-way' theory advocated by contemporary hard determinists such as Galen Strawson and Richard Double. On this view, free will of the kind required for moral responsibility is incompatible with both determinism and indeteterminsm, and so irrespective of whether the world turns out to be deterministic or indeterministic, we could not have free will of the relevant sort. Pereboom disagrees with this assessment, arguing instead that the kind of indeterminacy required for agent-causal libertarianism is compatible with the sort of free will required for moral responsibility. In his view, however, the evidence of our best scientific theories suggests that no such indeterminacy is to be found. This, of course, means that our lack of free will is a merely contingent fact: we might have been undetermined agent-causes, and hence free and responsible agents.Foot note 16
In what follows I will speak only of `hard determinism', taking thesis (2) above as the kernel of this position.Foot note 17 On this reading, the hard incompatibilist and the `no-free-will-either-way' theorist belong to the same camp as their classical determinist predecessors, and this is not entirely inappropriate. For what they all have in common is the belief that there is no free will, in the responsibility-entailing sense, although each group arrives at this belief from a different direction. It is this belief, furthermore, that distinguishes them from both compatibilists and libertarians.Foot note 18
II. Moral Responsibility, Praise and Blame
It is often thought that the first and most obvious casualty in the moral realm, once hard determinism is adopted, is the idea that human beings are sometimes morally responsible for their decisions and actions. Indeed, this result seems to simply fall out from the account of hard determinism given above, as thesis (2) states that there is no free will of the sort required for moral responsibility. But what is it for a person to be morally responsible? Roughly speaking, to be morally responsible for something (e.g., an action, an omission, a character trait) is to be worthy of a particular kind of reaction -- typically, praise or blame. But when are such reactions of praise and blame appropriate or called for? Answers to this question give rise to two competing conceptions of moral responsibility. According to one view, the merit-based view, praise or blame is an appropriate reaction toward someone if and only if they merit -- in the sense of `deserve' -- such a reaction, given their behaviour and/or character traits. This is perhaps the commonsensical view of moral responsibility, but there is an alternative: the consequentialist view, according to which praise or blame is an appropriate reaction if and only if a reaction of this sort is likely to bring about a desired change in the agent and/or their behaviour.Foot note 19
Pereboom states that hard determinists must relinquish our ordinary view of ourselves as morally responsible in the merit-based sense, that is, as deserving blame for immoral actions and as deserving praise for actions that are morally exemplary.Foot note 20 In this regard, I think Pereboom is entirely correct. Even so, the hard determinist need not renounce moral responsibility altogether, but may instead reconceive our commonsense notion of moral responsibility in consequentialist terms. Indeed, this is the route normally taken by determinists, both hard and soft. J.J.C. Smart, for example, states that
Threats and promises, punishments and rewards, the ascription of responsibility and the nonascription of responsibility, have...a clear pragmatic justification which is quite consistent with a wholehearted belief in metaphysical determinism.Foot note 21
Borrowing one of Smart's examples, we can see how this pragmatic or consequentialist justification would operate in practice by supposing that a student, Tommy, fails to do his homework and as a result his teacher deems him to be morally responsible or blameworthy. On the view we are considering, the teacher's reaction would be appropriate only if her act of blaming Tommy is likely to influence Tommy's attitudes and conduct in certain desirable ways.Foot note 22 On the other hand, such a reaction on the part of the teacher would not be appropriate if Tommy suffered from some illness which prevented him from doing his homework, since in that case no amount of exhortation or threats would succeed in modifying his behaviour. But if Tommy's failure to do his homework is simply due to laziness, then his behaviour may well be open to influence, and the threats or punishments of his teacher would form part of the environment that helps influence Tommy's behaviour. In such a scenario, ascriptions of blame would have a definite and valuable purpose. As Smart sees it, then, ascriptions of moral responsibility, and therefore of praise and blame, have a purely instrumentalist rationale of influencing other people's behaviour.Foot note 23
Unfortunately, as Richard Arneson points out, this instrumentalist account is «the position everyone loves to hate.»Foot note 24 Various objections have been raised against this account, but here I will only address two of the more common ones.Foot note 25 The first of these is derived from an objection that is often raised against consequentialist ethical theories in general, viz., the objection that such theories do not give sufficient weight to the needs or rights of individuals, and as a result these theories allow or even promote the unjust or cruel treatment of people in certain circumstances. Peter van Inwagen echoes this objection when considering the following utilitarian theory of what it is to assign blame or to say `X is Y's fault' (with `X' denoting a state of affairs and `Y' referring to some agent):
(6) X is an unfortunate state of affairs and Y brought about X and, for some group of people, it would maximize the general welfare if they were to do something unpleasant to Y and to describe their motive to Y and to the public as follows: We are doing this unpleasant thing to Y because Y brought about X.
Van Inwagen has us consider a mentally retarded person who, ignorant of his own strength and goaded by cruel bystanders, causes a serious injury to someone -- let's say he knocks down a parking inspector with a powerful left hook and as a result the inspector falls into a coma. The police and the courts, we are asked to suppose, are aware that the offender will be lynched if he is acquitted on grounds of incapacity, and so they conspire to deprive him of his legal rights and send him to prison for assault. The problem here, according to van Inwagen, is that if we were to adopt the utilitarian reading of `blame' encapsulated in (6), we would be permitted, if not obliged, to treat the offender in this paternalistic and harsh way, given that such treatment produces the best overall consequences.Foot note 26
Hypothetical cases such as this are well-known in the literature on utilitarian accounts of punishment. H.J. McCloskey, in particular, made much use of these cases in an attempt to show that the utilitarian is committed to endorsing unjust forms of punishment.Foot note 27 But various problems have been identified with McCloskey's examples, and these same problems affect van Inwagen's hypothetical scenario. I will briefly mention only three of these problems here.Foot note 28
Firstly, as many others have pointed out, the examples employed by non-utilitarians are often fanciful or bizarre, with the result that the ensuing debate is usually conducted (as Smilansky puts it) «in a somewhat unreal atmosphere».Foot note 29 Van Inwagen's example also has this `unreal' character. For one thing, it is strange that people would want to lynch the offender after having goaded him into committing the offence; moreover, we would expect the police to be able to protect the offender were he to be judged `not guilty'. It seems, therefore, that scenarios such as those imagined by van Inwagen are unlikely to arise in the real worldFoot note 30; and insofar as utilitarianism is put forward as an empirical (or contingently true) theory, it cannot be refuted merely by devising logically possible counterexamples.Foot note 31
A second and more telling criticism of McCloskey's examples, that is also applicable to van Inwagen's example, is that if such cases are thought to depict `real life' situations, it is very likely that the overall outcomes, in terms of costs and benefits, would be quite different than what is envisaged by the originators of these examples. Consider, for example, Sprigge's comments in response to an example of McCloskey's that closely parallels van Inwagen's example:
One line of objection to this conclusion [that the punishment of an innocent person is mandated from a utilitarian point of view] appeals to the likelihood that the facts will become known. I may urge parenthetically that in the real world such a likelihood is likely (surely) to be pretty strong. The utilitarian may then insist on a variety of evils which would result from its becoming known, such as a loss of confidence in the impartiality and fairness of the legal system, of a belief that lawful behaviour pays, etc.Foot note 32
Precisely the same point can be made with respect to van Inwagen's scenario, where the suspect behaviour of the police and the courts is likely to be exposed, sooner or later. This view is further bolstered by the fact that, if on one occasion we are prepared in the name of some greater good to tell lies which will cause great harm to someone or even ruin their life, then we risk blunting our moral sensibilities in such a way that we will have few qualms about using similar methods when far less justification for doing so is available. This, in turn, will set in motion a train of behaviour amongst the legal authorities that will almost certainly become public knowledge, thus undermining further the community's system of justice.
Thirdly, one may adopt a more sophisticated form of utilitarianism that is not vulnerable to the kinds of counterexamples put forward by the likes of McCloskey and van Inwagen. A `preference utilitarian', for example, would give great weight to a person's preferences -- including their preference to have their autonomy respected -- when judging the moral value of an action.Foot note 33 A utilitarian of this sort could therefore argue that respect for autonomy is one of the most important factors that goes into promoting `the general welfare'. In other words, to maximize the general welfare we would normally need to maximize the degree of autonomy had by individuals. But then, if to do the right thing usually requires maximizing the general welfare by maximizing autonomy, we could permissibly override a person's autonomy only in very exceptional circumstances -- specifically, when no other avenue is available for protecting the welfare of that individual or other members of the community. But clearly such circumstances do not hold in van Inwagen's hypothetical case, unless we are given good reason to think that the offender cannot be protected by the police if he were judged `not guilty' by the courts.Foot note 34
Another common response to the instrumentalist rationale for moral responsibility is sparked by the feeling that such a rationale runs deeply counter to our ordinary moral practices. We do not ordinarily think that ascriptions of moral responsibility and praise or blame merely have a pragmatic role of influencing behaviour. Indeed, we often praise and blame people without intending at all thereby to modify their future behaviour. Jonathan Bennett voices this objection well in relation to an instrumentalist position developed by Moritz Schlick:
The Schlickian description of what accountability is -- or of what the concept is for -- strikes everyone as incomplete and strikes most people as wrong. The latter will say that although a distinction based on the utility of a certain sort of therapy or behaviour-control might coincide with accountability/non-accountability, it cannot give the latter's essence, and that the Schlickian account of what the line is for does nothing like justice to the real nature of our praise- and blame-related responses. When we express indignation for someone's cruelty, or admiration for his unselfishness, we usually are not engaged in any sort of therapy: blame-related responses all involve something like hostility towards the subject; whereas a moral-pressure therapist, though he may have to feign ill-feeling for therapeutic purposes, can in fact be in a perfectly sunlit frame of mind. And -- to move briefly to the `welcome' side of the fence -- one may apply moral pressures to encourage a welcomed kind of behaviour while remaining in an ice-cold frame of mind, with no feelings of gratitude, admiration or the like.Foot note 35
It may be thought, therefore, that the instrumentalist rationale for moral responsibility is inadequate insofar as it rules out the following activities as unfitting or irrational:
(a) expressing feelings such as indignation, disgust, and contempt toward others on account of their misbehaviour;
(b) expressing feelings of admiration or respect for those who do what they ought to or go beyond the call of duty in performing superogatory acts;
(c) making people suffer for their past deeds;
(d) rewarding people for special achievements (and not merely for positive reinforcement);
(e) feeling remorseful about one's own past deeds and not excusing oneself; and
(f) taking pride in having done what one thinks one ought to have done.Foot note 36
Despite the initial appeal of such objections, there are various ways of accommodating the above facets of human life within a hard determinist worldview.
Let's look first at (e), feelings of remorse and guilt.Foot note 37 It is plausible to think that if someone were deprived of such feelings, he would be incapable of mending any relationships with people that he has hurt. In addition, he would lack any motivation, after having done something wrong, to restore his own moral integrity and thus to develop morally. And so hard determinism, insofar as it prevents one from feeling blameworthy and hence guilty, undermines our ability to form good interpersonal relationships and to mature as moral beings.
This may seem plausible, but in the final analysis it is mistaken. As many hard determinists have insisted, if you perform a morally wrong act and yet you reject the claim that you are blameworthy (perhaps because you subscribe to hard determinism), you may nevertheless feel profound sorrow and regret that you were the agent of wrongdoing. And such feelings of intense sadness and regret are just what an attitude of remorse ordinarily amounts to. Pereboom in this context quotes an insightful passage from Hilary Bok, who employs an analogy with heartbreak to capture the idea that feelings of guilt may have little to do with feelings of blameworthiness:
The relation between the recognition that one has done something wrong and the guilt one suffers as a result...is like the relation between the recognition that one's relationship with someone one truly loves has collapsed and the pain of heartbreak. Heartbreak is not a pain one inflicts on oneself as a punishment for the loss of love; it is not something we undergo because we deserve it... Similarly,...the recognition that one has done something wrong causes pain. But this pain is not a form of suffering that we inflict on ourselves as a punishment but an entirely appropriate response to the recognition of what we have done.Foot note 38
But as Pereboom points out, even if guilt cannot be analysed solely in terms of feelings of sadness and regret, these feelings are likely to generate the same response as the experience of guilt: e.g., a resolution not to perform the immoral action again, an attempt to make amends by seeking to alleviate the suffering caused to other people, offering expressions of remorse towards those that have been wronged, etc.Foot note 39 Thus, the phenomena of guilt and remorse do not provide us with any reason to think that hard determinism threatens interpersonal relationships and personal moral development.Foot note 40
Similar things can be said about the other items on the above list of (a)-(f). Consider, for example, the attitudes of indignation, disgust, and contempt as mentioned in (a). These are what P.F. Strawson has called `participant reactive attitudes', that is, attitudes to which people are subject by virtue of engaging in ordinary interpersonal relationships. The attitudes are reactive in the sense that they are «natural human reactions to the good or ill will or indifference of others towards us, as displayed in their attitudes and actions.»Foot note 41 Other reactive attitudes, beyond those given in (a), include gratitude, moral resentment, forgiveness, and love.
Some of these reactive attitudes, particularly love, are crucial to forming good and fulfilling relationships, but they are not clearly threatened by hard determinism. On the other hand, a hard determinist may have to concede that some reactive attitudes -- such as indignation and gratitude -- cannot be accommodated in a fully deterministic world. But as Pereboom points out, the attitudes that are incompatible with hard determinism typically fall into one of two categories: (i) they are not essential (or are even detrimental) to good relationships -- resentment and contempt are cases in point; or else (ii) they have analogues or substitutes which lack any presupposition of moral responsibility while offering whatever is required for sustaining interpersonal relationshipsFoot note 42 (for example, happiness and joy could play the same role as gratitude without involving any belief that some person is morally responsible or praiseworthy for some beneficial act that has brought about one's contented stateFoot note 43).
As for (c), making wrongdoers suffer for their past deeds, the hard determinist should not find it difficult to develop a reasonable position on managing and punishing criminal behaviour. A hard determinist, however, will not be able to justify punishment in retributive terms. The retributivist position sanctions the punishment of a wrongdoer on the grounds of desert: the wrongdoer deserves to have something bad happen to him -- e.g., pain, deprivation, death -- simply because he has done wrong. But hard determinism is incompatible with the notion of moral desert and must therefore renounce the retributivist justification for punishment. There are, nonetheless, a number of other options available to the hard determinist. For example, punishment might be endorsed by the hard determinist as a way of morally educating the wrongdoer or, alternatively, as a way of deterring the criminal and other prospective criminals from committing crimes. Even if these justifications for punishment are morally objectionable, they are not excluded by hard determinism. And even if all punitive measures against criminals were ruled out by hard determinism, there are a range of non-punitive or rehabilitative measures available, and these (as Pereboom indicates) may be more preferable and humane in any case.Foot note 44
It is not obvious, therefore, that hard determinism imperils the emotions and attitudes that underpin our social and moral lives, or that it jeopardizes our ability to respond to human evil in various punitive and non-punitive ways.Foot note 45 In fact, far from constituting a threat to the flourishing of human life, hard determinism might enable us to achieve a more fulfilling, meaningful, and moral life.
Compatibilists, going all the way back to the Stoics, have often promoted determinism on account of its perceived ethical advantages. Pereboom adopts a similar position, though from a hard incompatibilist perspective: «If we did in fact relinquish our presumption of free will and moral responsibility,» he writes, «then, perhaps surprisingly, our lives might well be better for it.»Foot note 46 Part of Pereboom's case in support of this view is predicated on the idea that hard determinism compels us to abandon expressions of moral anger, that is to say, the kind of anger that is directed toward someone who is represented (accurately or inaccurately) as having done wrong. Moral anger and closely related attitudes (such as indignation and contempt) are normally justified by way of the claim that the wrongdoer deserves blame in virtue of having committed some moral offense. This way of legitimizing moral anger, however, is clearly unavailable to the hard determinist. To be sure, the hard determinist may find a purely instrumentalist rationale for expressions of moral anger, while holding at the same time that the belief in moral desert that such anger presupposes is in fact false (as psychologists often tell us, false beliefs can sometimes be quite useful). Another option, however, is to jettison moral anger altogether, and this is the option Pereboom canvasses. This may provide substantial benefits, as we would be released from the harmful effects of moral anger -- in particular, the physical and emotional pain it causes, and the damage it wreaks on our most cherished relationships -- and this, in turn, would provide us with an increased sense of equanimity when we are wronged or treated unjustly.
But wouldn't the loss of moral anger engender an unwelcome degree of passivity? Moral anger motivates us to fight injustice and to address problems in our relationships with others. Without moral anger, therefore, there is a danger that we would make ourselves vulnerable to all manner of injustice, oppression, and abuse. Pereboom suggests a way of responding to this difficulty:
When the assumption that wrongdoers are blameworthy is withdrawn for hard incompatibilist reasons, the conviction that they have in fact done wrong could legitimately survive. Not implausibly, such a moral conviction could lead to a firm resolve to resist oppression, injustice, and abuse. As a result, hard incompatibilism might allow for the benefits that moral anger may also produce, while avoiding its destructive consequences.Foot note 47
Perhaps, then, one's moral convictions (for example, one's conviction that something morally wrong has happened) in conjunction with the attendant feelings (such as feelings of hurt or alarm, particularly when one is personally affected by the morally wrong action in question) would be sufficient to motivate one to act to put things right. How plausible this is will depend in part on whether such convictions can be rationally upheld in a hard determinist world, an issue to be addressed in the ensuing Sections.
Saul Smilansky also highlights the ethical advantages of hard determinism. On his view, «the acceptance of hard determinism appears to create the possibility of a `purer' morality than traditional free will based morality.»Foot note 48 Smilansky explains that traditional or commonsense morality has a dual nature: (i) a `substantive' part, according to which certain things morally ought to be done (or not done), and (ii) an `accountancy' part, according to which people ought to get positive reactions such as praise when they do things that morally ought to be done (and negative reactions such as blame when they do things that morally ought not to be done). If we accept hard determinism, we will of course have to give up the accountancy part of morality. But this, Smilansky writes, would be liberating and not destructive, as it would present us with a `purer' morality. For given that the hard determinist agent chooses to do the right thing for its own sake, without any hope for praise or any fear of blame, he would be
...in one very curious sense a `purer' moral agent than someone who is concerned with both parts of morality. For the hard determinist agent of the sort we are considering is not concerned with himself at all, he is solely focused on determining what he ought to do. By contrast, the `non-hard-determinist' moral agent is inherently concerned with himself, with what action would make him praiseworthy or blameworthy in the eyes of others or of himself... The hard determinist agent we are considering enquires after the right thing to do and goes on to do it despite realizing that he is not praiseworthy for doing so and would not be blameworthy were he to act differently... If you like, he does not care about saving his moral soul, but about morality alone.Foot note 49
This I find a very attractive view. Pereboom, on the other hand, finds it unconvincing. He notes, first, that «one might doubt Smilansky's claim that for an agent who believes in moral responsibility, a concern with one's blameworthiness or praiseworthiness is always present.»Foot note 50 To be fair to Smilansky, however, following the above-quoted passage he goes on to note that someone who believes in moral responsibility need not be egoistic: «When one wants to be a decent human being and act like a creature worthy of the attribution of moral worth it would normally be misleading to call one egoistic.»Foot note 51 Nevertheless, Smilansky continues to maintain that, with respect to the believer in moral responsibility, «a concern with the self's moral stature is...at least tacitly present.»Foot note 52 I think Smilansky has overstated the mark, as it would have sufficed for his purposes to make the weaker but more plausible claim that a person who believes they are morally responsible is much more likely to exhibit the relevant kind of self-concern than their hard determinist counterpart. This view is entirely compatible with Pereboom's claim that «it seems possible that one believes in moral responsibility, but that this belief plays no role in one's moral decision-making.»Foot note 53
Pereboom, however, voices a second criticism of Smilansky's view:
For the hard incompatibilist agent, an analog to concern for one's praiseworthiness or blameworthiness seems as likely to be present as does this concern itself for the agent who believes in moral responsibility. This analog can also be characterized as a concern that one be a morally worthy person.Foot note 54
Pereboom goes on to add that, in a hard determinist world, an agent's moral worth would be determined by, for example, the degree to which they regulate their behaviour in accordance with moral reasons, the degree to which they are disposed to examine their past behaviour from the moral point of view, and their willingness to change their behavior when tendencies to immorality are recognized. This kind of moral worth, on Pereboom's view, would be just as much the concern of the hard determinist agent as would be the analogous kind of moral worth for the believer in moral responsibility.
Pereboom may be correct in holding that the hard determinist need only sever the connection between an agent's moral worth and their degree of praiseworthiness, rather than abandoning the notion of an agent's moral worth altogether. What Pereboom overlooks, however, is Smilansky's point that, without the accountancy side of morality, «there is one less major `morally impure' source of motivation for acting as one should.»Foot note 55 In other words, the hard determinist agent, although concerned with their moral worth, will be motivated by, for example, the desire to become a better person or to promote the good. The believer in responsibility, on the other hand, is more likely to be motivated by more selfish reasons: viz., their wish to be rewarded or praised. Both the hard determinist and the non-hard determinist may display an equal degree of concern with their respective self's moral stature, but what is motivating them in each case is markedly different.
Smilansky, however, controversially assumes that the hard determinist can retain the `substantive' aspect of morality, which includes such notions as moral obligation, moral rightness and wrongness, and moral goodness. Pereboom, as we saw earlier, makes a similar assumption when promoting the renunciation of moral anger as overall beneficial. But can the assumption that the hard determinist may reasonably retain a conviction in obligation, rightness, and goodness be justified? It is to this problem that I now turn.
III. Moral Obligation, and Moral Wrongness and Rightness
What happens to `ought' judgements, and to our judgements that some actions are morally wrong while others are morally right, in a hard determinist world? It is commonly thought that the hard determinist must dispense with such judgements. Arguments in support of this view vary, but here is one inspired by Ishtiyaque Haji:Foot note 56
(7) Assume that action A is something that person S does in fact perform.
(8) K: Person S has a moral obligation to perform [not to perform] action A only if S can perform [refrain from performing] A.
Premise (8), or principle K (after its best-known exponent, Kant), is the `ought implies can' principle, the principle that one ought morally to do A only if one can do A. Although the principle is intuitively plausible, it has become very contentious in virtue of the varying interpretations of `can' that have been offered.Foot note 57 Haji has helpfully outlined three versions of this principle, each encapsulating a difference sense of `can':Foot note 58
Agent S ought to do A only if S can do A, where this entails that:
(i) S has the opportunity to do A -- i.e., nothing prevents or would prevent S from successfully exercising the relevant abilities (whatever they are) to do A; and
(ii) S is physically able to do A.
The first condition, relating to opportunity, expresses the idea that a person can do something only if they are `relevantly situated' or have the cooperation of the environment. I can save a drowning child, for example, only if I can perceive that there is a child who is struggling to keep afloat, or only if the waters are not infested with sharks which would eat the child before I even jump into the water. The second condition, relating to physical ability, is regularly associated with uses of `can' and `cannot', as when we say that James cannot read the traffic sign because he is blind.
The problem with conditions (i) and (ii) is that they fail to capture some important uses of `can' that are relevant to judgements of moral obligation. If, for example, you succeed in doing A but your accomplishment of A was strictly beyond your control -- because your accomplishment was, say, a matter of sheer luck or predetermined by God -- we would be reluctant to say that you could have done A in an obligation-conferring sense (that is, in a sense in which your ability to do A confers on you either the obligation to do A or the obligation not to do A). We might therefore opt for a stronger `ought implies can' principle:
Agent S ought to do A only if S can do A, where this entail that:
(i) S has the opportunity to do A -- i.e., nothing prevents or would prevent S from successfully exercising the relevant abilities (whatever they are) to do A;
(ii) S is physically able to do A; and
(iii) A's accomplishment is not `strictly out of S's control'.
Some have thought, however, that this version of the principle is also inadequate, since there is no provision made for various forms of knowledge -- specifically, `know-how' and `knowledge that' -- which undergird some morally relevant uses of `can'. We say, for example, that James ought to save the drowning child, given that he can do so -- and by this we mean, in part, that James knows how to swim and is not ignorant of the fact that the child is on the verge of drowning. We might therefore be led to an even stronger version of the `ought implies can' principle:
Agent S ought to do A only if S can do A, where this entails that:
(i) S has the opportunity to do A -- i.e., nothing prevents or would prevent S from successfully exercising the relevant abilities (whatever they are) to do A;
(ii) S is physically able to do A;
(iii) A's accomplishment is not `strictly out of S's control';
(iv) S has the relevant `know-how' to do A; and
(v) S is not ignorant of any germane facts (the `know-that' requirement).
Thankfully, for present purposes I do not need to decide which of these versions best accounts for the sense of `can' in `ought implies can' judgements. I will proceed, therefore, with the next step of the argument, which describes a standard principle of moral obligation:
(9) OW: Person S has a moral obligation to perform [not to perform] action A if and only if it is morally wrong for S not to perform [to perform] A.Foot note 59
Premises (8) and (9) entail:
(10) WAP: It is morally wrong for S to perform [not to perform] A only if S can refrain from performing [perform] A.
The plausibility of WAP may be highlighted by means of familiar Frankfurt-type cases. Suppose, for example, that I decide of my own accord to kill the President. Unbeknownst to me, however, Mad Scientist has the power to control my mind, so that if I were to hesitate to carry out my intentions, he would step in and stimulate parts of my brain in order to cause me to commit the crime. As it turns out, I carry out my plan without any hesitation, thus requiring no intervention from Mad Scientist. Nevertheless, I could not have acted otherwise; for instance, I could not have dropped the gun at the last minute, rather than pull the trigger. But given that I could not have refrained from killing the President even if I had wanted to, it seems that what I did was not wrong. Underlying this view, as Haji explains, is the idea that
Just as it is eminently reasonable to suppose that responsibility requires control -- no one, for example, can be morally responsible for an action if one does not have «responsibility-grounding control» over the action -- so it is reasonable to suppose that no one can perform an action that is morally right, or wrong, or obligatory unless one has appropriate «deontic-grounding control» over it.Foot note 60
If, therefore, I cannot refrain from killing the President, then I cannot exercise the requisite kind of control over my behaviour, and for this reason my behaviour cannot be deemed morally wrong.
The next premise draws out an obvious implication of the doctrine of hard determinism:
(11) D: If hard determinism were true, then no-one can do otherwise, that is, no-one can refrain from doing what they in fact do.Foot note 61
The remaining steps of the argument then proceed as follows:
(12) If hard determinism were true, then it is not the case that it is morally wrong for S to perform A -- that is to say, S's actions can never be judged to be morally wrong. (From 10 and 11)
(13) If hard determinism were true, then it is not the case that S has a moral obligation not to perform A -- that is to say, S is never morally obligated to do otherwise than what S in fact does.Foot note 62 (From 9 and 12)
What this argument indicates, assuming it to be sound, is that our commonsense system of morality -- and in particular our commonly held beliefs that at least some of our actions are morally wrong and that in at least some cases we ought to have acted otherwise -- are undermined by hard determinism. This does not mean that all our judgements regarding moral obligation and moral rightness or wrongness must be jettisoned. But it does mean that much of what we hold dear and common in the domain of morality is mere illusion.
I will now consider the attempt made by Pereboom to resist the conclusions of the above argument (or at least conclusion (12), as he is quite content to let moral obligation goFoot note 63). But before doing so, it may be helpful to briefly consider one possible line of response that those who are fond of compatibilism may be inclined to make. It may be held that a compatibilist reading of `S could have done otherwise' would remove any reason to think that determinism (whether hard or soft) is incompatible with obligation and rightness or wrongness. According to compatibilists, to say that S could have done otherwise is just to assert a hypothetical of the form, `S would have done otherwise if S had desired, chosen, or willed to do otherwise'. But if `can' is understood along these lines, then -- contra (11) -- people can do otherwise even in a hard determinist world.
Haji, however, demurs, and illustrates his objection with a Frankfurt-type case similar to that described earlier.Foot note 64 In a determinist world, he notes, we would all be in the same boat as the killer in the Frankfurt-type scenario because, like him, (i) we would not be able to do otherwise than what we in fact will to do, and (ii) we would not do otherwise even if we had so willed. Thus, even in the compatibilist's sense of `could have done otherwise', it remains the case that in a determinist world we could not act otherwise. But this does not seem right, for there is no reason to think that agents in a (hard or soft) determinist world would always find themselves in Frankfurt-style scenarios. And if that is so, then it follows that, even though in a determinist world we could not do anything other than what we in fact do, it remains possible that had we willed to do something other than what we in fact did, we would have done something other than what we in fact did.Foot note 65 In any case, I will let this criticism of premise (11) pass, and focus instead on the other steps of the argument.
Pereboom's general line of response to Haji's argument presented in (7)-(13) above is that the central premises of the argument -- viz., principles K and OW -- cannot plausibly be thought of as having «a justificatory status so strong that it immunizes them against disconfirming pressures from their unintuitive consequences.»Foot note 66 He goes on to explain that, «If the components of the [moral] theory derived from these principles [viz., K and OW] conform to our intuitions, that would provide theoretical support for them. But if such derived components do not conform to our intuitions, that would to some extent disconfirm these principles.»Foot note 67 To be sure, it is not (K & OW) alone that produces the counterintuive results that none of our actions are morally wrong and that we never ought to have acted otherwise. Rather, it is (D & K & OW) that produces these results. So, how is Pereboom to be understood here? As will be seen shortly, Pereboom's view is that K and OW are not as epistemically secure as might be thought at first glance, and that once we consider matters more deeply we will have no more reason to reject the claim that some of our actions are morally wrong as to accept this claim. Pereboom offers three arguments in support of this view.
His first argument begins with the following intuitively true principle, connecting moral goodness with moral rightness:
(14) GR: Sometimes, actions that bring about the greatest good overall in worlds accessible to S are right for S.
Note, first, that Haji's argument has nothing explicitly to say about moral goodness (and moral badness). Thus, as far as the considerations advanced by Haji go, we have no reason to think determinism to be incompatible with moral goodness, though we are given some reason to think that determinism rules out judgements of moral wrongness and rightness (or at least a large class of our ordinary judgements regarding moral wrongness and rightness). This, however, commits Haji to an invidious demarcation between judgements about moral goodness and judgements about moral rightness. In other words, if assume the truth of D for the sake of argument, and if we follow Haji in accepting (K & OW), then we must also accept the following unintuitive result:
(15) GR is false.Foot note 68
But, Pereboom adds, it is far from clear that Haji's position, as laid out in (I) below,
(I) Accept: D & K & OW & goodness & blameworthiness
and therefore reject: obligation & rightness/wrongness & GR
is at all superior to a view (like Pereboom's) that bids us to
(II) Accept: D & K & rightness/wrongness & goodness & GR
and therefore reject: blameworthiness & obligation & OW.
Each position has uninituitive results, but it is not obvious that position (I) is in better shape than (II). And so, at the very least, we are faced with a kind of détente. This problem, of course, would not arise if you thought that, in a hard deterministic world, there can be no moral goodness, and not just no moral rightness or wrongness. But as I hope to show in the following Section, judgements of moral goodness are not undermined by hard determinism. This argument of Pereboom's, then, has some degree of force.Foot note 69
Pereboom's second argument is structurally similar to his first argument, but targets Haji's view that blameworthiness survives determinism while wrongness does not. According to Haji, the principle that an agent is blameworthy for performing an action only if that action is in fact wrong, despite its intuitive plausibility, is seriously mistaken. He offers a number of considerations in support of this view, including hypothetical cases such as the following (which he dubs `Deadly's Defeat'):
In Deadly's Defeat, doing the best she can to murder her patient, Dr. Deadly does what credible evidence to which she has access indicates will kill the patient -- she injects the patient with medicine C. But...happily for the patient, there has been an innocent error in diagnosis. Contrary to what Deadly believes, her patient is suffering from a disease that can be cured only by taking C. Giving C results in the lucky patient's full recovery. Although it is arguably obligatory for Deadly to inject her patient with C, it appears that Deadly is to blame for the injection of this drug.Foot note 70
Cases such as this, according to Haji, strongly indicate that moral blameworthiness should be kept apart from objective wrongness and objective obligatoriness: Dr Deadly's behaviour may be open to blame, but he has done (albeit inadvertedly) what he ought to have done and what was right for him to do. Haji therefore endorses a `subjective view' of blameworthiness (and praiseworthiness), which roughly states that an agent is blameworthy for performing an action only under the condition that she has the belief that it is wrong for her to perform that action and this belief plays an appropriate role in the performance of that action -- the belief in question, however, need not be epistemically justified, let alone true.Foot note 71 This allows Haji to hold that, in a deterministic world, people can be blameworthy for what they do, even though their actions can never be morally wrong or obligatory. Pereboom, however, complains that Haji's view has the paradoxical result that when agents are blameworthy in a deterministic world, it is never because they have done something wrong. This, according to Pereboom, conflicts with another intuitively plausible moral principle, viz.:
(16) BW: Sometimes, when S is blameworthy for performing A, it was morally wrong for S to perform A.
And so, as before, if we assume the truth of D for the sake of argument, and if we follow Haji in accepting (K & OW), then we must also accept the following unintuitive result:
(17) BW is false.
And, once again, there is little reason to think that Haji's position,
(I*) Accept: D & K & OW & goodness & blameworthiness
and therefore reject: obligation & rightness/wrongness & BW
is preferable to the view that we should
(II*) Accept: D & K & rightness/wrongness & goodness & BW
and therefore reject: blameworthiness & obligation & OW.Foot note 72
The first thing to note is that Pereboom's argument is only aimed at someone, like Haji, who thinks that in a deterministic world there can be blameworthiness and praiseworthiness (of the non-consequentialist, merit-based variety). Some support for Haji's view is provided by Frankfurt-type cases where the murderer appears to be morally responsible and deserving of blame, despite lacking the ability to have done otherwise. As noted in Section II above, Pereboom (correctly, I take it) disagrees with this assessment, for in his view no-one in a deterministic world is ever deserving of praise or blame. But what is important to note in the present context is that Haji's view that blameworthiness may survive in a deterministic world is not essential to the argument he develops, in (7)-(13) above, in support of the claim that determinism undermines some central concepts of deontic morality (in particular, moral obligation and moral wrongness). Pereboom, therefore, in attacking Haji's views on blameworthiness, fails to engage with the argument of (7)-(13).
In any event, there are problems with both Haji's view of blameworthiness and Pereboom's response to this view. To begin with Pereboom's reply, there is no reason why Haji must reject BW. On Haji's view, when a person S is blameworthy for performing some action A, it is never because S has done something that is objectively wrong; rather, it is (partly) because S has done something that is subjectively wrong (or wrong from S's perspective). But Haji can still agree that, in some cases, a person S is blameworthy for performing A and it was objectively wrong for S to perform A. Haji, therefore, can endorse BW (at least if the `when' in BW is not read as `because').
However, Haji's subjectivist account of blameworthiness is also problematic, as it seems to be give insufficient weight to culpable ignorance. Consider, for example, a case where a Nazi officer kills a Jewish boy in front of its parents in the belief that it is right and obligatory to kill Jewish children. The Nazi's action would ordinarily be considered blameworthy, at least partly because he is culpable in forming the belief in question. In such cases, ignorance is no excuse: the offender ought to have known better and is therefore open to blame. But if Haji's account were correct, the Nazi officer could not be blameworthy since he does not believe that what he is doing is wrong. This, clearly, is a disappointing result of Haji's account.Foot note 73
Pereboom's third argument is intended to break the deadlock between positions (I) and (II), and their cognates, in favour of (II) and (II*). Pereboom's case here is built upon the premise that there is good reason to reject Haji's `standard principle of moral obligation':
OW: Person S has a moral obligation to perform [not to perform] action A if and only if it is morally wrong for S not to perform [to perform] A.
OW, being a biconditional, consists of two distinct claims:
(OW1) If S has a moral obligation to perform [not to perform] action A, then it is morally wrong for S not to perform [to perform] A, and
(OW2) If it is morally wrong for S not to perform [to perform] A, then S has a moral obligation to perform [not to perform] A.
Pereboom focus on the bracketed version of OW2, which reads as follows:
(OW2*) If it is morally wrong for S to perform A, then S has a moral obligation not to perform A.
To produce a counterexample to OW2*, one would need to devise a scenario in which it is morally wrong for you to do A and yet you do not have a moral obligation not to do A. Pereboom attempts to do just this by way of the following purported counterexample:
Suppose you say to an animal-abuser, «You ought not to abuse that animal,» but then you find out that he has a psychological condition (which he could have done nothing to prevent) that makes animal-abusing irresistible for him, so that he cannot help but abuse the animal. From my point of view, there is an appreciably strong pull to admitting that the «ought» judgment was false, but there is relatively little to denying that abusing the animal is morally wrong for him.Foot note 74
I am not convinced, as I doubt that moral wrongness can be entirely divorced from moral obligation. Indeed, it seems strange to say that, while having an obligation do something, A, presupposes having control over A-ing, its being wrong for one to do A does not. As Haji rightly asks, «Why the asymmetry in metaphysical presuppositions of control? After all, wrongness and obligatoriness are normative appraisals of the same family; they are deontic normative statuses and thus it would seem that the control-relevant presuppositions of the one should also be, barring cogent explanation to the contrary, those of the other.»Foot note 75 It seems unreasonable, then, to deem an action `morally wrong' while denying that one ought not engage in such behaviour. And so if one has no moral obligations, there is nothing morally wrong that one can do. That is why describing the animal abuser's actions as `morally wrong' would be just as mistaken as describing a shark attack on a human being as `morally wrong'.
Summing up the results of this Section, it appears that only Pereboom's first response to the Haji-inspired argument presented in (7)-(13) above has succeeded in providing us with some reason to doubt the claim that if hard determinism were true, none of our actions would be morally wrong. At best, therefore, we have just as much reason to believe that ascriptions of rightness and wrongness survive hard determinism as we have to reject this view. This is not an entirely positive result for the hard determinist, especially since no attempt has been made to respond to the above argument's further claim that hard determinism is not compatible with moral obligation.
There are, however, various ways out for the hard determinist that seem to have gone unnoticed by both Pereboom and Haji. One such way out involves extending the consequentialist account of praise and blame, as defended in Section II, to our practice of making `ought' judgements and judgements of rightness and wrongness. More specifically, given a consequentialist reading of moral wrongness, WAP (i.e., premise (10)) would turn out to be false. For whether an action counts as morally wrong will depend on the (foreseeable) consequences of that action, perhaps in conjunction with some other conditions (such as that the action was performed intentionally). But the requirement, as expressed in WAP, that the agent be able to refrain from performing the action in question will play no part in the consequentialist story of moral wrongness. Alternatively, one may have grounds for rejecting WAP that do not rely on any particular theory in normative ethics -- one could, for example, reject WAP on the grounds that the `ought implies can' principle presupposed by WAP is false. But then a range of options become available to the hard determinist. Even a deontological account of moral rightness, wrongness and obligation can be endorsed by the hard determinist, for without WAP no threat will be posed to these moral categories by the absence of alternate possibilities. With the demise of WAP, however, the argument presented above in steps (7) to (13) collapses as well.
IV. Moral Goodness
The last moral category I will consider is moral goodness. Can moral goodness be accommodated in a hard determinist world? Pereboom thinks it can. He holds that even if, under hard determinism, `ought' judgements are never true, it remains the case that judgements such as `It is morally good for person S to do A' and `It is morally bad for S to do B' can be true. He offers the following illustrations in support of this view:
Thus, for example, even if one is causally determined to refrain from giving to charity, and even if it is therefore false that one ought to give to charity, it still might be good to do so. Cheating on one's taxes might be a bad thing to do, even if one's act is causally determined, and thus, even if it is false that one ought not to do so.Foot note 76
At least one question this immediately raises is: What sense of `good' is operative here? One obvious option, in the light of what has been said earlier, is to give `goodness' a consequentialist reading. On this view, an action is considered morally good insofar as it has certain consequences (e.g., the greatest possible increase of pleasure over pain). But the hard determinist, as already mentioned, has a number of options available within normative ethical theory, even including a deontological understanding of moral goodness.
A slight problem, however, lurks here. Let's suppose that we adopt a purely deontological moral framework. Let's further suppose that, as Haji's argument in the previous Section suggests, we are not entitled to think that moral obligation and moral rightness or wrongness (characterised in non-consequentialist terms) can be made to fit into a hard determinist world. Could we then say of at least some actions in such a world that they are morally permissible for some agent? In other words, if an action is not right, wrong, or obligatory, does it follow that it is permissible?Foot note 77 Haji considers this question and answers that it would be more plausible to say that actions that are not right, wrong, or obligatory are `amoral' for a person, rather than permissible for a person. His reasoning in support of this view is that, if some of our actions are open to moral appraisal although they could never be wrong or obligatory, then each of our actions would either be amoral or morally permissible. But then if I were to kill someone in cold blood, my action would be, if not amoral, then morally permissible for me, «a result that is hard to swallow».Foot note 78
If Haji's reasoning were accepted, as I think it should, then we are faced with the following uncomfortable result: We can legitimately say of some action A that it is morally good, and yet we must deny that A is morally permissible. This sounds strange, however, only because we overlook the fact that judgements of permissibility are interchangeable with judgements of obligation (action A is not permissible for person S iff S ought not to do A; A is permissible for S iff it is not the case that S ought not to do A). That is why the collapse of our system of moral obligations also brings with it the end of our practice of judging actions to be permissible or impermissible. There should therefore be no hesitation in denying moral permissibility, given that we have already dispensed with moral obligation. But as pointed out by Pereboom, this need not lead us to reject moral goodness.
Given the truth of hard determinism, the ordinary folk picture of moral responsibility (and hence of praiseworthiness and blameworthiness), moral obligation, moral rightness or wrongness, and moral goodness -- since it presupposes a commitment to the notions of desert, the ability to do otherwise, and thus (compatibilist or libertarian) free will -- must be uprooted, or at least revised in significant ways. The nature and function of judgements involving these moral categories must therefore be thoroughly reconceived. But even if there is much in our commonsense system of moral beliefs that must be uprooted or seen anew, the hard determinist is not without the resources needed to rebuild our moral edifice. We have seen, for example, that one among many options available to the hard determinist is to (re-)interpret moral judgements in broadly consequentialist fashion, so that a judgement that, say, a particular action is morally right is to be analyzed in terms of the overall consequences that would most likely result if that action were performed.
There is no reason to think, however, that this reorientation in our moral outlook brought about by hard determinism would have disastrous practical consequences for our lives. Despite initial appearances, the assumption that we lack the kind of free will required for moral responsibility does not threaten the emotions and reactive attitudes that are crucial to our social life and moral development. Indeed, as some hard determinists have plausibly argued, the rejection of responsibility-entailing free will holds out the promise of a life that is morally deeper (because less self-centred) and more fulfilling (because less prone to destructive anger).Foot note 79
Department of Philosophy
School of Philosophy & Bioethics
Clayton, Victoria, 3800
Nick [dot] Trakakis [at] arts [dot] monash [dot] edu [dot] au
[Foot Note 1]
Peter van Inwagen, «Response to Slote,» Social Theory and Practice 16 (1990), p.394.
[Foot Note 2]
See Haji, «Moral Anchors and Control,» Canadian Journal of Philosophy 29 (1999), pp.175-203, and Deontic Morality and Control (Cambridge: Cambridge University Press, 2002), ch. 3. I should point out, however, that, on Haji's view, `deontic morality' is undermined not so much by the absence of free will but by the truth of determinism (I thank Ish Haji for clarifying this). See also Robert Kane, The Significance of Free Will (New York: Oxford University Press, 1996), ch. 6, where he discusses ten important goods that humans generally desire and which are often thought to be imperiled by the absence of libertarian free will.
[Foot Note 3]
For a list of some exponents and detractors of the idea that there can be no substantive morality without moral responsibility, see Bruce Waller, «Virtue Unrewarded: Morality Without Moral Responsibility,» Philosophia 31 (2004), pp.427-28, 442-43 (fn.5). Waller himself supports the view that «morality -- morality of almost any variety one favors -- can survive and flourish in the absence of moral responsibility» (p.428).
[Foot Note 4]
As is well known, Kant held that, despite the paucity of evidence in support of agent-causal libertarianism (which, in his view, is required for moral responsibility), we must postulate that we possess free will so as to uphold the validity of the moral law. Smilansky defends a similar position in a number of publications, but especially in Free Will and Illusion (Oxford: Clarendon Press, 2000).
[Foot Note 5]
Smilansky, Free Will and Illusion, p.170. Smilansky, I might add, does not think that hard determinism is the whole truth of the matter in regards to free will. Rather, he adopts the `dualist' view that both the hard determinist perspective and the compatibilist perspective capture part of the truth relating to free will, with neither perspective being in principle superior to the other (see ch. 6 of his Free Will and Illusion). Smilansky, however, holds that just as one cannot live well with the belief that there is no libertarian free will, so too one cannot live well with the dissonance created by acknowledging the validity of both compatibilism and hard determinism. Dualism, therefore, creates a further need for illusion regarding free will (Free Will and Illusion, pp.149-61).
[Foot Note 6]
For this reason, Smilansky stresses that his position is not to be equated with Hans Vaihinger's idea that, although we lack libertarian free will, we ought to behave as if we possessed such free will -- see Vaihinger, The Philosophy of `As If': A System of the Theoretical, Practical and Religious Fictions of Mankind, 2nd ed., trans. C.K. Ogden (London: Routledge & Kegan Paul, 1935), pp.42-48. As Smilansky states, «Vaihinger generally holds that awareness of the fictional character of an idea accompanies its use. My position is different -- lack of awareness (to varying degrees) of the illusory element is often the very condition of its productive existence» (Free Will and Illusion, p.147).
[Foot Note 7]
At least, this is what is suggested by comments made by Smilansky on pp.263-64 of Free Will and Illusion. The difficulty here, as Smilanksy has related to me in correspondence, concerns the matter of internalization: «I think that most human beings (including myself, even after all those years of dealing with the problem) have great difficulties in really internalizing the absence of libertarian free will, and what it implies about things like their moral worth» (personal correspondence, 10 April 2005).
[Foot Note 8]
A further unsavoury, if not paradoxical, aspect of Smilansky's `illusionism' is that, given the grave dangers involved in people becoming aware of the illusory nature of free will, it is important to keep the truth on this matter as hidden as possible. Smilansky, for example, writes that «it is important for her role as part of a community that the philosopher does not proclaim to everyone her findings about free will» (Free Will and Illusion, p.274). Even if this kind of deception or dishonesty can be justified (as Smilansky thinks it can, in pp.258-68 of Free Will and Illusion), one wonders why Smilansky publishes papers and books proclaiming the non-reality of libertarian free will, rather than making every effort to conceal this truth through publications that defend libertarianism with rhetorically effective (albeit invalid and unsound) arguments.
[Foot Note 9]
A principle of this sort has been defended by many philosophers of religion in the context of the problem of evil, where it is argued that it is morally permissible for God to allow a human being S to undergo intense, involuntary and underserved suffering only if the good secured by S's suffering is shared by S in a conscious way. See, for example, Eleonore Stump, «The Problem of Evil,» Faith and Philosophy 2 (1985), p.411; William Alston, «The Inductive Argument from Evil and the Human Cognitive Condition,» Philosophical Perspectives 5 (1991), p.48; and Marilyn Adams, Horrendous Evils and the Goodness of God (Melbourne: Melbourne University Press, 1999), pp.29-31. For some dissenting views, see Peter van Inwagen, «The Magnitude, Duration, and Distribution of Evil: A Theodicy,» Philosophical Topics 16 (1988), p.184, and C. Behan McCullagh, «Evil and the God of Love,» Sophia 31 (1992), p.55.
[Foot Note 10]
The second and third criticisms made above are indebted to Derk Pereboom, Living Without Free Will (Cambridge: Cambridge University Press, 2001), pp.198-99.
[Foot Note 11]
I will be relying in the main on Pereboom's defence of hard determinism in Living Without Free Will. For more succinct expressions of Pereboom's position, see his «Determinism Al Dente,» Noûs 29 (1995), pp.21-45, and «Living Without Free Will: The Case For Hard Incompatibilism,» in Robert Kane (ed.), The Oxford Handbook of Free Will (Oxford: Oxford University Press, 2002), ch. 21.
[Foot Note 12]
See William James, «The Dilemma of Determinism,» in The Will to Believe and Other Essays in Popular Philosophy (New York: Longmans, Green, and Co., 1917), p.149.
[Foot Note 13]
Pereboom also cites Joseph Priestley (1733-1804) as a hard determinist (see Pereboom, Living Without Free Will, p.129). However, it is clear from Priestley's major writings on determinism -- viz., The Doctrine of Philosophical Necessity Illustrated (1777), and his correspondence with Richard Price, published as A Free Discussion of the Doctrines of Materialism and Philosophical Necessity (1778) -- that he rejected the libertarian conception of free will (or what he calls `the doctrine of philosophical liberty'), but accepted a compatibilist understanding of free will. He notes, for example, that
I allow to man all the liberty or power that is possible in itself, and to which the ideas of mankind in general ever go, which is the power of doing whatever they will, or please, both with respect to the operations of their minds and the motions of their bodies, uncontroulled by any foreign principle or cause.
(The Doctrine of Philosophical Necessity Illustrated, Section I, in The Theological and Miscellaneous Works of Joseph Priestley, New York: Klaus Reprint, 1972, vol. 3, p.459, emphases in the original.)
Elsewhere, Priestley writes that
To me it seems sufficient that men be voluntary agents, or that motives, such as hopes and fears, can influence them in a certain and mechanical manner to make it in the highest degree right and wise in the Divine Being to lay such motives before them, and consequently to place them in a state of moral discipline, or a state in which rewards and punishments are distributed so as to correspond to certain characters and actions.
(A Free Discussion of the Doctrines of Materialism and Philosophical Necessity, Introduction, in The Theological and Miscellaneous Works of Joseph Priestley, New York: Klaus Reprint, 1972, vol. 4, p.11, emphases in the original.)
[Foot Note 14]
Pereboom concedes that Frankfurt-type cases succeed in showing that it is not the lack of alternate possibilities in a deterministic world that prevents us from holding agents in such a world morally responsible and blameworthy for their actions. See Harry G. Frankfurt, «Alternate Possibilities and Moral Responsibility,» Journal of Philosophy 66 (1969), pp.829-39. For Pereboom, what undermines moral responsibility in a deterministic world is not the lack of alternate possibilities, but the lack of causal histories that render agents ultimate sources of their actions.
[Foot Note 15]
See Pereboom, Living Without Free Will, pp.127-28.
[Foot Note 16]
See Pereboom, Living Without Free Will, pp.128-33, where Pereboom compares his position with the views of other contemporary hard determinists.
[Foot Note 17]
Cf. Robert Kane, «Introduction: The Contours of Contemporary Free Will Debates,» in Kane (ed.), The Oxford Handbook of Free Will, p.28, where both thesis (1) and thesis (2) are taken to be the kernel of the hard determinist position.
[Foot Note 18]
It may be noted that it is entirely consistent with the hard determinist view, as expressed in (1), (2) and (3) above, that humans do possess free will, albeit not free will of the sort required for moral responsibility. That is why Bruce Waller can argue in Freedom Without Responsibility (Philadelphia, PA: Temple University Press, 1990) that «our contemporary, nonmiraculous, naturalist (determinist) world system leaves no room for moral responsibility -- though it leaves quite adequate space for individual freedom» (p.4).
[Foot Note 19]
I take the labels `merit-based' and `consequentialist' from Andrew Eshleman, «Moral Responsibility,» §1, in The Stanford Encyclopedia of Philosophy (Fall 2004 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/fall2004/entries/moral-responsibility/>. The `consequentialist' view of moral responsibility will also be referred to as the `utilitarian' or `pragmatic' view, without thereby wishing to draw any distinction between these three labels.
[Foot Note 20]
See Pereboom, Living Without Free Will, pp.139-41.
[Foot Note 21]
J.J.C. Smart, «Free Will, Praise and Blame,» Mind 70 (1961), p.302.
[Foot Note 22]
As Joel Kupperman points out, «This is not to say that people, including philosophers, normally praise or blame as a device to modify behavior; to equate the function of an utterance with a speaker's intention would be like saying that people in agony normally scream in order to get help» (Character, New York: Oxford University Press, 1991, p.60, emphasis in the original).
To prevent any unnecessary misunderstandings, I stress that `blameworthiness' in this context simply consists in the appropriateness of outward expressions of blame by words, gestures, or actions -- there is no implication of someone's being morally worthy (or deserving) of being judged in a certain way on account of their behaviour.
[Foot Note 23]
Apart from Smart, other proponents of this pragmatic or consequentialist account of moral responsibility have included Henry Sidgwick, The Methods of Ethics, 7th edition (London: Macmillan, 1907), pp.71-72; Moritz Schlick, Problems of Ethics, trans. David Rynin (New York: Dover Publications, 1939), ch. 7; Richard B. Brandt, «A Utilitarian Theory of Excuses,» Philosophical Review 78 (1969), pp.337-61; Daniel Dennett, Elbow Room: The Varieties of Free Will Worth Wanting (Oxford: Clarendon Press, 1984), pp.163-65; Michael Slote, «Ethics Without Free Will», Social Theory and Practice 16 (1990), pp.369-75; and Joel J. Kupperman, Character, ch. 3.
In «Denying Moral Responsibility: The Difference It Makes,» Analysis 49 (1989), pp.44-47, Bruce Waller points out some interesting implications of such pragmatic accounts of praise and blame. Waller notes that the hard determinist who aligns his views with the findings of contemporary behavioural science (with its account of how behaviour is shaped by various contingencies and schedules of reinforcement) will respond quite differently to good and bad behaviour in comparison with a believer in (a non-utilitarian conception of) moral responsibility. In fact, such a hard determinist will dispense rewards and punishments in the opposite way to that in which someone who accepts moral responsibility would dispense rewards and punishments. More specifically, the behavioural hard determinist will offer more generous rewards (positive reinforcements) to the vicious individual (whenever that individual's actions resemble the desired good behaviour) than to an individual who has been conditioned to act virtuously, since greater positive reinforcement is required to reform the reprobate's behaviour than to sustain the virtuous individual's behaviour. By contrast, someone who rejects hard determinism would offer praise whenever it is deserved, and the virtuous (on this view) always deserve greater praise than the vicious. «Behavioural hard determinism,» writes Waller, «rejects moral desert claims, and the schedule of rewards (positive reinforcement) it proposes is fundamentally incompatible with those based on moral desert/responsibility» (p.46, emphasis in the original). Furthermore, «the hard determinist justification of positive and aversive conditioning will be the opposite of the justification given for reward/punishment by believers in moral responsibility» (p.46, emphasis in the original). Strangely, in Freedom Without Responsibility (pp.135-40), Waller offers these observations as an objection to the pragmatist conception of moral responsibility. I would have thought that Waller's behaviourist account fits quite nicely with the utilitarian view of moral responsibility.
It is important also not to overlook the fact that the hard determinist has the option of offering various non-utilitarian accounts of moral responsibility. A hard determinist may, for example, defer to virtue ethics in developing an account of moral responsibility, so that judgements about a person's blameworthiness (or praiseworthiness) for performing a certain action are reduced to judgements about that person's character. Such a view was, in fact, endorsed by Spinoza, as Michael Slote illustrates in «Ethics Without Free Will», pp.375-79. See also Pereboom, Living Without Free Will, pp.148-52, and Ben Vilhauer, «Hard Determinism, Remorse, and Virtue Ethics,» Southern Journal of Philosophy 42 (2004), pp.547-64, where Vilhauer attempts to accommodate the phenomenon of remorse within hard determinism by incorporating a circumscribed consequentialism about punishment into a broad-based virtue ethic. On Vilhauer's approach, see also fns 38 and 39 below. For a more pessimistic view of the compatibility of virtue ethics and determinism, see Haji, Deontic Morality and Control, ch. 11.
[Foot Note 24]
Richard J. Arneson, «The Smart Theory of Moral Responsibility and Desert,» in Serena Olsaretti (ed.), Desert and Justice (Oxford: Clarendon Press, 2003), p.233.
[Foot Note 25]
A further objection to the instrumentalist account is that agents, on this view, would be treated as blameworthy (by, for example, expressing indignation toward them) even though they do not deserve to be treated in this way -- and this, it might be claimed, is unfair and morally wrong (such an objection is voiced by Pereboom, Living Without Free Will, p.156). However, if the person who is treated as blameworthy stands to gain greatly (in moral terms) from being so treated, then perhaps the treatment is morally justified despite being undeserved. An equally plausible option, one advocated by Pereboom, is to do away with praise and blame, and instead engage in moral admonishment and encouragement. As Pereboom states,
One could explain to an offender that what he did was wrong and then encourage him to refrain from performing similar actions in the future. The hard incompatibilist can maintain that by admonishing and encouraging a wrongdoer one might communicate a sense of what is right, and a respect for persons, and that these attitudes can lead to salutary change.
(Living Without Free Will, p.157)
For a comprehensive list of, and response to, the traditional objections to the instrumentalist account of moral responsibility, see Manuel Vargas, «Moral Influence, Moral Responsibility» (unpublished).
[Foot Note 26]
See van Inwagen, «Response to Slote,» p.386.
[Foot Note 27]
See H.J. McCloskey, «An Examination of Restricted Utilitarianism,» Philosophical Review 66 (1957), pp.468-69; «A Non-Utilitarian Approach to Punishment,» Inquiry 8 (1965), pp.255-59; and «Utilitarian and Retributive Punishment,» Journal of Philosophy 64 (1967), pp.91-102.
[Foot Note 28]
It may be worth noting, however, that the account of moral responsibility defended here does not state that the attitudes and practices characteristic of moral responsibility are justified because of the utility such attitudes and practices may have for some group or community. Rather, the view is that our responsibility-characteristic practices, particularly the practices of praising and blaming, are justified when they are likely to influence or pressure individuals to behave in morally desirable ways. It may be possible, however, to raise the objection from injustice even against this latter account of moral responsibility, as is indicated in fn. 25 above.
[Foot Note 29]
Smilansky, «Utilitarianism and the `Punishment' of the Innocent: The General Problem,» Analysis 50 (1990), p.257. See also T.L.S. Sprigge, «A Utilitarian Reply to Dr. McCloskey,» Inquiry 8 (1965), pp.272-75.
[Foot Note 30]
Van Inwagen has, of course, the option of filling out his example in various ways, thus rendering it more `believable'. But this is likely to raise the problems mentioned below with respect to cases that are `true to life'.
[Foot Note 31]
Admittedly, it is more common to view utilitarianism and other normative ethical theories as theories that attempt to specify necessary and sufficient conditions for the application of concepts such as moral obligation and moral rightness. In line with this view, it is open to the non-utilitarian to say that the principle of utility enshrined by utilitarianism (roughly, the principle that an act is right or good if and only if it has the greatest utility -- e.g., it produces the most favourable balance of happiness over unhappiness) is to be thought of as describing what would be right and wrong under any conceivable or logically possible circumstances. But then the utilitarian account of punishment would hold that there is no logically possible world in which it is right or good (in the utilitarian sense of these terms) for people to be unjustly punished, and this view can be defeated simply by way of logically possible counterexamples. Perhaps this is what lies behind McCloskey's view that,
Against the utilitarian who seeks to argue that utilitarianism does not involve unjust punishment, there is a very simple argument, namely, that whether or not unjust punishments are in fact useful, it is logically possible that they will at some time become useful, in which case utilitarians are committed to them. («A Non-Utilitarian Approach to Punishment,» pp.254-55)
However, even if the principle of utility is to be construed as a theory about what would be right or wrong under any conceivable circumstances, it is unlikely that commonsense morality should be thought of in the same terms. As Timothy Sprigge notes, «Plain men will probably admit that if the empirical nature of the world had been very different then different moral sentiments would often have been appropriate» («A Utilitarian Reply to Dr. McCloskey,» p.273). But if the principles of commonsense morality have force only in the actual and nearby worlds, then that would explain our abhorrence when faced with far-fetched cases of the punishment of innocents. Alleged logically possible `counterexamples', in other words, only show up the limitations of commonsense morality, not of utilitarianism. See, however, McCloskey's different assessment of the matter in «Utilitarian and Retributive Punishment,» pp.93-94.
[Foot Note 32]
Sprigge, «A Utilitarian Reply to Dr. McCloskey,» p.275.
[Foot Note 33]
See Jonathan Glover, Causing Death and Saving Lives (London: Penguin, 1977), pp.80-82, and Peter Singer, Practical Ethics, 2nd ed. (Cambridge: Cambridge University Press, 1993), pp.99-100. A further option is to adopt a `rule' or `restricted' form of utilitarianism, according to which it is the practice of putting into jail those found guilty after a fair legal trial, rather than any particular instances of this practice, that can be justified on utilitarian grounds. See, however, McCloskey's «An Examination of Restricted Utilitarianism» for some forceful arguments against this form of utilitarianism.
[Foot Note 34]
As this indicates, the utilitarian should leave open the possibility that, although the punishment of innocents is unjust, in some circumstances it is permissible, or even obligatory, to override the dictates of justice. Interestingly, McCloskey himself endorses this position -- see his «A Non-Utilitarian Approach to Punishment,» p.251.
[Foot Note 35]
Jonathan Bennett, «Accountability», in Zak van Straaten (ed.), Philosophical Subjects: Essays Presented to P.F. Strawson (Oxford: Clarendon Press, 1980), p.20, emphases in the original.
[Foot Note 36]
Cf. Richard Brandt's similar list in «Determinism and the Justifiability of Moral Blame,» in Sidney Hook (ed.), Determinism and Freedom in the Age of Modern Science (New York: New York University Press, 1958), p.140.
[Foot Note 37]
Admittedly, (e) above only mentions feelings of remorse, not feelings of guilt, and the two should not be confused. There are, however, obvious connections between the two; hence my decision to discuss the two together.
[Foot Note 38]
Hilary Bok, Freedom and Responsibility (Princeton, NJ: Princeton University Press, 1998), pp.168-69; quoted in Pereboom, Living Without Free Will, p.206. See, however, Ben Vilhauer's criticisms of Bok's analogy (on pp.557-58 of Vilhauer's «Hard Determinism, Remorse, and Virtue Ethics»), which lead Vilhauer to a different analogy, one which explains remorse on the model of suffering in sympathy with friends and loved ones (he writes, for example, that «sympathetic suffering prompts us to try to relieve the suffering of the friend or loved one, much as remorse prompts wrongdoers to try to make amends» [p.551]).
[Foot Note 39]
See Pereboom, Living Without Free Will, p.206. Ben Vilhauer plausibly notes that remorse does not always have such beneficial behavioural consequences, and when it doesn't, the hard determinist will be unable to appeal to consequentialist considerations in justifying the appropriateness of feeling remorse. The hard determinist, furthermore, does not have the option of invoking the notion of desert, since the hard determinist cannot say that the remorseful person deserves to suffer emotionally on account of their wrongful actions. Vilhauer, however, contends that the option remains of appealing to virtue ethics. As he puts it,
[P]art of the solution to our problem about remorse is to stop thinking of it as punishment and, instead, to think of it as just one member of a set of emotional engagements, all of which depend upon suffering in sympathy with others, and none of which involve desert. Like love and friendship, remorse is a virtuous state of character that depends upon suffering in sympathy with people one cares about. («Hard Determinism, Remorse, and Virtue Ethics,» p.552.)
[Foot Note 40]
Similar views on guilt are expressed by Henry Sidgwick, The Methods of Ethics, p.71, and Bruce Waller, Freedom Without Responsibility, pp.164-69. Waller's discussion of guilt is set within ch.12 of his book, where he provides some very good reasons for thinking that the denial of moral responsibility does not entail the rejection of all genuine morality.
[Foot Note 41]
P.F. Strawson, «Freedom and Resentment,» in his Freedom and Resentment and Other Essays (London: Methuen & Co., 1974), p.10, emphasis in the original. Strawson divides the reactive attitudes into those that are associated with demands on others for oneself (`participant' or `personal' reactive attitudes), those that are associated with demands on others for others (`generalised' or `vicarious' reactive attitudes), and those that are associated with demands on oneself for others (`self-reactive attitudes') -- see pp.14-15 of Strawson's «Freedom and Resentment». The argument of the main text is not restricted to any one of these three varieties of the reactive attitudes. (I should note that Strawson, like most compatibilists, does not think that determinism threatens merit-based moral responsibility, whereas I am assuming for the sake of argument that it does.)
[Foot Note 42]
See Pereboom, Living Without Free Will, pp.199-204.
[Foot Note 43]
Cf. Pereboom, Living Without Free Will, pp.200-201, where a slightly different view on this matter is presented.
[Foot Note 44]
See Pereboom, Living Without Free Will, ch. 6.
[Foot Note 45]
It may also be argued that even if hard determinism imperils certain aspects of commonsense morality, this may be offset by the resources hard determinism has in resolving some problems lying at the heart of our everyday moral thinking. Michael Slote takes such a view, arguing that the phenomenon of moral luck gives rise to a nest of inconsistencies within commonsense morality, but this can be overcome by going beyond our ordinary moral thinking and adopting a utilitarian account of moral responsibility. See Slote, «Ethics Without Free Will,» pp.372-75.
[Foot Note 46]
Pereboom, Living Without Free Will, p.213.
[Foot Note 47]
[Foot Note 48]
Smilansky, «The Ethical Advantages of Hard Determinism,» Philosophy and Phenomenological Research 54 (1994), p.356. See also Smilansky, Free Will and Illusion, ch. 10, where some further benefits of hard determinism are identified. It should be borne in mind that, according to Smilansky, these benefits of hard determinism are but a small consolation in the face of the disastrous personal, moral and social impact of accepting the absence of libertarian free will.
[Foot Note 49]
Smilansky, «The Ethical Advantages of Hard Determinism,» pp.358-59, emphasis in the original.
[Foot Note 50]
Pereboom, Living Without Free Will, p.152.
[Foot Note 51]
Smilansky, «The Ethical Advantages of Hard Determinism,» p.358.
[Foot Note 52]
[Foot Note 53]
Pereboom, Living Without Free Will, p.152.
[Foot Note 54]
[Foot Note 55]
Smilansky, «The Ethical Advantages of Hard Determinism,» p.359.
[Foot Note 56]
The abbreviations of the moral principles outlined below, such as K and OW, are taken from Haji and Pereboom. Haji's argument is developed in Moral Appraisability: Puzzles, Proposals, and Perplexities (New York: Oxford University Press, 1998), ch. 3, and in «Moral Anchors and Control,» pp.182-88.
[Foot Note 57]
The controversial status of the principal that `ought implies can' is also due to differing views as to what kind of implication relation holds between `ought' and `can'. On this, see Walter Sinnott-Armstrong, «`Ought' Conversationally Implies `Can',» Philosophical Review 93 (1984), pp.249-61. For a spirited defence of principle K and related principles, see Haji, Deontic Morality and Control, chs 2-4.
[Foot Note 58]
See Haji, «Moral Anchors and Control,» pp.177-82, 202, and Deontic Morality and Control, pp.16-24.
[Foot Note 59]
Although Haji considers both K and OW intuitively plausible, he also claims that these principles enjoy theory-based support, since both figure as theorems within (what Haji takes to be) some of our best current theories of moral obligation -- for example, those developed by Michael Zimmerman in The Concept of Moral Obligation (Cambridge: Cambridge University Press, 1996) and Fred Feldman in Doing the Best We Can (Dordrecht: D. Reidel, 1986).
[Foot Note 60]
Haji, Deontic Morality and Control, p.3.
[Foot Note 61]
I should note that Haji's argument is targeted against soft determinism, though I think no damage is done to his argument if it is steered towards hard determinism.
[Foot Note 62]
A further consequence is that, since you can never be morally obligated to act otherwise, there would be no point in conferring any moral obligations on you.
[Foot Note 63]
Given that the above argument is applicable to both moral obligation and moral wrongness/rightness, it is somewhat strange that Pereboom concedes the argument's force in relation to the former category while disputing its cogency in relation to the latter category. I may note, however, that Pereboom has recently reported (via personal correspondence) that he now thinks that determinism only undermines some, but not all, uses of `ought' (a view prefigured somewhat on pp.147-48 of Pereboom's Living Without Free Will).
[Foot Note 64]
See Haji, «Moral Anchors and Control,» p.189.
[Foot Note 65]
I must confess that I was initially persuaded by Haji's objection, but Pereboom (in personal correspondence) brought me to my senses.
[Foot Note 66]
Pereboom, Living Without Free Will, p.144.
[Foot Note 67]
[Foot Note 68]
Pereboom notes that what holds for GR also holds for the following intuitively true principle:
GR': There is some correlation between an action's bringing about the greatest good overall in worlds accessible to S and that action's being right for S.
See Pereboom, Living Without Free Will, p.145, fn. 45.
[Foot Note 69]
Haji has replied that Pereboom's view that GR is true is question-begging:
[I]f one finds GR intuitively plausible, presumably one does so independently of having reflected on the view that there is a requirement of alternative possibilities for rightness, wrongness, and obligatoriness, and hence that determinism is incompatible with deontic morality. It is not clear, consequently, that having engaged in the relevant sort of reflection, one would still find GR intuitively attractive. (Deontic Morality and Control, p.71)
In other words, GR appears plausible only when principles such as K and OW are overlooked or presupposed to be false. But I doubt that one needs to reject or overlook K and OW in order to appreciate the intuitive plausibility of GR (though this is not to say that GR ought to be accepted as true). I may add here that Haji does not reject GR outright, but thinks that GR inchoately expresses the truth of the matter, which is better expressed in terms of Zimmerman's `world utilitarian' principle (stated on p.72 of Haji's Deontic Morality and Control).
[Foot Note 70]
Haji, Moral Appraisability, p.146. Haji adds that we are to «construe this case as one in which Deadly is under no delusion that she takes herself to be doing wrong when she injects the patient with C» (p.152). There is a problem with the case of Deadly's Defeat that seems to have gone unnoticed. Haji assumes that it is obligatory for Deadly to inject her patient with medicine C. But given that the doctor is (non-culpably) ignorant of the fact that the patient requires C in order to be cured, how can the doctor have an obligation to administer C to the patient? This suggests that, in order to determine what is objectively right or obligatory for a person, we must take into account that person's epistemic predicament (e.g., what facts are accessible to that person).
[Foot Note 71]
See Haji, Moral Appraisability, chs 8 and 9, and Deontic Morality and Control, ch. 10.
[Foot Note 72]
Haji claims, however, that Pereboom is not entitled to accept BW. Haji first distinguishes two varieties of BW:
BW1: Sometimes, when S is blameworthy for performing A in a deterministic world, it was morally wrong for S to perform A in that world.
BW2: Sometimes, when S is blameworthy for performing A in an indeterministic world, it was morally wrong for S to perform A in that world.
Clearly, as Haji observes, Pereboom is only concerned with BW1. Haji then goes on to state that a hard incompatibilist of the Pereboom variety «should reject this principle [i.e., BW1] (even though he might find it intuitive) because hard incompatibilism entails that both determinism and its denial are incompatible with blameworthiness» (Deontic Morality and Control, p.73). But this is mistaken, for Pereboom (unlike Haji) can accept BW1, understood as the claim that If someone is blameworthy for doing A in a deterministic world, their performance of A in such a world is morally wrong. For Pereboom, qua hard incompatibilist, this conditional would be true since its antecedent is false and its consequent is true.
[Foot Note 73]
The disappointment is not removed by the defence offered by Haji against objections of this sort. In response to cases like that involving the Nazi officer, Haji writes that, «I strongly doubt that (in such cases) S ought (morally) to have realized that failing to acquire BF [the belief that torturing people for fun is wrong] is wrong, failed in this obligation, and was culpable for doing so» (Moral Appraisability, p.165; similar remarks are made in Haji, «On Psychopaths and Culpability,» Law and Philosophy 17 (1998), p.137). Unfortunately, Haji offers no explanation for the doubt he expresses here. It seems, however, that the doubt is quite misplaced, for most people would agree that (a) the Nazi officer ought to have realized that it was wrong of him to fail to acquire the germane belief (i.e., the belief that killing people solely on account of their race is wrong), (b) he failed in this obligation, and (c) was culpable for doing so.
[Foot Note 74]
Pereboom, Living Without Free Will, p.147.
[Foot Note 75]
Haji, «Moral Anchors and Control,» p.185. See also Haji, Deontic Morality and Control, pp.52-53, including his fn. 12.
[Foot Note 76]
Pereboom, Living Without Free Will, p.143.
[Foot Note 77]
As indicated, the notions of `right', `wrong', `obligatory', `permissible', and `good' are to be understood (in this and the following paragraph) in non-consequentialist terms.
[Foot Note 78]
Haji, «Moral Anchors and Control,» p.187. I might add that it will not help to bring the category of moral badness into the picture by holding that, even though the killing is not amoral, it remains impermissible because it is morally bad. For badness would then be intimately linked to obligation: to say that action A is bad would (in this case at least) entail that A is not permissible, which entails that one ought not to do A. The working assumption here, however, is that there are no moral obligations.
[Foot Note 79]
For helpful comments on a previous draft, I must thank Daniel Cohen, Ish Haji, Saul Smilansky, and Derk Pereboom. I am also grateful for the feedback provided by audiences at Monash University and at the 2005 Conference of the Australasian Association for Philosophy (held at the University of Sydney), where an earlier version of this paper was presented.