Doxastic Voluntarism: A Sceptical Defence

DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 1 Danny Frederick* International Journal for the Study of Skepticism 3 (1): 24-44 (2013) Abstract Doxastic voluntarism maintains that we have voluntary control over our beliefs. It is generally denied by contemporary philosophers. I argue that doxastic voluntarism is true: normally, and insofar as we are rational, we are able to suspend belief and, provided we have a natural inclination to believe, we are able to rescind that suspension, and thus to choose to believe. I show that the arguments that have been offered against doxastic voluntarism fail; and that, if the denial of doxastic voluntarism is part of a strategy to defeat scepticism, it is inept, because knowledge presupposes doubt. Keywords Alston, belief; doubt; doxastic voluntarism; knowledge, scepticism, Williams. 1. Introduction Doxastic voluntarism is generally denied by contemporary philosophers. It is the view that, at least in some circumstances, we have voluntary control over our beliefs. There are stronger and weaker versions of this view. William Alston (1988) distinguishes four kinds of voluntary control that it might be claimed we sometimes have over our beliefs. Two of these are direct forms of control. • Basic voluntary control is the ability to bring it about that one believes, or does not believe, a specific proposition at will, as a basic mental act, not by means of doing something else (1988, 260-61). • Non-basic immediate voluntary control is the ability to bring it about that one believes, or does not believe, a specific proposition as a non-basic act but ‘in one fell swoop, i.e., during a period of activity uninterruptedly guided by the intention to produce that belief’ (1988, 274), without having to return to the attempt a number of times after having been occupied with other matters (1988, 269). We have such control in ordinary cases of opening a door, informing someone of something, and turning on a light. ‘To succeed in any of these requires more than a volition on the part of the agent; in each case I must perform one or more bodily movements and these movements must have certain consequences, causal or conventional, in order that I can be said to have performed the non-basic action in question…[But] We suppose that if the agent will just voluntarily exert herself the act will be done’ (1988, 269). The two indirect forms of control are: long-range voluntary control, involving interrupted activities designed to bring about belief in a specific proposition, such as selective exposure to evidence, hypnotism, or seeking the company of believers and avoiding non-believers; indirect voluntary influence, in which activities, such as seeking evidence to resolve a question one way or the other, bring about a belief but without us intentionally bringing about that specific belief. Alston denies that we have either form of direct voluntary control over our beliefs (1988, 263-274). He insists that we only rarely have long-range voluntary * Webpage: http://independent.academia.edu/DannyFrederick; Email: [email protected] DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 2 control; and, while he concedes that we have indirect voluntary influence, he maintains that this does not amount to being able to select the specific propositions we will believe (1988, 273-80). I will not discuss the indirect forms of control. I agree with Alston (1988, 263) that, while we do not in fact have basic voluntary control of our beliefs, it seems possible that we could have or that some person might; but I will not defend that possibility here. My concern is with non-basic immediate voluntary control. My argument has negative and positive components. On the negative side, I show that the arguments against the possibility of non-basic immediate voluntary control are unsuccessful. On the positive side, I argue that, for each of the propositions we believe, there are circumstances in which we have non-basic immediately voluntary control over that belief; that we normally also have non-basic immediate voluntary control over whether we are in those circumstances; and that, as a consequence, we normally have non-basic immediate voluntary control over all our beliefs. There is a qualification: the beliefs must be ones of which we are aware. Although in recent years some philosophers have defended weak or limited versions of doxastic voluntarism (for example, Funkhouser 2003, Ginet 2001, Heller 2000, Nickel 2010, Weatherson 2008), none has come close to recognising the actual extent of our voluntary control over our beliefs. In section 2, I review Alston’s arguments against basic and non-basic immediate voluntary control, criticise them, reject most of them and amend the rest. In section 3, I offer an account of doxastic voluntarism. In section 4, I criticise Williams’ arguments against doxastic voluntarism. In section 5, I consider Hieronymi’s argument against the possibility of basic voluntary control of belief, but only to note the weaknesses in her argument that are relevant to our discussion. In section 6, I conclude. 2. Alston Perhaps the most influential case against doxastic voluntarism is due to Alston. He argues, in sequence, against basic and non-basic immediate voluntary control of belief, but he uses essentially the same arguments against each (1988, 269). His procedure is to invite us to share his intuitions about counter-examples which seem clear to him. To save on words, I will use ‘choose to believe’ as shorthand for bring it about that one believes, either as a basic act or as a non-basic but immediate act. First, Alston maintains that we cannot choose to believe propositions which are obviously false. For example, no one could just choose to believe the proposition that the US is still a colony of Britain, even if offered a substantial reward for believing it (1988, 263-64). Second, he avers that we cannot help but believe propositions which are obviously true. For example, he claims that it is perfectly clear that I have no power at all to refrain from the belief that the tree has leaves on it when I see a tree with leaves on it just before me in broad daylight with my eyesight working perfectly. Other examples, he says, would be provided by ordinary beliefs formed by introspection, memory, and simple uncontroversial inferences (1988, 264). Third, with respect to propositions which are neither obviously true nor obviously false, he says that, given a set of contrary propositions, we cannot help but believe the one that seems most likely to be true (1988, 266). In cases where rival propositions seem equally likely, and where no other rival seems more likely, he DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 3 affirms that we cannot believe any of the propositions. Of course, in such a situation in which we have to act, we might resolve to act as if one of the propositions were true, as would a field commander in wartime faced with questions about the current disposition of enemy forces, where the information at his disposal does not tell him just what that disposition is. And in a situation in which we are theorising, we might suppose for the sake of argument that one of the propositions is true, as when a scientist adopts a working hypothesis which she subjects to tests. But acting-as-if and supposing are not believing (1988, 264-68). Some of Alston’s examples seem persuasive. But others seem misconceived. Further, even insofar as the persuasive cases are correct, this is for reasons other than the ones that Alston gives. It may be easiest to begin by showing why some of the examples seem plainly wrong. I will then discuss the other examples. 2.1 ‘Obvious Truths’ Alston claims that I have no power at all to refrain from the belief that the tree has leaves on it when I see a tree with leaves on it just before me in broad daylight with my eyesight working perfectly. Yet it is characteristic of some philosophers that they do refrain from belief in precisely such circumstances. Descartes, for example, wondered whether he might be dreaming, or suffering some form of illusion, in such circumstances; and he resolved to withhold belief until he could find an argument to defeat these doubts (1641/1931). But non-philosophers also sometimes refrain from belief in such circumstances. For example, someone familiar with the work done on film sets might wonder whether what she sees is a tree with leaves or just a construction of papier-mâché and plastic, or perhaps even a hologram; and a budding botanist might wonder whether it is not in fact some species of bush rather than a tree. The same kind of response can be made to a similar example that Alston offers. He says (1988, 270): When I look out my window and see rain falling, water dripping off the leaves of trees…I no more have immediate control over whether I accept those propositions than I have basic control. I form the beliefs that rain is falling, etc. willy-nilly. There is no way I can inhibit these beliefs. At least there is no way I can do so on the spot, in carrying out an uninterrupted intention to do so. How would I do so? The answer seems clear: I consider that there may be potential alternative explanations. If I have it in me, I might even come up with some potential alternative explanations. For example, I might wonder whether someone is having fun with a hosepipe or sprinkler, or whether it is not water but some chemical being sprayed from an aeroplane by a pilot who has mistakenly flown to the wrong location and thinks he is spraying crops. Or I might wonder whether it is not rain that is falling but acid rain or some compound liquid that it would take an imaginative scientist concerned with atmospheric changes to conjecture. There is no interpretation of perceptions that may not be doubted by someone who is serious about the pursuit of knowledge, as is shown by many scientific discoveries. For example, Galileo acknowledged that the observed variations in brightness of the planets as revealed to the naked eye was inconsistent with the variations in their distance from the earth as predicted by Copernican astronomy; but he doubted DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 4 what he saw with his own eyes and developed a telescope which revealed variations in brightness consistent with the Copernican theory (Feyerabend 1975, 141-43). As Galileo put it, ‘It is, therefore, better to put aside the appearance, on which we all agree, and to use the power of reason either to confirm its reality or to reveal its fallacy’ (1632/1953, 256). Even so simple an observation statement as ‘Here is a glass of water’ implies that the receptacle will exhibit the law-like behaviour of glass and that its contents will behave in the law-like way that water does; and these implications may be inconsistent with other observation statements (Popper 2002, 76, 86). Indeed, even if all scientists are agreed that the observation statement ‘Here is a glass of water’ is true because the liquid in the glass has passed all the scientific tests for being water, they may still be mistaken. Urey’s discovery of heavy hydrogen in 1931 showed that what scientists had previously taken to be water was in fact a mixture of two physically different substances (Popper 1966, 374-75). There are many other examples. When we see the sun on the horizon, ‘in broad daylight with eyesight working perfectly,’ the sun is actually below the horizon, but it appears to be on the horizon because its light is refracted by the earth’s atmosphere. The star we think we see is not there, because it has ceased to exist during the time it takes light to travel to us. An apparent fish is actually a mammal. The apparent movement of the sun is the motion of the earth. The apparent gold is iron pyrites. Many apparent stars are actually asteroids. The two lines in the Müller Lyer illusion that look unequal are in fact equal. The apparent witch is just a strange (or disliked) woman. Apparent design is just an evolved product of natural selection. Things we perceive as solid objects are mostly empty space. And so on. Anyone familiar with such facts can normally doubt what he seems to perceive by his senses. The same applies to Alston’s other candidates for ‘obvious truths.’ We all often doubt our memories, especially if we have learned enough psychology to know some of the tricks that memory can play (Tavris & Aronson 2007, 68-96). Even simple inferences such as double negation, modus ponens and conjunction elimination have been doubted or denied by some logicians in order to try to solve serious logical problems (Priest & Thomason 2007, 96-98). We can also refrain from belief in the deliverances of introspection, if we are aware that philosophers dispute each other’s accounts of the phenomenology of thought (Schwitzgebel 2008), or that psychologists have discovered that we are often mistaken about the influences on our own thinking (Nisbett & Wilson 1977), or even if we have simply experienced how difficult it can be to tell whether it is anger or fear that we are feeling. What these sorts of examples indicate is that there are no obvious truths, though we may think that some things are obvious to the extent that we are ignorant, unimaginative or uncritical. 2.2 Propositions That Seem Most Likely Alston also seems plain wrong when he says that in cases in which one proposition seems more likely than its rivals we cannot help but believe it. Funkhouser (2003, 185) notes that it sometimes happens that an agent believes a proposition while sincerely conceding that the evidence points against that belief. He instances religious beliefs or beliefs about the goodness of friends or family members. He also refers to delusional disorders in which believers are indifferent to contrary evidence. In all these cases he regards the agents as irrational. DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 5 However, even if we concede the irrationality of the agents in all of Funkhouser’s examples, withholding belief from the proposition that seems most likely is not in itself irrational. For, the hypothesis that seems most likely could be false. For example, Edison’s production of his electric light refuted the unanimous scientific opinion that such a light was impossible (Kuhn 1977, 238). Indeed, scientific advance depends upon people withholding belief from the proposition that seems most likely. In the early part of the seventeenth century, the available evidence seemed clearly in favour of the geocentric theory; yet Galileo not only withheld belief from the geocentric theory, he also set to work on developing its heliocentric rival (Kuhn 1957; Feyerabend 1975, 69-161). In the later part of that century, Kepler’s laws of planetary motion seemed clearly to be the most likely of all those available, while Galileo’s laws seemed clearly to be the most likely of all those available for terrestrial motion. But so far was it from being the case that informed and rational people could not help but believe these most likely propositions, that Newton set out to develop a better theory. And he was successful: his new theory contradicted Kepler’s and Galileo’s laws, yet it explained all that they explained and other things besides, such as the motions of the tides (Popper 1983, 139-45, 19091). Given their aim of making a contribution to the growth of knowledge, and given that scientific progress comes principally from new theories which contradict their successful predecessors, Galileo and Newton were not irrational in withholding belief from the hypotheses that seemed most likely. We are not bound, psychologically or rationally, to believe the most likely proposition. Indeed, if we were, scientific progress would be baffling. New theories do not just pop into people’s minds fully formed: it takes a great deal of creative and painstaking work to develop them from the germ of an idea into an interconnected set of propositions and models with explanatory power. How could anyone devote a substantial part of his life to such a project if he believed the existing successful theories which his incipient new theory contradicts? 2.3 No Proposition Seems Most Likely Alston seems partially right when he says that, in cases in which no proposition seems more likely than its rivals, we cannot believe any of them. This seems to hold for the sorts of examples he gives, such as the scientist testing a hypothesis and the military commander. But there are other cases for which it seems not to hold. For example, although I seem to be sitting at my computer, I can doubt whether this is really so: perhaps I am dreaming or hallucinating; perhaps I am a brain in a vat; perhaps I am all that exists, with everything else being mere appearance; perhaps I am actually in a straightjacket in an asylum and suffering from an insane delusion; and so on. There does not appear to be any consideration that favours the proposition that I am sitting at my computer, rather than suffering an illusion, since any putative consideration may be part of the illusion. Given all the available evidence – indeed, it seems, given all the possible evidence – each of the two propositions seems as likely as the other. Yet, when I am sitting at my computer, I can and usually do believe that I am sitting at my computer. 2.4 ‘Obvious Falsehoods’ Alston seems to be right that we cannot choose to believe such ‘obviously false’ propositions as that the US is still a colony of Britain. But this is not because DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 6 the propositions are obviously false. The fact that anything may be doubted means not only that there can be no obvious truths but also that there can be no obvious falsehoods: even some self-contradictions are held to be true by dialetheic logicians because this, they say, offers a more plausible overall account of logical paradoxes (Priest 2004). 2.5 Summary On Alston’s view, our beliefs are determined willy-nilly by evidence that presents itself to us: • where evidence for a proposition is decisive, we are compelled to believe the proposition; • where evidence against it is decisive, we cannot believe it; • where evidence is not decisive, we are compelled to believe the proposition that seems most likely; • if no proposition seems more likely than all of its rivals, then we cannot believe any of them. In contrast, I have pointed out (and will argue further below), first, that because there are always potential alternative explanations, there is no such thing as decisive evidence, and no evidence can appear decisive to a thoroughly critical enquirer. As a consequence, it is always open to us to withhold belief, at least as long as we retain our critical faculty (and are not, for instance, suffering some form of psychopathology). Second, we are not constrained, psychologically or rationally, to believe the hypothesis that seems most likely to us. Indeed, if we were, the fact of scientific progress would be baffling. Third, there are cases in which we can and typically do believe a proposition that seems no more likely than any of its rivals; and there are cases in which people believe a proposition that seems less likely than its rivals. 3. Doxastic Voluntarism My thesis is as follows. There are some propositions that we have a natural tendency to believe. Under normal circumstances, we can subject any of these propositions to gratuitous doubt and thereby suspend belief in it. Because the doubt is gratuitous, the natural tendency to believe the proposition is inhibited rather than destroyed. Further, because the doubt is gratuitous, we are able simply to dismiss the doubt. In that position, we can, as a basic act, dismiss the doubt, in which case our natural tendency will immediately bring about our belief in the proposition; or we can, as a basic act, maintain the doubt, in which case we will continue to withhold belief in the proposition. In that position, then, we have non-basic immediate voluntary control over the belief: we can choose either to believe by dismissing the doubt, or to withhold belief by maintaining the doubt. Furthermore, since we can get ourselves into that position simply by entertaining a gratuitous doubt, for each of the propositions we have a natural tendency to believe we can normally bring it about that we believe it ‘in one uninterrupted activity,’ if we ‘just voluntarily exert ourselves to do it,’ by suspending belief and then rescinding the suspension. So, normally, where we have a natural tendency to believe a proposition, we have non-basic immediate voluntary control over that belief. Further still, where we have such control, we can choose to believe the proposition for the sake of a reward. DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 7 There is an important qualification. We have many beliefs of which we are not aware; and we cannot doubt a proposition without being aware of it. Of course, we often become aware of a belief because a surprising observation or a novel theory proposed by someone else conflicts with it; that is, we become aware of the belief because it has been called into doubt. For example, perhaps few people were aware that they believed that simultaneity is absolute until Einstein proposed otherwise. But we cannot autonomously call a belief into doubt unless we are already aware that we have the belief. So, in the case of beliefs of which we are not aware, we can have a natural tendency to believe a proposition without having voluntary control over the belief. For simplicity, I will suppress this qualification in what follows. I say that we have a ‘natural tendency to believe’ a proposition if we have either a natural inclination to form a belief in it or a natural inclination to maintain a belief in it. I take it to be uncontroversial that we have a natural inclination to form beliefs in the following three ways. First, we are naturally inclined to adopt the beliefs that are generally accepted or inculcated in our culture or subculture, including those passed down as part of formal and informal education and training. Second, we are naturally inclined to form beliefs that are prompted by our perceptions, memories and introspections; though what beliefs these happen to be depends upon our interpretative scheme. Part of this scheme is inherited biologically: a person’s brain conveys to his conscious mind filtered and stereotyped information that it ‘guesses’ will be useful, either in general or, more particularly, in the light of current desires or interests; and some of the stereotypes are inborn, for example, particular neurons appear to have the function of ‘looking for’ particular geometric shapes in the visual field, such as rectangles (Popper & Eccles 1977, 91-92, 261-71). Other parts of a person’s interpretative scheme are supplied by his culture, including even specific subcultures such as academic disciplines (Kuhn 1970, 62-65). For example, where a layman observes an oscillating iron bar, with a mirror attached, sending a beam of light to a celluloid ruler, a physicist observes the electrical resistance of a coil (Duhem 1914/1954, 145); indeed, the layman might not even recognise the bar as iron or realise that the spot of light on the ruler had been sent there by the mirror (Watkins 1984, 266). Third, we seem naturally inclined to form a belief in a proposition that is both plausible and substantially more plausible than any of its available rivals. We also seem to have a natural inclination to maintain any of the beliefs we have already acquired. This inclination may be stubborn, with a person continuing to maintain a belief even after becoming aware that the belief was acquired accidentally or mistakenly. For example, a young American may continue sincerely to believe that Christianity offers the only path to paradise even though she has learned from her comparative religion class that, had she been raised in Iran, she would sincerely have believed that Islam provides the only path to paradise (Funkhouser 2003, 187). Further, in experiments, people who formed beliefs given some slender and fictitious data retained those beliefs even after the data was revealed to them to be fictitious (Funkhouser 2003, 188). Indeed, this inclination to maintain existing belief may be so stubborn that the belief is maintained even by people who acknowledge that the evidence is all against it (as noted in subsection 2.2): the tendency to maintain an already acquired belief may be stronger than the tendency to believe the proposition that seems most likely. Further, the stubbornness of the inclination may be exhibited in recidivism: people sometimes reject a belief in the face of what they sincerely DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 8 acknowledge is a refutation, but then, at a later stage, they fall back into the old belief because the acknowledged refutation is not present to consciousness, though it need not have been forgotten. Much of the time, our natural inclinations to form beliefs lead us straight into belief, without us thinking about it. For example, when I have a perception as of the telephone ringing, I normally automatically believe that the telephone is ringing. However, normally, any of these inclinations to form beliefs can be inhibited by adopting a critical attitude. For example, if we attend critically to our perceptions as they occur, we can withhold belief, as when I attend to a developing scene which I think may be illusory. Further, where a belief has already resulted from one of these natural inclinations, we can, by turning our critical attention to it, suspend belief in it. We withhold or suspend belief in a proposition by entertaining a doubt about it, that is, by taking seriously a hypothesis that is inconsistent with it. But doubts are of different kinds. Some doubts are gratuitous. We can distinguish two kinds. In speculative doubts we entertain a possible alternative hypothesis which is not testable. This seems to be the case with the doubt (subsection 2.3) as to whether I am sitting at my computer or suffering some form of global illusion. Thus, although I believe I am currently sitting at my computer, I can, by entertaining speculative doubts, suspend that belief. If I continue to entertain these doubts, then, although I may act as if I am sitting at my computer, I will truly deny that I believe that I am. I might even continue to say that I am at my computer, but explain this in terms of the convenience of being able to talk with the vulgar while thinking with the learned (Berkeley 1710/1910, 138). However, if I dismiss the doubts, I will thereby believe that I am at my computer, because I have a natural inclination to believe propositions that are prompted by my perceptions. I imagine this has been the standard response of generations of students to Descartes’ Meditations, since all must have shared his doubts (if only temporarily) but few can have been persuaded by his anti-sceptical arguments. The second kind of gratuitous doubt is non-speculative. In this case, the alternative hypothesis that is taken seriously is again only a possibility, but it is one which, in principle at least, is testable. This seems to be the case with the doubt (subsection 2.1) that is generated about the rain by taking seriously the alternative hypothesis of an airborne crop-sprayer. The doubt remains gratuitous so long as the alternative hypothesis is not tested. Gratuitous doubts may be dismissed at will, as a basic act; but if they are not dismissed, the business of life will tend to distract us from them, so that we end up slipping back into belief unintentionally. Thus, even the sceptic and the idealist will normally believe in the existence of the external world when he is not considering his philosophical position: he will slide back into an animal belief which contradicts his considered opinion (recidivism). But he can normally simply choose to resume his philosophical reflections; and whenever he is asked whether he believes in the existence of the external world he may truly reply that he does not, because the question recalls to his mind speculative doubts or his considered opinion. Non-gratuitous doubts are generated when an alternative hypothesis seems to pass a test. An alternative hypothesis typically generates predictions that are inconsistent with the predictions generated by the customary interpretation. So, once we entertain such a hypothesis, we are primed to notice things that we might not otherwise have noticed. If we thereby notice something that is incompatible with the DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 9 customary interpretation, this surprising successful prediction of the alternative hypothesis will typically undermine the prior natural tendency to believe the customary interpretation. What started as a gratuitous doubt that enabled us to choose whether or not to believe becomes a non-gratuitous doubt that may render us unable to believe the customary interpretation. However, although the explanatory success of the alternative hypothesis cannot simply be dismissed, it might, once attention switches to other things, be absent from consciousness, in which case we may fall back into believing the customary interpretation. Thus, non-gratuitous doubts, so long as they are kept in mind, typically render us unable to believe. They do not permit doxastic voluntarism. It might be objected that if one currently believes a specific proposition, one would be able to suspend belief in it only if one had doubts of a sufficient cogency. Non-gratuitous doubts are strong enough for this; but they are too strong to enable doxastic voluntarism because we are not able simply to dismiss them. On the other hand, gratuitous doubts may be too frivolous. For example, if I appear to see rain in apparently ordinary circumstances, can I really suspend belief in the proposition that it is raining, merely by considering that the appearance of rain might be due to a misguided crop-sprayer? Would I not simply regard that hypothesis as a joke and continue to believe that it is raining? This objection virtually concedes my point. If I take seriously the hypothesis that there is a crop-sprayer up there, then I suspend belief in the rain. But because that hypothesis has not passed a test I can simply dismiss it, simply put it out of my mind, which will leave me believing that it is raining. What would I have to do to take the hypothesis seriously? I might set out to test it, for example, by taking a look in the sky or by listening for an aeroplane engine. But such actions are neither necessary nor sufficient. They are not sufficient, because they could be undertaken by someone who does not take the hypothesis seriously, for instance, someone who just wants to refute the hypothesis to stop other people taking it seriously. They are not necessary because taking a hypothesis seriously just means taking it seriously, entertaining it as a serious possibility: it need not be manifested in any behaviour. Further, for me to be able to suspend belief in a proposition, it is not necessary that I have in mind a specific potential alternative explanation for why that proposition seems true. It is enough that I take seriously the proposition that there may be such an alternative explanation. This is a speculative gratuitous doubt, since a bare existential proposition is not falsifiable (Popper 2002, 47-50). Gratuitous doubts enable us to suspend belief because we can take them seriously; and they enable us to choose belief because we can dismiss them (refuse to take them seriously). It might be protested that I have missed the point of the objection. Let it be agreed that we can suspend belief if we can take seriously an untested alternative hypothesis or even the bare possibility that there is some alternative explanation. The question remains as to whether we can take these seriously. However, I showed in section 2 that we can. Anyone who is familiar with the way in which, in science, one highly successful explanation has been overturned by a better one, only to be overturned in its turn, can take seriously the prospect of an alternative to any empirical proposition. Anyone familiar with the history of paradoxes in mathematics and logic, and the rival systems for dealing with them, can take seriously the prospect of an alternative to any claimed a priori truth. Indeed, anyone familiar with Descartes’ sceptical doubts can take seriously the prospect of an alternative to any DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 10 proposition at all. But one does not need familiarity with this intellectual history in order to take seriously the possibility of an alternative to a currently accepted proposition (otherwise that history might never have happened): all one needs is to exercise one’s critical rationality. When Alston (1988, 270) considers how one might exercise non-basic immediate voluntary control over normal perceptual, introspective, or memory propositions, he lists a number of possibilities, including ‘I could rehearse some sceptical arguments.’ He dismisses the possibilities by saying that ‘unless I am a very unusual person, none of these will have the least effect.’ Three points should be noticed here. The first is that he concedes that rehearsing sceptical arguments would amount to non-basic immediate voluntary control over belief if it were successful in inhibiting belief. The second is that he is right that merely ‘rehearsing’ sceptical arguments will be ineffective, because we need to take the sceptical arguments seriously. The third is that his supposed counter-examples to doxastic voluntarism depend upon his refusal to take sceptical arguments seriously. Yet, not only is it the case that we can take sceptical arguments seriously, and thus entertain gratuitous doubts, it is also the case that a refusal to do so would be a serious obstacle to the growth of knowledge, as we saw in section 2. Gratuitous doubt, whether speculative or non-speculative, enables us to choose whether to believe or not to believe a proposition if we have a natural tendency to believe that proposition: it enables us to choose to yield (or not) to that inclination. We cannot choose to believe a proposition that we have no natural tendency to believe. As a consequence, we cannot choose to believe the negation of a proposition that we have a natural tendency to believe (except in the special case in which we also happen to have a different natural tendency to believe that negation): the ability to withhold belief is not the ability to believe the negation. Suspension of belief is doubt, not another belief. Alston’s specious claim that it is impossible to believe for the sake of a reward is thus mistaken. Of course we cannot simply choose to believe any proposition for the sake of a reward. We cannot choose to believe that the US is still a colony of Britain, because it seems less plausible than its negation, and we do not have a natural tendency to believe a proposition that seems implausible or unlikely. But, as we can choose to believe a proposition that we have a natural tendency to believe, then we can do so in order to obtain a reward. For example, suppose I have a rich friend who thinks that my persistent worry that I might be a brain in a vat is making it difficult for me to enjoy life. She offers me a fortune if I will choose to believe that the spatio-temporal world that I appear to inhabit, including my body, is real. I want the fortune so much that I agree to choose that belief. How do I do it? I stop entertaining doubts about it. That example is fanciful in that my friend would have no way of knowing whether I had stopped doubting: for all she knows, I might just be putting on an act. We might imagine that a brain-scanner has been developed that can check what we believe; but that might also seem fanciful. However, we have no need of fanciful examples. For maintaining doubt where we have a natural tendency to believe involves a constant effort. As soon as our concentration shifts elsewhere, we are liable to fall into believing what we have a natural inclination to believe. The only way to avoid this is by holding fast to one’s doubts, keeping them in mind through all of our activities. It seems unlikely that it is possible to do this for an extended period DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 11 (unless one is suffering some form of psychosis). But even to the extent to which it is possible, it is easier not to do it. It is easier to dismiss the doubts and carry on believing. But if we choose to believe for this reason, we are choosing to believe for a reward, namely, the reward of avoiding the effort of doubt. Yielding to belief can therefore be rational in that it makes efficient use of our scarce attention. The situation with belief has some similarity to that of laughing (Danto 1970, 46-50). We cannot just laugh as a basic act (though we can feign a laugh as a basic act). We can laugh only when we have an inclination to laugh, which is usually when we find something amusing. But even then we do not have to laugh: we can instead inhibit it. So, provided the option of inhibiting the laughter occurs to us, we can, as basic act, inhibit the laughter, and thus not laugh, or we can omit to inhibit the laughter, in which case our natural inclination to laugh will immediately bring about our laughing. In that situation, we have non-basic immediate voluntary control over our laughter. Further, provided we have an inclination to laugh, we can get ourselves into that situation simply by thinking about whether to inhibit our laughing. So, as long as we have an inclination to laugh, we can normally bring it about that we laugh in one uninterrupted activity, if we just voluntarily exert ourselves to do it, by thinking about whether to inhibit it and then omitting to inhibit it. Thus, where we have an inclination to laugh, we normally have non-basic immediate voluntary control over our laughter. Another similar case would be urination. We cannot simply urinate at will: we need a sufficiently full bladder, which gives us a natural inclination to urinate. Once we have that, we can choose either to urinate or not. Of course, beyond a certain point we lose control. Similarly in the case of laughter: sometimes we cannot help laughing, even if we try our utmost not to. There seem also to be circumstances in which we lose our ability to suspend our natural tendencies to believe. Psychopathological cases of delusion, obsession or compulsion involve impairment of rationality; and in extreme circumstances, such as terror, our critical faculties are overridden and sheer animal belief enslaves us. For example, if I see a panther, my fear of being eaten may render me, for the time being, unable to doubt that there is a panther there; though, once back in relative safety, I may reflect on whether it really was a panther that I saw. This account of doxastic voluntarism is consistent with recent research in cognitive science which distinguishes two reasoning processes in humans: system 1, which humans share with other animals, comprises processes which are rapid, parallel and automatic, only the final product of which is posted in consciousness; but system 2, which is distinctively human, is slow, sequential, volitional and of limited capacity, but permits abstract hypothetical thinking and enables us to override or inhibit the default responses emanating from system 1 (Evans 2003, 2008). 4. Williams Bernard Williams’ first argument (1973, 148) against doxastic voluntarism is connected with the characteristic of beliefs that they aim at truth. If I could acquire a belief at will, I could acquire it whether it was true or not; moreover I would know that I could acquire it whether it was true or not. If in full consciousness I could will to acquire a ‘belief’ irrespective of its truth, it is unclear that before the event I could seriously think of it as a belief, i.e., as something purporting to represent reality. At the very least, there must be a restriction on what is the case after the event; since I could not then, in full DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 12 consciousness, regard this as a belief of mine, i.e. something I take to be true, and also know that I acquired it at will. This argument does not tell against doxastic voluntarism. Indeed the preceding account of doxastic voluntarism has more or less endorsed each of Williams’ propositions, which I will re-word to try to make them clearer: • beliefs purport to represent reality (or aim at truth), • if we choose to believe a proposition, we do so irrespective of its truth, • while we are considering the choice of whether to believe, we cannot regard the proposition as one that we believe, • after the choice, we cannot regard the proposition as one that we believe if we recall to consciousness that we chose to believe it. Williams’ argument can appear to have force against doxastic voluntarism only if we fail to take account of the switch from a state of doubt to a state of belief. While a person may sometimes choose either to continue to doubt or to believe a proposition, he cannot believe it while continuing to doubt it. If, after the choice, he resumes his doubts, he then ceases to believe. But Williams’ final proposition is not quite right, because to recall the doubts is not necessarily to resume them, since we could recall the choice to dismiss the doubts without again taking the doubts seriously. In the same paper, Williams offers two other arguments against doxastic voluntarism. One of them has two premises: • if a person believes a proposition, she would be surprised to discover that the proposition is false; • if she believed a proposition at will, she would not be surprised to find that the proposition is false. Each of these premises seems false. First, education is often a process of discovering that propositions that one previously believed are rejected as false by scientists. As one goes through this process one may wonder what homely proposition will be next to find its way into the bin. One might consider some candidates but continue to believe them. Later, when one discovers that one of these is indeed rejected by science, one need not be surprised at all. Second, we saw in sections 2 and 3 that in the case of a proposition that seems more likely than any of its rivals we may suspend belief or choose to believe it. If we choose to believe it, we may be surprised to discover it is false, given that it seemed most likely. Williams’ other argument runs: • a person acquires an empirical belief that p only because the state of affairs that p generates in the person perceptions that lead to the belief; • if one could believe empirical matters at will, there would be no regular connection between the environment, the perceptions and the belief; • therefore, believing an empirical matter at will is impossible. In the light of our discussion one might be tempted to respond that, since we can choose to believe only things that we have a natural tendency to believe, the second premise is false, because in those cases there is a regular connection between the environment, perceptions and beliefs. It should however be clear that this temptation should be resisted; for there is no such regular connection. Williams’ first premise is false. First, as we noted in sub-section 2.1 and in section 3, our perceptions depend not only upon the external stimuli but also upon the interpretation supplied by our DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 13 prior theories, some of which are biologically ‘hard wired’ into our cognitive equipment but some of which vary with cultures and subcultures. Second, while empirical beliefs are those that can be tested against statements we take to report our observations, the content of empirical beliefs, especially in the sciences, is universal, going way beyond the content of any finite accumulation of observation statements (Popper 2002, 440-46). Third, many empirical beliefs contradict other empirical beliefs. These may be beliefs of the same person at different times, as with astronomers who gave up Kepler’s laws for Newton’s, or physicists who gave up Newton’s theory for Einstein’s. They may be beliefs of different persons at the same or different times, as illustrated by the myriad disputes during the progress of science (Popper 1983, 69-71). They may even be beliefs of the same person at the same time, as with someone who discovers a contradiction within his empirical beliefs but does not know how to resolve it (Raz 2010, 15). Assuming that the law of noncontradiction holds for these beliefs, at most one of the conflicting beliefs may be true; in which case many empirical beliefs are false (that is, it is often the case that not-p while people have the empirical belief that p). Indeed, evolutionary theory suggests that many of our empirical beliefs are false. A simple but false theory that gives approximately accurate results in the situations in which it is used will have all the practical advantages of a true but more complex theory but without requiring the greater cognitive resources and time needed to acquire, understand and apply the more complex theory. It is a commonplace of practical life that engineers, navigators, accountants and others often use an approximate ‘rule of thumb,’ instead of more exact theories or techniques, to economise on labour and time; and scientists at NASA use Newton’s theory for planning space explorations, despite the fact that Newton’s theory is false if relativity theory is true, because the calculations using relativity theory would be more complex but practically indistinguishable. From an evolutionary point of view, it is advantageous for a creature to have false but simple beliefs that are good enough to be useful in the particular environmental niche in which it has evolved. The resources saved by not building more powerful cognitive equipment can then be used instead to build stronger wings, faster legs, and so on. Genes which build creatures with such false but apt beliefs, or with a tendency to acquire them, would therefore become more numerous in the gene pool, other things being equal (for examples that seem to illustrate this point, see Dawkins 1989, 99-105 and passim; also see Evans 2003 and 2008 for a useful summary of, and detailed references to, the ‘heuristics and biases’ literature which relates specifically to human beings). This suggests that evolutionary theory is itself false; but that does not impugn its explanatory value or predictive utility. 5. Hieronymi Pamela Hieronymi (2006) argues that it is ‘conceptually impossible’ that we have basic voluntary control over our beliefs (she calls this ‘immediate’ voluntary control, but I will stick with Alston’s terminology to avoid confusion). She distinguishes between constitutive and extrinsic reasons for a belief. Constitutive reasons for believing a proposition bear on the truth of that proposition. If one finds such constitutive reasons convincing, one thereby believes the proposition; otherwise (if one finds them inconclusive) one cannot believe the proposition, though one may act as if one does. Either way, believing the proposition, or coming to DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 14 believe it, is not under one’s basic voluntary control. Extrinsic reasons count in favour of believing a proposition independently of whether the proposition is true; they bear on the question whether it would be good to believe the proposition. Even if one finds the extrinsic reasons convincing, that does not constitute one’s believing the proposition; something further is required for that. This something further may be under our voluntary control; but bringing it about that we believe a proposition is always done by doing something else; it cannot be done as a basic act. Thus, whether our reasons are constitutive or extrinsic, our believing or coming to believe is not under our direct voluntary control. Hieronymi seems to reject the possibility of intentionally believing or coming to believe a specific proposition for no reason, though she admits that she has no argument against it. Hieronymi’s conclusion, that we do not have basic voluntary control over belief, is compatible with my claim that we do have non-basic immediate voluntary control. In particular, nothing in her argument excludes believing, or coming to believe, for an extrinsic reason (such as to avoid the burden of continuing to doubt), by dismissing gratuitous doubts (she does not consider the possibility). We can, though, express the following scruples about her argument. First, it is normally open to us to find any constitutive reasons unconvincing, by raising doubts about them, that is, by taking seriously the possibility of an alternative. Conclusive is not something a reason can be. Second, if one finds the available constitutive reasons for a proposition unconvincing, one can still believe the proposition, so long as one has a natural tendency to believe it and one’s doubts about it are gratuitous. Third, intentionally believing or coming to believe a proposition for no reason is possible and may be commonplace. The students of Descartes who are at first captivated by his speculative doubts but later dismiss them do not, I imagine, dismiss them for the extrinsic reason that it is too burdensome to continue to doubt; and they cannot dismiss them for constitutive reasons (at least, so long as they have understood the arguments); rather, they dismiss them without considering reasons. It might be objected that the students do not intentionally re-acquire their suspended beliefs; rather, they intentionally dismiss the doubts and thereby bring about belief as an unintended consequence, or even as a double effect. But even if this were generally true, there might still be some students who intend to resume their old beliefs by dismissing their doubts about them. Fourth, I think that the part of Hieronymi’s argument (not summarised above) for her claim that direct voluntary control of belief is ‘conceptually impossible,’ rather than something ordinary humans are unable to do, is unsuccessful for various reasons, though I do not have space here to explain that. 6. Conclusion The arguments against doxastic voluntarism fail. They may seem plausible because we cannot choose to believe an arbitrary proposition. But in the case of propositions that we have a natural tendency to believe, we normally have the power to suspend belief or to yield to it. We have a natural tendency to believe propositions that have been absorbed or inculcated via our participation in a culture or subculture, including those that are vouched for by figures we regard as authorities. We also have a natural tendency to believe propositions that are strongly suggested to us by our sensory experiences, memories and introspections; but what propositions these happen to be depends upon our biologically inherited framework of interpretation as DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 15 well as on the more explicit theories in terms of which we customarily interpret the world, some of which are acquired as part of our early socialisation, others through education and training, and some may be our own creations. But so long as we retain our critical rationality, we are able to doubt the propositions we have a natural tendency to believe; and, so long as our doubt remains gratuitous, we can instead choose to believe them. If we do choose to believe, this may be done for ‘extrinsic’ reasons, such as, for example, that it avoids the effort of doubt. The motivation for the opposition to doxastic voluntarism seems to be a fear that doubt destroys knowledge, so that, if we are to safeguard knowledge, we must have beliefs that are indubitable, either inherently or given specific evidence. The propositions we have a natural tendency to believe are the most plausible candidates for indubitability; so these are decreed indubitable, so that our belief in them is not under our immediate voluntary control (at least, once we attend to them or attend to the evidence for them). As for propositions that we do not have a natural tendency to believe, we do not seem psychologically able to choose to believe those. It then follows that in no case can we choose to believe, so doxastic voluntarism is false. However, the strategy not only fails, it is also inept, because it is through doubt, and particularly through doubting propositions we have a natural tendency to believe, that scientific discoveries are made. So far is it from being the case that knowledge depends on the denial of doubt, that it is doubt that is the beginning of knowledge. As Kant (1787/1933, B xiii) put it: Reason…must approach nature in order to be taught by it. It must not, however, do so in the character of a pupil who listens to everything that the teacher chooses to say, but of an appointed judge who compels the witnesses to answer questions which he has himself formulated. Knowledge – at least, rational knowledge – depends not on unquestioning acceptance, but on critical and sceptical enquiry; it involves standing back from our beliefs, or accepted views, and evaluating them (Popper 1966, 380-81; 1982, 81-85; see also Kant 1785/1948, 448). Thus, the arguments against doxastic voluntarism are not only faulty, they are futile. Rational knowledge, as opposed to mere animal belief, is inseparable from choice. References Alston, William. 1988. ‘The Deontological Conception of Epistemic Justification.’ In James Toberlin ed., Philosophical Perspectives, Volume 2, 257-99. Atascadero: Ridgeview. Berkeley, George. 1710 /1910. A Treatise Concerning the Principles of Human Knowledge in A New Theory of Vision and Other Writings, 87-195. London: Dent Dutton. Danto, Arthur. 1970. ‘Basic Actions.’ In Alan White, ed., The Philosophy of Action, corrected edition, 43-58. Oxford: Oxford University Press. Dawkins, Richard. 1989. The Selfish Gene, second edition. Oxford: Oxford University Press. Descartes, Rene. 1641/1931. Meditations on First Philosophy in Philosophical Works, Volume 1, trans. Elizabeth Haldane and G R T Ross, 131-199. Cambridge: Cambridge University Press. Duhem, Pierre. 1914/1954. The Aim and Structure of Physical Theory, second edition, tr. P. P. Wiener. Princeton: Princeton University Press. Evans, Jonathan. 2003. ‘In Two Minds: Dual-Process Accounts of Reasoning.’ Trends in Cognitive Sciences, 7 (10): 454-59. DOXASTIC VOLUNTARISM: A SCEPTICAL DEFENCE 16 Evans, Jonathan. 2008. ‘Dual-Processing Accounts of Reasoning, Judgment, and Social Cognition.’ Annual Review of Psychology, 59: 255-78. Feyerabend, Paul. 1975. Against Method. London: New Left Books. Funkhouser, Eric. 2003. ‘Willing Belief and the Norm of Truth.’ Philosophical Studies, 115 (2): 179-95. Galileo, Galilei. 1632/1953. Dialogue Concerning the Two Chief World Systems, tr. Stillman Drake. Berkeley: University of California Press. Ginet, Carl. 2001. ‘Deciding to Believe.’ In Knowledge, Truth, and Duty, ed. M. Steup, 63-76. Oxford: Oxford University Press. Heller, Mark. 2000. ‘Hobartian Voluntarism: Grounding a Deontological Conception of Epistemological Justification.’ Pacific Philosophical Quarterly, 81: 130–41. Hieronymi, Pamela. 2006. ‘Controlling Attitudes.’ Pacific Philosophical Quarterly, 87 (1): 45– 74. Kant, Immanuel. 1785/1948. Groundwork of the Metaphysic of Morals, tr. H. J. Paton in his The Moral Law, 53-123: London: Hutchinson. Kant, Immanuel. 1787/1933. Critique of Pure Reason, second edition, tr. N. Kemp Smith, corrected edition. London: Macmillan Kuhn, Thomas. 1957. The Copernican Revolution. Cambridge, Mass: Harvard University Press. Kuhn, Thomas. 1970. The Structure of Scientific Revolutions, second enlarged edition. Chicago: University of Chicago Press. Kuhn, Thomas. 1977. ‘The Essential Tension.’ In his The Essential Tension, 225-39. Chicago: University of Chicago Press. Nickel, Philip. 2010. ‘Voluntary Belief on a Reasonable Basis.’ Philosophy and Phenomenological Research, 81 (2): 312-34. Nisbett, R.E. & Wilson, T. D. 1977. ‘Telling More Than We Can Know.’ Psychological Review 84/3: 231-59. Popper, Karl. 1966. The Open Society and its Enemies, Vol. 2, fifth (revised) edition. London: Routledge and Kegan Paul Ltd Popper, Karl. 1982. The Open Universe. London: Routledge. Popper, Karl. 1983. Realism and the Aim of Science, London: Routledge. Popper, Karl. 2002. The Logic of Scientific Discovery. London: Routledge. Popper, Karl & John Eccles. 1977. The Self and its Brain. London: Springer-Verlag. Priest, Graham. 2004. ‘What’s So Bad About Contradictions?’ In The Law of NonContradiction, ed. Graham Priest, J. C. Beal and Bradley Amour-Garb, 23-38. Oxford: Clarendon Press. Priest, Graham & Neil Thomason. 2007. ‘60% Proof.’ Australasian Journal of Logic, 5: 89100. Raz, Joseph. 2010. ‘Reason, Reasons and Normativity.’ In R. Shafer-Landau (ed.), Oxford Studies in Metaethics, Volume 5, 5-23. Oxford: Oxford University Press. Schwitzgebel, Eric. 2008. ‘The Unreliability of Naive Introspection.’ Philosophical Review 117 (2): 245-273. Tavris, Carol, & Eliot Aronson. 2007. Mistakes Were Made (But Not By Me). New York: Harcourt. Watkins, John. 1984, Science and Scepticism. Princeton: Princeton University Press. Weatherson, Brian. 2008. ‘Deontology and Descartes’ Demon. ‘ Journal of Philosophy, 105 (9): 540-569. Williams, Bernard. 1973. ‘Deciding to Believe.’ In his Problems of the Self, 136-51. Cambridge: Cambridge University Press.