Weighing the aim of belief
- First Online:
- Cite this article as:
- Steglich-Petersen, A. Philos Stud (2009) 145: 395. doi:10.1007/s11098-008-9239-8
- 223 Downloads
The theory of belief, according to which believing that p essentially involves having as an aim or purpose to believe that p truly, has recently been criticised on the grounds that the putative aim of belief does not interact with the wider aims of believers in the ways we should expect of genuine aims. I argue that this objection to the aim theory fails. When we consider a wider range of deliberative contexts concerning beliefs, it becomes obvious that the aim of belief can interact with and be weighed against the wider aims of agents in the ways required for it to be a genuine aim.
KeywordsEpistemic normativity Aim of belief Deliberation
Many authors have been attracted to a theory of belief, according to which believing that p essentially involves having as an aim to believe p truly.1 The explanatory hopes motivating such a theory are well known. First of all, it might explain the standard of correctness for belief. Many authors think that beliefs are governed by a standard of correctness saying that believing p is correct only if p is true. While some have thought this to express a basic feature of belief, not standing in need of further explanation, aim-theorists hope to explain this standard in terms of an aim of belief. In short, believing p is correct only if p is true because only true beliefs achieve the aim involved with believing. Secondly, aim theorists hope to explain the norms of epistemic justification governing belief. Beliefs ought not merely to be true—they also ought to be formed in ways that ensure or make it likely that they are true. A natural explanation of such epistemic norms is that following them promotes the aim of believing truly. Epistemic norms thus turn out to be instrumental norms, deriving their normative force from their ability to guide us to achieve our aims. This would also explain the peculiar authority epistemic norms have over other kinds of considerations in forming beliefs, and the way these norms function in motivating beliefs. If it is a conceptual truth about believing, that someone believing p aims at believing p truly, it is simply impossible to consciously and deliberately believe that p despite evidence to the contrary. The aim theory promises, in other words, a simple, unified, and prima facie unproblematic explanation of epistemic normativity.
Despite its obvious explanatory appeal, however, the aim theory of epistemic normativity has recently fallen into widespread disfavour. For many of the most forceful critics, including Shah (2003), Kelly (2003), and in particular Owens (2003), there is a single heavy-weighing consideration counting against the aim theory: if our adherence to epistemic norms is explained by some aim of ours, we should expect that we, on at least some occasions, when deciding how to adhere to the norms generated by that aim, weigh the aim of belief against other aims. But epistemic deliberation is not like that, they say. When we deliberate about what to believe, we focus exclusively on epistemic considerations bearing on the truth of the proposition being considered for belief.
By contrast, deliberation over aim-motivated activities typically involves weighing of aims. Suppose, for example, that I have the aim of becoming more fit, and I am considering what to do in order to become fit. It would be reasonable to expect, then, that although my deliberation is guided by the aim of fitness, other aims influence what I in the end decide to do. Perhaps the best way of becoming fit is to spend at least 3 hours at the gym every day, and eat only oatmeal. But I may have other things to do, and have a strong aversion to oatmeal, so the best way of becoming fit may prevent me from pursuing other aims of mine (i.e. spending my time doing something interesting and eating tasty food). So in the end I am likely to go for a compromise, striking a balance between the aim of fitness and my other aims. The crucial point is that this type of deliberation where several aims are weighed against each other in no way compromises the fact that I have the aim of fitness, nor does it compromise the fact that my going to the gym at some more moderate frequency might be explained by that aim. In fact, if my deliberation did not exhibit this feature of weighing different aims against each other, it would be difficult to seriously think of my deliberation as deliberation over how to achieve my aims. Aims, it seems, are essentially the kind of things that can conflict and be weighed against each other. If a state or consideration cannot enter into this kind of weighing, it is difficult to seriously think of that state or consideration as an ‘aim’. If an act is not the outcome of aims being weighed against each other, it is difficult to think of that act as aim-directed.
But the supposed aim of belief seems to lack this exact property: when I deliberate over whether to believe that the earth is flat, only evidence bearing on the truth or falsity of that proposition seems to matter. It doesn’t matter the least bit to what I ought to believe, what other, non-epistemic aims I may or may not have, what belief I would find most pleasant, etc.—only considerations bearing on the truth of the proposition matter, to the exclusion of other considerations. It does not make sense to weigh aims when deciding what to believe. So although there might be a trivial, metaphorical sense in which beliefs ‘aim at truth’, namely that beliefs are correct only if true, no serious explanatory meaning can be ascribed to that claim. Or so the objection goes.
Call this objection to the aim theory the ‘exclusivity objection’, because it is the exclusive relevance of truth in deliberation over belief that seems to prevent us from seeing beliefs and belief-formation as being ‘aimed’ at truth in any interesting sense. In the remaining part of the paper, I defend the aim theory against this important and influential complaint. My main line of argument will be that while the critics are correct in claiming that any genuine aim must be able to interact with the wider aims of agents, they are wrong in denying that the aim of belief can interact with other aims in the required ways. The critics have focused on a particular context of deliberation over belief, namely that where one considers whether to believe a particular proposition or not. But when we consider a wider range of deliberative contexts concerning beliefs, in particular those where one considers whether to adopt or discard the aim of belief, it becomes obvious that the aim of belief can interact and be weighed against the wider aims of agents in the ways required for it to be a genuine aim. The critics have been looking for interaction of aims in the wrong deliberative context. Making this point will involve some development of the aim theory, but first I turn to a more detailed analysis of the exclusivity objection, as it appears in Owens (2003).
1 Owens on the aim of belief
Owens’ argument for the claim that the aim of belief does not interact with the wider aims of believers, and therefore cannot be understood as a genuine aim, is centred around a detailed comparison between believing and a slightly different mental act, guessing, in regard to two aspects of those attitudes: (i) the authority of epistemic considerations in deciding the rationality of the attitude, and (ii) the kind of control we are able to exercise when deciding what attitude to adopt. According to Owens, guessing, as opposed to believing, aims at truth. This is reflected in the way epistemic considerations enter into what it is rational for us to guess. The truth aim of guessing interacts with the wider aims of the guesser, so epistemic considerations do not exert authority when deciding what to guess, or when deciding how strong evidence to base one’s guesses on. When deciding what it is rational for one to believe, on the other hand, epistemic considerations exert authority to the exclusion of other sorts of considerations, making it impossible to see the rationality of belief as a function of the weighing of several aims. This, according to Owens, should make us think that the rationality of belief is not aim dependent in any interesting sense at all. This is also reflected in the kind of control we have in forming the respective attitudes. While it makes sense to say that it is under our control to make a guess when we seem sufficiently likely to achieve the aim of guessing, where ‘sufficient’ is determined by our wider aims, Owens does not think it makes sense to regard believing as controlled in that sense.
φ-ing that p aims at the truth if and only if someone who φs that p does so with the purpose of φ-ing that p only if p is true. (Owens 2003, p. 289)
The truth-aim is formulated negatively here to avoid an implausible commitment to seek out all truths, and is taken to apply to beliefs individually. It is also taken to be an aim that one can pursue more or less aptly. Someone can pursue the aim unsuccessfully, or while being misguided about evidence etc., without thereby ceasing to count as having the truth-aim. The claim is now that while guessing clearly aims at truth in this sense, believing does not.
If this talk of purpose is to be more than an empty metaphor, the notion of purpose invoked must be one that does explanatory work for us outside the domain of epistemic norms. And, surely, it is our ordinary notion of purpose which the truth-aim theorist means to be employing. Now such purposes interact with each other in certain familiar ways, so if a subject really does form a belief with the purpose of forming it only if it is true, his pursuit of that goal should be constrained by his other goals and objectives in (something like) the usual fashion. That is just what happens in the case of guessing which is why we are fully justified in treating guessing as a purposive activity. But we failed in our attempts to treat the ‘purpose’ allegedly shared by all believers as something that interacts with their other objectives in familiar ways. So what is left of the idea that they share a purpose? The truth aim theorist can’t evade this problem simply by declining to explain the differences between believing and guessing, since what these differences suggest is that our notion of a purpose does not apply to belief at all. (Owens 2003, pp. 295–296)
The point can also be put as one about control. According to Owens, we have no grounds for thinking that believing is purposive, unless we by reflecting on how best to achieve the aim of belief can move ourselves to form beliefs. Many authors have attempted to account for the fact that we cannot exercise control in forming beliefs in response to non-evidential considerations in terms of a truth-aim. Just like one cannot at once be said to attempt to score a goal, and attempt to shoot past the goal, one cannot be said to form a belief in a way that one knows will not achieve the aim involved with believing. But Owens thinks that the aim theorist is committed to explaining normal cases of belief formation in terms of reflection over how to achieve that aim as well. And according to Owens, this requires that it makes sense to understand the believer as acting in order to achieve that aim in a way that integrates with his wider aims and purposes. Again, guessing is the apt comparison. I cannot make myself guess something I know to be false, because guessing aims at getting it right. I cannot make myself guess that I am 7 feet tall, for instance, because I know that I am not. But it does make sense to understand my control over guessing as a matter of weighing the aim of guessing against other aims. In a quiz situation, where I am presented with more and more evidence bearing on the question, but where the money prize for getting it right decreases the longer I hold back my guess, I can weigh the aim of guessing—getting it right—against the aim of getting as much money as possible, and decide to plunge for a guess at a time with less than perfect evidence, but a high reward for getting it right. According to Owens, such control based on considerations about when I am likely to achieve my aims, is not possible in the case of belief.
Although Owens has made the exclusivity objection most forcefully, other authors, in recent years most notably Shah (2003) and Kelly (2003), have put forward similar arguments, moving from the fact that relevant considerations in deliberation over belief are restricted to truth-relevant considerations, to the claim that beliefs cannot be seen as truth-aiming in anything but a metaphorical sense. Shah’s main point is that deliberation over what to believe is transparent to the truth of the target proposition, such that whenever one is considering whether to believe that p, one must decide that question exclusively by deciding whether p is true (or supported by evidence). Shah thinks that such transparency is incompatible with the aim account insofar as this account is committed to there being a further consideration necessary in order to decide the rationality of believing p, namely whether believing some truth would further one’s aims.2 Kelly argues that the aim account of epistemic normativity fails, since epistemic norms give rise to non-hypothetical requirements. On the aim account, epistemic requirements are hypothetical in the sense of depending on the contingent aims of agents. But according to Kelly, epistemic requirements are not like that. If I want to convince someone that he should believe that p, I don’t have to first convince him that he should have some aim that is best served by believing that p. All I have to do is to provide evidence that p is true. Kelly notes that one way to answer this objection would be to make the truth-aim constitutive or defining of belief in the fashion under consideration in this paper, such that anyone believing p necessarily does so with the aim of thereby believing a truth. But just like Owens, Kelly thinks that this in effect would be to abandon the central appeal of the aim account, which was to make epistemic norms depend upon which aims an agent should or would prefer to have realised (2003, p. 632).
While Shah and Kelly’s objections to the aim account are not identical to Owens’ objection, they are thus founded upon the same fundamental worry: if the correctness and rationality of beliefs are to be explained by reference to an aim of the believer, why does the question of which aims the agent should adopt and pursue seem irrelevant in deliberation over whether to believe some proposition? For Owens, the objection takes a more specific and interesting form: if the rationality beliefs are to be explained by an aim of the believer, why are the wider aims of the believer irrelevant for what he should believe? How can the aim of belief be isolated in deliberation from the other, sometimes more pressing aims and purposes of the believer, and still be properly counted as an aim? In the remaining parts of the paper, I explore possible ways of resisting this objection. Although my arguments will be directed primarily at Owens’ version of the objection, they should be taken as bearing on the more general class of arguments sketched above as well.
2 Weighing the aim of belief
There are two main strategies that the aim theorist might adopt in defending the aim theory against Owens’ objection. He can either deny that the aim of believers fails to interact in the ways necessary to be thought of as a genuine aim. Or he can deny that the ability to interact in those ways is indeed necessary for something to be thought of as a genuine aim. In the following, I will pursue the first of these strategies. I will thus concede the very general point that the kind states described multiply as ‘aims’, ‘goals’, and ‘purposes’ necessarily are such that they can interact with other such states, and can be weighed against each other in deliberation, in the ways described by Owens. But contrary to what Owens argue, I will claim that the aim involved with believing does interact with the wider aims of believers in ways sufficient for the aim-theory to be genuinely explanatory.
To begin with, we need a firmer grip on the general claim. What sort of interaction with, and weighing against other aims in deliberation, must genuine aims be able to form part of? In order to answer this question, it will be useful to survey the different ways in which aims might interact in deliberation and action.
The most straightforward case is that in which two or more aims are mutually compatible, such that they can all be pursued either independently of each other, or in the same act. Trivial instances of this occur when the aims are simply irrelevant to each other. If the achievement of one aim in no way infringes or detracts from the feasibility of achieving another distinct aim, those aims can be said to be practically irrelevant for each other in deliberation. Since in theory the achievement of any aim comes at a cost of scarce resources, such situations may seem rare or even impossible, but it nonetheless seem very common for aims to at least appear completely irrelevant to one another in deliberation. For example, my aim of calling my mother today and my aim of having fish for dinner tonight seem completely irrelevant to one another in deliberation, and they don’t interact in any way whatsoever. One might speculate that the aim of believers is systematically irrelevant to any other aims in this way, but as I shall argue later, there is a more straightforward sense in which the aim involved with believing interact with other aims.
A more interesting case of interaction between mutually compatible aims is that in which the aims interact in the way they are each pursued. For example, my aim of clean dishes is compatible with my aim of minimising environmental impact insofar as there are ways of washing the dishes (albeit, perhaps, less effectively), with little or no environmental impact, for instance by refraining from using hot water and detergent. The resulting action may be a suboptimal way of pursuing the aims when taken in isolation, but it nevertheless makes sense to say that the action takes both aims into account. This is also the case in Owens’ example of guessing. The aim involved with guessing is compatible with the aim of winning a cash prize, insofar as the prospect of such a prize can motivate plunging for a guess with less than ideal evidential support for that guess. There is a possible action serving both aims, even if the action sometimes, as in the guessing case, is a less than ideal way of pursuing each aim in isolation. Owens seems correct in claiming that the aim of believers cannot interact with other aims in this specific sense. But there are other possible ways of interaction.
Particularly important are the situations in which the pursuits of two or more of one’s aims are mutually incompatible. This too is a familiar situation. My aim of spending the summer perfecting my golf swing is incompatible with my aim of spending the summer catching up on my research duties. I cannot pursue both, so I have to choose which of the aims is more important to me. This need not be because the aims considered as types are incompatible. In the case at hand, there is no necessary inconsistency in pursuing both aims. But as a contingent matter of fact, what it takes to achieve those two aims might turn out to be mutually incompatible actions. I cannot perfect my golf swing and see to my research duties at the same time, although the fortunate few talents might be able to. This does not mean that there cannot be weighing or interaction of aims—I weigh the aims in order to choose which aim to pursue and which to discard; it merely means that the resulting action cannot be seen as an attempt to combine the pursuit of those two aims—one of them has to go.
This type of weighing, i.e. weighing of incompatible aims resulting in discarding one or more of them rather than combining them in action, seems sufficient to satisfy what Owens has in mind when demanding that aims must be weighable against each other. While it is reasonable to claim that it is essential to something being an aim that it can meaningfully be said to interact with and be weighed against other aims, we should not exclude the possibility that sometimes, due to contingent de facto incompatibility, such weighing is of the kind that results in discarding aims rather than combining them in action. Weighing which of one’s aims is more important is a fundamental and important element in practical deliberation. If a state or consideration is systematically recalcitrant to even this form of interaction with other considerations, let us grant that Owens is right in claiming that it cannot be a genuine aim. But we should at the same time grant that ability for such kind of interaction is sufficient to pass the test.
My hypothesis is that while the aim of belief uniformly fails in being capable of the kind of weighing against other aims that results in a decision that combines the pursuit of both aims at once, the aim of belief is capable of the kind of weighing that results in discarding either of the aims because of mutual incompatibility. I will argue that this is in fact a fairly common occurrence. If this is the case, we have an explanation of the exclusive relevance of truth that is compatible with a non-trivial version of the aim theory. No doubt, the most common reason why the aim of belief fails to interact with other aims in deliberation is that they are practically irrelevant to most other aims in the way defined above, but cases of mutual incompatibility resulting in discarding either the aim of belief or some competing aim provide a clearer case for our purposes, so those are the cases I will focus on.
In order to make this case, we need to consider a wider range of doxastic deliberative contexts. For Owens, the main focus is on a very particular context of deliberation in regard to belief, namely deliberation over whether to believe a particular proposition p or to believe its negation. Owens’ main observation is that deciding such deliberation over whether to believe p or not-p cannot rationally be seen as a matter of deciding how to best achieve one’s aims, because no aim of the deliberator apart from the aim of truth can be relevant. Shah’s analogous observation is such deliberation is transparent to the question of truth, and thus as a matter of psychological necessity excludes consideration of whether adopting the belief in question would satisfy the wider aims of the believer. As far as the restriction to truth-relevant considerations go, this seems correct. When we focus narrowly on conscious deliberation over belief in a particular proposition or its negation, our deliberation does seem constrained to truth-relevant considerations in the way observed by Owens.
This is exactly what we should expect on the aims account since the content of the aim of belief restricts the conditions under which the aim is successfully pursued to those in which the proposition believed is true.3 In forming a belief in a particular proposition, one aims at accepting the proposition in question if and only if that proposition is true. We can think of other aims that are restricting in the same fashion. I might form the aim of trying out a new restaurant if and only if it receives good reviews. When subsequently deciding whether to try the restaurant or not, only one consideration will be relevant as far as achieving the aim goes, namely whether or not it received good reviews. But it would be a mistake to conclude from the fact that only one consideration is relevant to the successful pursuit of these restrictive aims, that they are isolated from one’s wider aims in deliberation. We can see this by considering other and less specific doxastic deliberative contexts where it is entirely rational for an agent to weigh the aim of belief against the agent’s wider aims. In particular, when one considers whether to take up the aim of forming a belief about some particular topic or proposition, it seems entirely relevant to weigh the value of adopting and carrying out that aim against the value of carrying out other, non-epistemic aims that the epistemic one is incompatible with. Similarly, it is often rational to give up the aim of belief in regard to some proposition, and thus give up forming beliefs about that proposition, when the consequences of carrying out that aim conflict with one’s other aims.
Consider first deliberation over whether to take up the aim of belief. Suppose that you are deliberating about whether to take up the aim of forming a belief about some proposition or subject matter.4 In order to do so, you must of course think that it is possible for you to form such a belief. For some propositions or subject matters, this possibility might be ruled out in advance. It is not straightforwardly possible, for instance, to form a belief about whether Cesar woke up on his third birthday lying on his belly. For many other propositions, forming a belief about them might be practically feasible, but only at a certain cost. Forming a belief about whether global warming is caused by human activities, for instance, is clearly possible, but requires dedicating time and resources to attend to the evidence—resources that could be dedicated to the pursuit of alternative aims. In cases such as these, it can be entirely rational to decide not to take up the aim of forming a belief about the subject matter in question, and instead dedicate resources to pursuing other aims deemed more significant.
The next case is that in which one gives up an already adopted aim of forming a belief about some proposition or subject matter. Sometimes, giving up the aim of belief can be the result of purely truth-relevant considerations. After carefully considering the available evidence as to whether p is true, I might decide that it is inconclusive and thus come to suspend belief as to whether p. This need not involve giving up the aim of, eventually, forming the belief, but it might. More interesting for our purposes are the cases where the aim is given up because one comes to realise that achieving the aim would be incompatible with, or detract from the feasibility of some other aim. Again, the typical case will be that where one realises that the costs of achieving the aim of forming the relevant belief are such that one would be better off pursuing some other aim. But it seems equally common to give up the aim of forming a belief because the anticipated consequences of doing so are incompatible the pursuit of other aims. A teacher might set out to find out which of her pupils was responsible for breaking a window, but coming to realise that she would then have to face the unpleasant task of scolding the guilty pupil, she might decide that she is better off not forming the belief at all. This is not an instance of forming a belief for practical reasons, such as the unpleasantness of holding a particular belief, since in this case, no belief is formed at all. Consciously forming a belief for practical, non-evidential reasons is ruled out by the phenomenon pointed out by Owens and Shah. But it is nevertheless an instance of weighing the aim of belief against practical aims, and deciding in favour of the practical one.
These examples presuppose that one can have the aim of belief in regard to some proposition or topic, without yet having formed a belief about that proposition or topic, i.e. that the aim one adopts when deciding to find out about something is the very aim that makes the attitude one subsequently forms in regard to that proposition into a belief. But on reflection, that is exactly what we should expect. I can adopt the aim of washing the dishes prior to actually doing something with that aim (i.e. washing the dishes), and give up the aim without ever acting on it; yet, when I do wash the dishes, I do it with that very aim. Similarly for believing, I might adopt the aim of forming a belief about some proposition, without actually forming that belief, and even give up the aim without ever having formed the belief; yet, when I do form the relevant belief, I do it with the aim originally adopted. A clearer way of stating this relationship becomes available if we conceive of beliefs as a kind of acceptance of a proposition, namely an acceptance of p with the aim of thereby accepting a truth.5 On this account, one has the aim of forming a belief about some proposition p if one aims at accepting p if and only if p is true. An acceptance of p is a belief if it is formed and regulated in an attempt to comply with that aim. This makes obvious the identity between the aim one has in aiming at forming a belief, and the aim that makes the subsequently formed belief into a belief. Of course, one might have other, extrinsic aims in forming one’s beliefs, which are not identical to the intrinsic aim defining those beliefs.6 For example, one might have the aim of only forming beliefs about subjects that are of practical importance to oneself. This is an aim one may or may not have in forming beliefs, which is different from the intrinsic truth-aim one necessarily has in believing a proposition. But the possibility of such extrinsic aims does not prevent that one can have the aim intrinsic to belief in regard to some proposition prior to forming a belief about that proposition. If this identity can be granted, and the above examples of weighing the aim of belief against wider practical aims in deliberation over whether to take up or discard the aim of belief are compelling, then the aim of belief is weighable and capable of interaction with other aims. Even if there are a number of other objections to the teleological account of belief, this would answer at least one influential worry.
3 Reasons for believing versus reasons for aiming
It is important to realise that accepting the above possibilities of weighing the aim of belief against one’s wider pragmatic aims does not commit one to accept what has been called ‘pragmatic’ or ‘state given’ reasons for belief, or to accept that such pragmatic reasons can be weighed against and compete with evidential reasons.7 Pragmatic reasons for belief are reasons for believing a particular proposition on account of desirable features or consequences of having that belief. For example, the pleasantness of believing that we have a soul which continues to live after the death of our bodies might be held to count in favour of holding that belief, even if the evidence speaks against it. It remains extremely controversial whether we ought to regard such features or consequences of holding a belief as genuine reasons for having the belief. The perhaps most influential argument against counting such practical considerations as reasons for belief is that we are unable to respond directly to such considerations in deliberation over belief.8 But part of the appeal of the aims account of belief is that it explains this exclusive focus on truth whenever one deliberates over whether to believe a particular proposition. In short, when deliberating over whether to believe a particular proposition, we focus exclusively on considerations that are relevant to the truth of that proposition, since only such considerations are relevant to whether the aim of believing the proposition if and only if it is true will be satisfied.9 None of this is affected by the possibility of weighing the costs and benefits of pursuing the aim of belief against the costs and benefits of pursuing other aims, in deliberation over whether to adopt or discard the aim of belief. And if, as I have argued, such weighing is possible, Owens’ objection against the teleological account of belief fails.
I provide a more detailed answer to Shah’s version of the exclusivity objection in Steglich-Petersen (2006).
For more on the teleological explanation of the exclusive focus on truth in deliberation over belief, see Steglich-Petersen (2006).
If ‘taking up the aim of belief about p’ sounds artificial, replace it with the equivalent, but more familiar sounding and straightforwardly aim-describing ‘deciding to find out about p’, ‘deciding to form an opinion as to whether p’, or ‘deciding to arrive at a conclusion as to whether p’.
An interesting account of how pragmatic reasons for belief might be weighed against evidential reasons has recently been suggested by Reisner (2008).
Shah and Velleman (2005) argue that the teleological account is unable to explain the exclusive focus on truth in deliberation of belief, in a way that is compatible with the fact that sub-personal processes of belief formation that are not truth-guided. However, as argued I argue in Steglich-Petersen (2006), this is no problem for the teleological account.
I am grateful to Jane Heal, Nicolas Espinoza, Veli Mitova, Johanna Seibt, Carl Erik Kuhl, Katrine Krause-Jensen, Raffaele Rodogno, and an anonymous referee for valuable criticisms and suggestions.