The normative ground of the evidential ought

The normative ground of the evidential ought Anne Meylan, University of Zurich* 1. The evidential ought: instrumentalism vs. intrinsicalism. Besides being subject to certain obligations or permissions to act, we are also subject to certain duties and permissions to believe. More precisely, many philosophers have defended the view that we are subject to the following evidential ought: One ought to believe in accordance with one's evidence.1 Take the case of a teenager observing a mother carrying her buggy on the metro staircase in clear daylight. The young man is not only compelled to help her. According to evidentialists—viz. the upholders of the evidentialist ought—he also ought to believe a certain number of things given the evidence he has, e.g. that there is a woman in the staircase, that she is carrying a buggy and not a bunch of flowers, etc. 2 Although evidentialists agree on this, a more fundamental question keeps dividing them. It concerns the normative ground of the evidential ought: in virtue of what ought one to believe in * Many thanks to the editors of this volume for their helpful comments and suggestions on an earlier version of this paper. 1 The contemporary literature contains many distinct formulations of the evidentialist ought. Steglich-Petersen (2018), for instance, favours the following: One ought to believe that p only if one has adequate evidence that p. I do not discuss these various different formulations since they do not matter for my purpose in this paper. I shall also ignore the important question whether the possession of adequate evidence is sufficient to give rise to duties to believe or whether it is only sufficient for permissions. 2 Upholders of evidentialism are numerous. See e.g. Adler 2002; Conee and Feldman 2004; Kelly 2002; Nolfi 2018; Shah 2006; Skorupski 2011; Vahid 2010; Way 2016. 1 accordance with the evidence? Or, to phrase the same question differently: from where does the evidential ought derive its normative force? The philosophical literature contains two answers to this question, which correspond to two mutually exclusive conceptions of the normative ground of the evidential ought: instrumentalism, and what is often called intrinsicalism. Here are recent descriptions of these two conceptions. "Instrumentalism says that there is a reason to believe in accord with the evidence because doing so is an instrumentally rational way of achieving the ends, goals, or interests one has... Intrinsicalism disagrees. According to intrinsicalists, the reason there is to believe in accord with the evidence is independent from any relationship doing so bears to one’s ends, goals, or interests. Instead, says intrinsicalism, it’s a brutely epistemic normative truth that there is a reason to believe in accord with the evidence." Sharadin, 2018, 3792. What is it about evidence that explains the existence of reasons—such as they are—to believe in accordance with it? Two rival explanations dominate the literature. The first explanation is, at bottom, practical. It is that there is reason to believe in accordance with one's evidence because this an excellent means of fulfilling the goals that one has or should have. I'll call this instrumentalism. The second explanation is not instrumentalist. It states that there is reason to believe in accordance with one's evidence in virtue of a brutely epistemic normative truth relating belief to evidence (or to some other epistemic property such as truth, or epistemic rationality). I'll call this view intrinsicalism. Cowie 2014, 4003–4004. To recapitulate, instrumentalism3 is the view that the normative force of the evidential ought is grounded in the epistemic purposes of the believer, for instance, in the aim of achieving knowledge 3 Instrumentalism has been extensively defended by Steglich-Petersen. See Steglich-Petersen 2006a, 2006b, 2009, 2011, 2013, 2018. 2 or of getting a true belief about a particular topic. According to instrumentalism, it is in virtue of our epistemic goals that one ought to believe in accordance with one's evidence (the assumption here is that beliefs that satisfy the evidential ought contribute to the achievement of our epistemic goals).4 In contrast, intrinsicalism5 denies the instrumentalist claim that the evidential ought derives its normative force from the fact that its satisfaction contributes to achieving our epistemic purposes. What is the normative ground of the evidential ought, according to intrinsicalism, then? As we just saw, the usual answer is that: "There is a brute epistemic value in believing in accordance with one's evidence". Cowie, 2014, 4005. But what does this really mean? Intrinsicalists' negative claim is that the normative source of the evidential ought has nothing to do with the fulfilment of our epistemic purposes. But what is their positive claim exactly? What is the normative source of the evidential ought according to intrinsicalism? This paper aims at improving our understanding of intrinsicalism by ruling out a natural answer that one might be tempted to give to this question. 4 Instrumentalism faces the objection (see e.g. Côté Bouchard 2015; Kelly 2003; Shah 2003; Owens 2003) that epistemic normativity does not seem to be hypothetical in the way it suggests. The fact that the temperature of the room has dropped is an epistemic reason to believe that there has been a transfer of air between outside and inside, and this is the case even if the believer does not aim at achieving knowledge about whether there has been a transfer of air. For various replies see e.g. Cowie 2014; Leite 2007; Steglich-Petersen 2009, 2018. 5 The vast majority of evidentialists are also upholders of a kind of intrinsicalistm. See, for instance, Adler 2002; Kelly 2003; Shah 2003, 2006; Engel 2004; Parfit 2011; Fassio 2011; Whiting 2013. 3 2. Intrinsic and instrumental rightness Bearing in mind that intrinsicalism and instrumentalism are considered to be mutually exclusive positions, a natural and immediate way of conceiving intrinsicalism is by invoking the classical opposition between intrinsic and instrumental rightness. The Natural Conception of Intrinsicalism Intrinsicalism, as the name suggests, is the view that believing in accordance with one's evidence is intrinsically right (as opposed to instrumentally right). According to this natural conception, the meaning of "there is a brute epistemic value in believing in accordance with one's evidence" is thus simply that believing in accordance with one's evidence is intrinsically right. Now, what do philosophers—mainly ethicists—normally mean when they say that a thing is intrinsically right? An intrinsically right thing is right “in itself” in the sense that it derives its rightness from what it is and not from the rightness of something else. Typically, an intrinsically right thing does not derive its rightness from the rightness of its consequences. A pleasant experience is a classical—even though debated—example of an intrinsically right experience. A pleasant experience—some philosophers think—draws its rightness from what it is. In contrast, an instrumentally right thing is a thing that derives its rightness from the rightness of its consequences. For instance, punishing your dog is an instrumentally right action if it has some benefits, perhaps, on its social behaviour. The question whether such or such a thing—pleasant experience, punishment, etc.—is instrumentally or intrinsically right is one that frequently recurs in the domain of ethics. Getting back to the disagreement between instrumentalism and intrinsicalism, given what I have just said is the most natural conception of intrinsicalism, the bone of contention between the two views is thus the following: While instrumentalism takes my believing in accordance with my evidence to be 4 instrumentally right, that is, to derive its normative force from its consequences, intrinsicalism takes my believing in accordance with my evidence to be intrinsically right, that is, to derive its normative force from what it is (and not from its consequences). According to intrinsicalism, to believe things in accordance with the evidence we have is right in itself. According to this way of looking at things, the disagreement between instrumentalism and intrinsicalism constitutes a specific instantiation—in the field of beliefs—of the classical question whether a determinate entity is instrumentally or intrinsically right. Just as ethicists disagree about whether, say, a pleasant experience is instrumentally or intrinsically right, the debate between instrumentalists and intrinsicalists revolves around the question whether my believing in accordance with my evidence is instrumentally or intrinsically right. If this were what the disagreement between intrinsicalism and instrumentalism amounted to, then we’d be on familiar territory. But this natural understanding of what constitutes the bone of contention between instrumentalism and intrinsicalism is incorrect. The antagonism between instrumentalism and intrinsicalism does not come down to a disagreement as to whether it is instrumentally or intrinsically right to believe things in accordance with the evidence. And this is because the fact that a belief is held in accordance with the evidence does not make beliefs intrinsically right. The "kind of rightness" that the fact that a belief is held in accordance with the evidence provides for this belief cannot be identified with the intrinsic rightness that is traditionally opposed to instrumental rightness. Or so, at least, I shall argue in Sections 3 and 3.1. Before this, I would like to address a subsidiary worry. The reader might wonder why I have moved from my original question that concerned the ground of the evidential ought to a discussion about the instrumental or intrinsic rightness of my believing in accordance with my evidence. As the quotations above make clear, the disagreement between instrumentalism and intrinsicalism cuts across various normative notions that apply to beliefs: ought, value, rightness, etc. Nothing prevents me from formulating my original question—viz. the one that pertains to the normative 5 ground of the evidential ought—in terms of other normative notions. Rather than asking in virtue of what one ought one to believe in accordance with the evidence, I could have asked: In virtue of what is my belief epistemically valuable when my belief is in accordance with my evidence? Or In virtue of what is my believing (epistemically6) right when I believe in accordance with my evidence? I will henceforth, for sake of simplicity, express the question that divides the instrumentalists and the intrinsicalists in these latter terms, that is, in the terms of what makes it right to hold beliefs in accordance with the evidence. But this does not reflect any commitment regarding, for instance, the primacy of the normative notion of rightness over other normative notions. 3. Intrinsic rightness transfers from actions to purposes We have seen that the immediate and natural way of conceiving of intrinsicalism is by relying on the traditional ethical distinction between instrumental and intrinsic rightness and taking intrinsicalism to be the view that believing in accordance with one's evidence is intrinsically right, that is, right in itself. What I would like to show now is that this cannot be—on pain of being defeated—the intrinsicalist view. Indeed, the view that believing in accordance with one's evidence is intrinsically right does not make sense. To understand why this view is untenable I need, first, to recall a previously mentioned distinction. There are instrumentally right and intrinsically right actions. Suppose that having a healthy diet is intrinsically right. The action of buying your food at the organic market is, under this supposition, instrumentally right in that it promotes a healthy diet. Suppose now that having pleasure is another intrinsically right end and that jogging is pleasant. The action of jogging is under this supposition 6 For sake of brevity, I won’t use the qualifier “epistemic” below, except on one occasion where it is is required to differentiate epistemic from practical rightness. 6 an intrinsically right action in that it is an action that achieves by itself the end of having pleasure. It does not help achieve this end. It achieves it. The following argument also relies on the following hardly deniable claim: we often act for the sake of something we want or some purpose we have.7 Now, there are two different ways of acting for the sake of something we want or some purpose we have. The first way is when the action performed directly achieves our purpose. When this is so, we act for the sake of acting in this way and not for the sake of some other purpose we have. For instance, if I like jogging and want to go for a jog this morning, when I go for a jog this morning, I jog for the sake of jogging and not for the sake of some other purposes I would like to achieve by jogging (e.g. health). The second way to act for the sake of a purpose is when the action performed does not directly achieve the agent's purpose but is an action that is a means to achieving this purpose (at least in the agent's eyes). If I go to the organic market for the sake of having a healthy diet, I do not perform this action for the sake of performing it but for the sake of some other purpose I have, namely having a healthy diet. Now, the following relation holds between, on one hand, the intrinsic and instrumental rightness of our actions and, on the other, the rightness of the purposes we have when we act: When an action is intrinsically right, what makes this action intrinsically right is one and the same thing as what would make the agent's purpose a right purpose (supposing the agent has such a purpose).8 For instance, the fact that makes jogging intrinsically right—the fact that it is pleasant for me— also makes it right to want to jog. Why should I want to jog? Because it is pleasant for me. In contrast, when an action is instrumentally right, what makes it instrumentally right is not one and 7 For an enlightening discussion regarding how the purposes for the sake of which we act motivate, see Alvarez 2010, 95–97. 8 See Alvarez 2010, 107 for the claim that intrinsic rightness transfers from the purposes to the reasons for which we act. 7 the same thing as what would make the agent's purpose by acting in this way a right purpose (supposing the agent has such a purpose). What makes the action of buying food at the organic market instrumentally right—the fact that it is a means to having a healthy diet—does not also make it right to want to have a healthy diet. Why should I want to have a healthy diet? The following answer: "because buying food at the organic market is a means to achieving a healthy diet" clearly does not make sense. Put as a slogan: Intrinsic rightness transfers from actions to purposes while instrumental rightness does not. This is not a surprising result. Purposes for the sake of which we act cannot, by definition, be right means to ends since they are themselves the ends. But this result is, as we shall see now, revealing when one moves from actions to beliefs. 3.1 Believing in accordance with the evidence: no transfer of rightness Recall that the point I am trying to make is that the view that believing in accordance with one's evidence is intrinsically right is not tenable. Now, the reason why it is not tenable is that the rightness of my believing in accordance with my evidence does not transfer to any plausible purposes for the sake of which I might believe things in accordance with my evidence. What are the plausible purposes or ends for the sake for which a subject might believe things in accordance with the evidence? Here are some possibilities: • the believer might want to get the truth and only the truth and believe things in accordance with the evidence for the sake of this purpose; • she might want to know and believe things in accordance with the evidence for the sake of this purpose; • she might want to get the truth regarding a certain topic and believe things in accordance with the evidence for the sake of this purpose; • she might want to know whether p and believe things in accordance with the evidence 8 concerning p for the sake of this purpose; • etc. Even though the fact that such or such a belief is held in accordance with the evidence makes it right to hold it, this fact does not make any of the aforementioned potential purposes right too. The fact that the belief that p is held in accordance with the evidence concerning p makes it right to believe that p but it does not also make wanting to know whether p a right purpose to have. The rightness of this purpose has nothing to do with the fact that the belief that p is held in accordance with the evidence concerning p. For instance, what makes it right to want to hold a true belief about whether the weather is going to be beautiful next weekend might be the fact that true beliefs are valuable states or the fact that a true belief about the weather will help you planning your hiking tour. But the fact that your belief about the weather is (supposedly) held in accordance with the evidence does not make your wanting to hold a true belief about the weather a right purpose to have. The evidential support of the belief that p is a right making property of your belief that p but it is not a right-making property of any of the purposes listed above. Searching for the truth about p or aiming at knowledge about whether p is not right in virtue of the (potential) evidential support of the belief that p. Note that the previous enumeration of our epistemic purposes does not include the “want to obtain the belief that p”. I consider whether the fact that the belief that p is held in accordance with the evidence makes it right to want to obtain the belief that p. In this paragraph, to repeat, the point that I have tried to make is the following: the rightness that characterizes beliefs held in accordance with the evidence does not transfer to the various purposes for the sake of which a subject might end up believing things in accordance with the evidence. But if the fact that the belief is held in accordance with the evidence is what makes beliefs intrinsically right—as suggested by the hypothesis under scrutiny—the same fact should also make the aforementioned potential purposes right. If the fact that the belief is held in accordance with the evidence is what makes the belief intrinsically right, rightness should transfer from the belief to the 9 purposes because intrinsic rightness—as opposed to instrumental rightness—does transfer in this way (as the previous section taught us). Thus, the kind of rightness provided by the fact that a belief is held in accordance with the evidence cannot be identified with the intrinsic rightness that is traditionally opposed to instrumental rightness. The reader might object that it is not this "classical" intrinsic rightness that most of the intrinsicalists have in mind when they claim that "there is a brute epistemic rightness in believing in accordance with one's evidence". What many intrinsicalists mean by this is that believing in accordance with the evidence is required by the nature of beliefs.9 The view that the ground of the evidential ought lies in what beliefs essentially are is often called "constitutivism".10 I discuss this objection in the conclusion. Before this, let me address a less serious worry. The previous argument relies on the following reasoning: Since what makes it intrinsically right to perform a certain action F—e.g. the fact that it pleasant—makes it intrinsically right to want to F, analogously, what makes it intrinsically right to hold the belief that p—the fact that it is held in accordance with the evidence—should also make one or the other of the aforementioned epistemic purposes (getting the truth, knowledge, etc.) intrinsically right. The reader might object that the analogy between actions and beliefs is not precise enough. In the case of actions, what makes it intrinsically right to F makes it right to want to F. There is a strict correspondence between the purpose (I want to F) and the action itself (the action of F-ing). But this correspondence is absent in the case of beliefs since the epistemic purposes that I am considering do not include wanting to believe that p. A more precise formulation of the analogy is the following: 9 See Velleman 2000; Shah 2003; Wedgwood 2007. 10 See Cowie 2014, 4010; Cowie and Greenberg 2018; Steglich-Petersen 2018, 5; 10 Since what makes it intrinsically right to jog—the fact that it pleasant—makes it intrinsically right to want to jog, analogously, what makes it intrinsically epistemically right to hold the belief that p—the fact that it is held in accordance with the evidence— should make it epistemically intrinsically right to want to believe that p. Is this “should” satisfy? Does the fact that the belief that p is held in accordance with the evidence also makes it epistemically right to want to believe that p? A positive answer to this question seems self-contradictory. The fact that the belief that p is held in accordance with the evidence cannot make wanting to believe that p epistemically right since wanting to hold a specific belief —wanting to believe that p—is not right from this point of view. The formulation is self-contradictory because aiming at holding a specific belief is precisely not aiming at acquiring a belief that is evidentially supported. When you aim at holding a specific belief you are ready to ignore the evidential support that speaks against the belief in question. Note that things are different with a belief’s practical rightness. Suppose pleasantness makes not only action but also certain beliefs intrinsically practically right and that the belief that p is pleasant. What makes it intrinsically practically right to believe that p—the fact that it is pleasant—also makes it intrinsically practically right to want to believe that p. Quite unsurprisingly, beliefs' intrinsic practical rightness transfers in exactly the same way as actions' intrinsic rightness transfers. 4. Conclusion One might think that, just as ethicists disagree about whether, say, a pleasant experience is instrumentally right or instrinsically right, the debate between instrumentalists and intrinsicalists comes down to a disagreement about whether the satisfaction of the evidential ought is instrumentally or intrinsically right. The purpose of this paper has been to show that this is not the case. Intrinsicalism cannot be—on pain of being defeated—the view that believing in accordance with one's evidence is intrinsically right in the classical sense of the term, that is, the sense of intrinsic rightness on which it is opposed to instrumental rightness. If "there is a brute epistemic rightness 11 in believing in accordance with one's evidence", as intrinsicalists hold, this brute epistemic rightness is not identical with intrinsic rightness. As I have already said, the reader might object that intrinsic rightness is not what most intrinsicalists have in mind. Their claim is rather the following: beliefs are essentially such that they require that one ought to believe things in accordance with one's evidence. Beliefs are mental states such that one ought to hold them in accordance with the evidence. There are thus two ways of being an intrinsicalist. The first kind of intrinsicalism is the one that I have discussed at length and criticized in this paper. Here I shall call it "intrinsic intrinsicalism" in order to distinguish it from the second kind. Intrinsic intrinsicalism, to reiterate, states that believing in accordance with the evidence is intrinsically right, that is, right in itself just as a pleasant experience can be right in itself. I shall dub the second kind of intrinsicalism "constitutive intrinsicalism". According to constitutive intrinsicalism, believing in accordance with the evidence is constitutively right, that is, right in virtue of what beliefs essentially are. How does intrinsic intrinsicalism differ from constitutive intrinsicalism? In contrast to constitutive intrinsicalism, the view criticized in this paper does not imply that the evidential ought derives its normative force from what beliefs essentially are. Intrinsic and constitutive rightness are independent kinds of rightness as made clear by the fact a property can make a thing intrinsically right without making it constitutively right. Suppose, once again, that pleasantness makes phenomenal experiences intrinsically right. Pleasantness is not also something that make phenomenal experiences constitutively right. Experiences are not mental states such that they ought to be pleasant. Nothing in what experiences essentially are requires that they be pleasant. Unpleasant experiences are perfectly proper experiences that are intrinsically wrong. Analogously it is perfectly possible to claim—as intrinsic intrinsicalism does—that beliefs held in accordance with the evidence are intrinsically right without claiming that they are constitutively right. Intrinsic 12 intrinsicalism is thus clearly distinct from constitutive intrinsicalism and the criticism presented in Sections 3 and 3.1 targets only the former view. Constitutive intrinsicalism is left untouched by my objection. By clarifying why intrinsic intrinsicalism is not a tenable view, the present article has shown that there is only one possible way of being an intrinsicalist. Intrinsicalists need to be constitutivists. References Adler J. 2002. Belief’s Own Ethics, Cambridge/London, MIT Press. Alvarez, M. 2010, Kinds of Reasons An Essay in the Philosophy of Action, Oxford, Oxford University Press. Conee E. and Feldman R. 2004. Evidentialism, Oxford University Press. Côté-Bouchard Ch. 2015. “Epistemic Instrumentalism and the Too Few Reasons Objection”, International Journal of Philosophical Studies 23, 337–55. Cowie Ch. 2014. “In defence of instrumentalism about epistemic normativity”, Synthese 191:16, 4003–4017. Cowie C. and Greenberg A. 2018. “Constitutivism About Epistemic Normativity”. In C. Kyriacou, R. McKenna (eds.), Metaepistemology, Palgrave Innovations in Philosophy, Palgrave Macmillan,173– 196. Engel P. 2004. “Truth and the Aim of Belief”. In D. Gillies (ed.), Laws and Models in Science, London, King’s College Press, 77–97. Fassio D. 2011. “Belief, Correctness and Normativity”, Logique et Analyse 54, 471. Kelly Th. 2002. “The Rationality of Belief and Other Propositional Attitudes,” Philosophical Studies 110:2, 163–96. Kelly Th. 2003. “Epistemic Rationality as Instrumental Rationality: A Critique”, Philosophy and Phenomenological Research 66:3, 612–640. Leite A. 2007. “Epistemic Instrumentalism and Reasons for Belief: A Reply to Tom Kelly’s”, Philosophy and Phenomenological Research 75:2, 456–464. Nolfi K. 2018. “Why Evidence (and Only Evidence) Can Justify Belief”. In C. McHugh, J. Way and D. Whiting (eds.), Normativity: Epistemic and Practical, Oxford University Press. Owens D. 2003. “Does Belief Have an Aim?”, Philosophical Studies 115, 283–305 Parfit D. 2011. On What Matters, vol. 1, Oxford University Press. Shah N. 2003. “How Truth Governs Deliberation”, Philosophical Review 112, 447–482. Shah N. 2006. “A New Argument for Evidentialism”, The Philosophical Quarterly 56:225, 481–498. Sharadin N. 2018. “Epistemic instrumentalism and the reason to believe in accord with the evidence”, Synthese 195, 3791–3809. Skorupski J. 2011. The Domain of Reasons, Oxford University Press. 13 Steglich-Petersen A. 2006a. “No Norm needed: On the aim of belief”, Philosophical Quarterly 56, 225, 499–516. Steglich-Petersen A. 2006b. “Voluntarism and Transparent Deliberation”, South African Journal of Philosophy 25: 2, 171–176. Steglich-Petersen A. 2009. “Weighing the aim of belief”, Philosophical Studies 145: 3, 395–405. Steglich-Petersen A. 2011. “How to be a teleologist about epistemic reasons”. In A. SteglichPetersen and A. Reisner (eds.), Reasons for Belief, Cambridge University Press, 13–33. Steglich-Petersen A. 2013. “Transparency, Doxastic Norms, and the Aim of Belief”, Teorema 32, 59–74. Steglich-Petersen A. 2018. “Epistemic Instrumentalism, Permissibility, and Reasons for Belief”. In C. McHugh, J. Way and D. Whiting (eds.), Normativity: Epistemic and Practical, Oxford University Press. Vahid H. 2010. “Rationalizing beliefs: evidential vs. pragmatic reasons”, Synthese 176, 447–462. Velleman D. 2000. “On the aim of belief”. In D. Velleman (ed.), The Possibility of practical reason, Oxford University Press. Way J. 2016. “Two Arguments for Evidentialism”, The Philosophical Quarterly 66:265, 805–818. Wedgwood R. 2007. The nature of normativity, Oxford University Press. Whiting D. 2013. “Truth: the Aim and Norm of Belief”, Teorema 32, 121–36. 14